This thread was automatically split due to performance issues. You can find the rest of the thread here: http://reefcentral.com/forums/showthread.php?postid=13635654#post13635654
DIY Red Dragon Pumps, gathering ideas.
- Thread starter JohnL
- Start date
hahnmeister
In Memoriam
<a href=showthread.php?s=&postid=13634715#post13634715 target=_blank>Originally posted</a> by alpine
Ok so how did this thread go from diy red dragons to maxijets volutes?
Also I thought the sole purpose of creating smaller bubbles,was for more surface area and better foam production.Now Hahn is saying that making bubbles to small can cause problems?
IIRC I could swear before the dc red dragons came out that Klaus had posted in a thread saying the pumps would be capable of 10k rpm.
Now you guys are saying that spinning the pump to fast can cause problems.I dont see how rpm is relevant in this thread since the pumps we are suppose to be talking about do not spin that fast.When will the shananigans end and we all get back to the original thread?
Bean & Hahn,while your debates are very informative.You guys need to realize when your debates have become a whos right and whos wrong discussion,and just call it a draw.After a while you two really start to derail some threads.
Sorry for the infighting. Bean always likes to correct people though and say whats right and wrong, and I just find it annoying (as well as the lack of any corrective information) so I throw it right back as useless jibber-jabber. To me, it seems he is more interested in telling the person they are wrong than actually contributing anything of value (very dismissive), and that burns my gord.
As for DC pumps and skimmers... its just my opinion, but something I will add to my response to YF just before this one is that once again 'its all relative'... the actual speed of the OD of the impeller is relative to its diameter... so perhaps tossing around rpm values is a bad idea, since a 6500rpm pump with a 1" diameter pinwheel is going to hit 28.3 ft/s at the OD, and a 3" diameter pinwheel going 1800 rpm is only 23.55 ft/s at the OD. So talking about rpm's itself is the error... as its all relative to the size of the pump. Smaller impellers would benefit from a higher rpm, where larger pumps are going to hit a wall with rpm's that is much lower. I think everyone can agree to that.
My own 'rpm tinkering' is with a smaller pump. I see larger DC pumps are more of a liability than advantage... it depends on the waveform of the power going to the pump, but one of the major advantages of a DC pump is that the torque is constant (if you have a square wave input). The problem is that then you are using a higher RMS wattage (RMS power = peak power with many DC pumps), and so you might have awesome startup power, but you are burning up wattage. So then you turn back to a sinewave input like regular AC power, and gain back the efficiency. But most DC motor controls wont let you startup with square wave, and then switch over to sinewave for operation... not to mention, you just nullified a major reason for going with DC then (all thats left is higher rpm's... but past a certain diameter * rpm, you see a decline in efficiency anyways).
So IMO, the best application for DC pumps is something small. I have been tinkering with that 6500rpm DC motor that is only 24v/48watts max with promising results... but my impeller diameter is only about 1.25"... so at 6500rpm, its OD is only at 35.4 ft/second. But on a 5000rpm motor with a 3" diameter impeller... thats 65.4 ft/s at the outside of the impeller! So Im sure that factors into why you aren't going to see any huge DC motors pulling down massive rpm's in skimmer applications. I would submit that its not the rpm's, but a figure of rpm's * impeller diameter that is what we should be comparing. And thats cool because you can get some great quality small Ti shaft pumps made in Hong Kong on the cheap these days...
alpine
New member
That sounds like some good info there Hahn I will keep in the memory banks for future use.I have thought about picking up a dc pump off ebay to play with.
Tarzan my last post was not directed at you in anyway.Also the pump I posted seems to be that missing link between the 2905 and the 4200. Unfortuneately its not made by laguna,so theres no telling how well it will serve our purposes.Also this thread is about diy red dragons,and if it looks like one thats fine with me.
Tarzan my last post was not directed at you in anyway.Also the pump I posted seems to be that missing link between the 2905 and the 4200. Unfortuneately its not made by laguna,so theres no telling how well it will serve our purposes.Also this thread is about diy red dragons,and if it looks like one
thats fine with me.
Young Frankenstein
New member
hahner.....I have a proposition...........
I chose you as my personal adviser on the next NW super turbo design..........do you accept ?
I chose you as my personal adviser on the next NW super turbo design..........do you accept ?
hahnmeister
In Memoriam
Lol... sure... why not. But what are you going for? My designs are pretty much out there already...
the 'next step' would be getting volutes moulded since there is only so much you can do with a CNC lathe even...
the 'next step' would be getting volutes moulded since there is only so much you can do with a CNC lathe even...
Young Frankenstein
New member
:thumbsup: I can doit, I can make a model out of plaster or a mold with my dremel, and get the material from IPS to do something that was not done before, give me some idea.
I have mentioned large diameter impellers for years, for their tip speed on a low rpm motor.
Centrifugal chiller compresors use this principle. Very large impeller (36"), low motor speed. Smallish impeller (12"), reduction gear to 18000rpm.
The problem, as you mentioned is the drag on the motor.
I was thinking of a thin impeller disc, very few pins right at the center inlet then very few or no pins until the pins at the edge of the impeller. I know this flat, pinless surface will still cause drag (ala Tesla turbine), but it should let the motor turn the impeller.
On my long list of ideas to try in my spare time
Centrifugal chiller compresors use this principle. Very large impeller (36"), low motor speed. Smallish impeller (12"), reduction gear to 18000rpm.
The problem, as you mentioned is the drag on the motor.
I was thinking of a thin impeller disc, very few pins right at the center inlet then very few or no pins until the pins at the edge of the impeller. I know this flat, pinless surface will still cause drag (ala Tesla turbine), but it should let the motor turn the impeller.
On my long list of ideas to try in my spare time
Young Frankenstein
New member
Thats very interesting..........could you put this in paper and show us a diagram........nothing fancy a Barr napkin will do.<a href=showthread.php?s=&postid=13641163#post13641163 target=_blank>Originally posted</a> by H20ENG
I have mentioned large diameter impellers for years, for their tip speed on a low rpm motor.
Centrifugal chiller compresors use this principle. Very large impeller (36"), low motor speed. Smallish impeller (12"), reduction gear to 18000rpm.
The problem, as you mentioned is the drag on the motor.
I was thinking of a thin impeller disc, very few pins right at the center inlet then very few or no pins until the pins at the edge of the impeller. I know this flat, pinless surface will still cause drag (ala Tesla turbine), but it should let the motor turn the impeller.
On my long list of ideas to try in my spare time
hahnmeister
In Memoriam
On the downside to this, the better skimmer pumps seem to benefit from oversized volutes, while leaving the impeller diameter about the same (more flow bias, but because you are moving more water/air the pump's PF seems to go back up to where it might be as a water-onlly pump).
Making the impeller larger in diameter doesn't seem to do much though, past a certain diameter that is about 1/2 the diameter of the volute. I know that centrifugal force SHOULD mean that the water gets pressed to the outside... smaller bubbles as well, and that larger bubbles and less water would flow in the areas under less centrfugal force... but if these pumps are well designed for flow, they are going to be less like a centrifuge where water cant flow outwards... they so it seems this isnt a very valid consideration (with good enough flow, the pump shouldnt develop 'layers' of different air/water ratios in the mixture like say... a maxwel demon)... I know, a bunch of rambling blah blah. But I am not sure why making the impellers larger than a given diameter in comparison to the volute diameter seems to do nothing... as if the impeller past a certain point is just 'spinning its wheels'. All I can come up with is that since needlewheels have their 'pressure bias kicked out from under them' from the venturi on the inlet and the air in the volute... having the impeller larger than a certain diameter does nothing except possibly improve the pressure bias (pressure biased pumps tend to have larger diameter impellers compared to the volute... impellers almost as large as the volute itself). But with needlewheels, making the impellers larger in diameter, past a certain point, just takes up volume in the volute that could otherwise ge going to volume for air... so you actually start to see a performance hit.
Bottom line: Even though these needlewheels are rather 'pressure limited' by design, there must be a ratio of impeller diameter to impeller well diameter... where a larger diameter impeller will rob the the pump of its air to water ratio, but on a taller skimmer body, this would happen anyways if the pump cant generate enough pressure to suck in air at the venturi. This is pretty much 'duh' for those who know about pumps, but with a larger diameter impeller, you are likely pushing the inside diameter of the volute outwards to be rather large. Now, this means that you should increase the volute depth as well (putting a 1" outlet on a volute that is 6" in diameter is very restrictive, so you make the pump outlet larger... say 2", but now you also have to increase the volute thickness as well as diameter... and now you have opened the pandoras box on pump design. There are limits to everything, and using larger volutes and impellers means there are other things to consider as well... design considerations that depend on these parameters. Just like water pumps, there are 'ideals' where you cant just make up for rpm's with peripheral velocity... thats why you actually have that 'golden range' of 2000-4000rpm for water pumps, not a 'peripheral velocity' ideal... because the way that the impellers relate to the volute diameter is one thing, but also the volute thickness, pressure/flow, etc. Im just interested in finding what the 'golden range' for these '2-phase mixing pumps' is since you have air in the water buggering everything up. With a larger diameter, lower speed pump, you are likely to have less pressure handling because of the internal friction that the volute has on the water. A 6" diameter volute vs. a 3" diameter volute means the water has 2x the distance to cover as it spins around the pump, encountering drag from the viscosity of the water on the volute. The water has to travel 2x as far from the inlet as well... encountering drag from the face of the volute on the way unless you have an enclosed needlewheel impeller (I think only H&S has that). So there is likely an ideal diameter for volute, impeller, and ideal rpm... its not just something you can exchange one for another with... trade off 50% of your rpm's for 50% more impeller diameter. These 'ideals' also vary depending on the back pressure that the skimmer puts on the pump. Simply put, the idea pump for a 2' tall skimmer is not the ideal for a 4' one... not just with the venturi diameter, but also the rpms, impeller diameter, etc.
Making the impeller larger in diameter doesn't seem to do much though, past a certain diameter that is about 1/2 the diameter of the volute. I know that centrifugal force SHOULD mean that the water gets pressed to the outside... smaller bubbles as well, and that larger bubbles and less water would flow in the areas under less centrfugal force... but if these pumps are well designed for flow, they are going to be less like a centrifuge where water cant flow outwards... they so it seems this isnt a very valid consideration (with good enough flow, the pump shouldnt develop 'layers' of different air/water ratios in the mixture like say... a maxwel demon)... I know, a bunch of rambling blah blah. But I am not sure why making the impellers larger than a given diameter in comparison to the volute diameter seems to do nothing... as if the impeller past a certain point is just 'spinning its wheels'. All I can come up with is that since needlewheels have their 'pressure bias kicked out from under them' from the venturi on the inlet and the air in the volute... having the impeller larger than a certain diameter does nothing except possibly improve the pressure bias (pressure biased pumps tend to have larger diameter impellers compared to the volute... impellers almost as large as the volute itself). But with needlewheels, making the impellers larger in diameter, past a certain point, just takes up volume in the volute that could otherwise ge going to volume for air... so you actually start to see a performance hit.
Bottom line: Even though these needlewheels are rather 'pressure limited' by design, there must be a ratio of impeller diameter to impeller well diameter... where a larger diameter impeller will rob the the pump of its air to water ratio, but on a taller skimmer body, this would happen anyways if the pump cant generate enough pressure to suck in air at the venturi. This is pretty much 'duh' for those who know about pumps, but with a larger diameter impeller, you are likely pushing the inside diameter of the volute outwards to be rather large. Now, this means that you should increase the volute depth as well (putting a 1" outlet on a volute that is 6" in diameter is very restrictive, so you make the pump outlet larger... say 2", but now you also have to increase the volute thickness as well as diameter... and now you have opened the pandoras box on pump design. There are limits to everything, and using larger volutes and impellers means there are other things to consider as well... design considerations that depend on these parameters. Just like water pumps, there are 'ideals' where you cant just make up for rpm's with peripheral velocity... thats why you actually have that 'golden range' of 2000-4000rpm for water pumps, not a 'peripheral velocity' ideal... because the way that the impellers relate to the volute diameter is one thing, but also the volute thickness, pressure/flow, etc. Im just interested in finding what the 'golden range' for these '2-phase mixing pumps' is since you have air in the water buggering everything up. With a larger diameter, lower speed pump, you are likely to have less pressure handling because of the internal friction that the volute has on the water. A 6" diameter volute vs. a 3" diameter volute means the water has 2x the distance to cover as it spins around the pump, encountering drag from the viscosity of the water on the volute. The water has to travel 2x as far from the inlet as well... encountering drag from the face of the volute on the way unless you have an enclosed needlewheel impeller (I think only H&S has that). So there is likely an ideal diameter for volute, impeller, and ideal rpm... its not just something you can exchange one for another with... trade off 50% of your rpm's for 50% more impeller diameter. These 'ideals' also vary depending on the back pressure that the skimmer puts on the pump. Simply put, the idea pump for a 2' tall skimmer is not the ideal for a 4' one... not just with the venturi diameter, but also the rpms, impeller diameter, etc.
Young Frankenstein
New member
Hahner what are your thoughts on this pin design ? it made my pump work more efficient.
hahnmeister
In Memoriam
On a side note: I have been working on something with some EE/motor guys, maybe you guys have some thoughts...
Most DC pumps are the same as AC pumps, its just the controller thats added to the DC that varies the frequency and waveform(oversimplification, but the basic idea).
So how about making a 'ballast' for an AC skimmer pump?
Why?
Well... when they start, they are full of water, yet when they operate, they suck in air. With some of these 'large diameter volutes' on 'flow biased' pumps, the impeller diameter is too large to start... not just too heavy to start spinning, but their 'interface' encounters too much drag in pure water... so you get sputtering of the pumps at startup. If you make the impeller smaller for startup, then when it gets air in the volute for normal operation, the impeller 'slips' ahead and the power factor/efficiency of the pump takes a dive... possibly killing the pump, or at least making it run hotter because its 'electrical end' wants to do more work than the 'wet end' will allow.
This is why I came up with the idea of just making the volutes larger in diameter/thickness and keeping the impeller diameter small. The impeller shouldn't have more work to do at startup because its size is kept close to original, but after startup, the impeller is doing more work because you have modified the 'wet end' of the pump to move more water than when its just a water pump. So in this way, you can try to keep the power factor/efficiency of the pump closer to its original spec while running as a needlewheel, but not make it so challenged that it cant start on its own. You can kill a good pump simply by leaving it sputter for an hour or so...
But there are other ways to help the impeller start and then transition to normal operation...
You could use an impeller with a 'built in clutch'... many smaller pumps have this... the impeller is allowed to spin a certain number of degrees before it engages the impeller... the energy required to accelerate the magnet is seperate from the impeller head, and then the momentum from the mass of the magnet helps start the impeller head. The amount of rotation varies... some only spin 90 degrees, but many spin almost a full 360... say 180-270 degrees. This is something that is very important to preserve when modifying the Laguna 900 (AC110) and Aquaclear 70/802 pumps.
The other method is to increase the startup torque by using a stronger magnet, or one that is larger diameter. This decreases the space between the impeller magnet and the stators/windings in the motor... allowing the pump more torque like an external/air driven pump where the magnets can be just a hair's thickness away from the stators as they spin. This also tends to impede the flow of water around the magnet for cooling, so things like bypass lines (anti-lime loop), back vanes, and balancing holes get used to increase the flow around the magnet.
But I came up with another way... a pump 'ballast'. By changing the waveform of the electricity going to the pump from the wall current, much like a DC pump, we can give it an ideal 'startup' phase, and an ideal 'operation' phase. This is different from a PWM or DC controller though because it is only changing the incoming power during the initial startup... say... 1-5 seconds after the power comes on... then this 'power bypass' switches over to the regular wall current for normal operation. The impeller could be then designed for ideal efficiency at normal operation without being limited by its startup 'sputtering' considerations.
The downside... one is not for all. Each motor, even of similar wattage, is very different so each model of pump requires a different method of starting. So a 'starter ballast' for an eheim 1262 will most likely not work for a Laguna 1500... even if both are the same wattage.
The upside... one box/pump combo could run on any skimmer without as much regard for the pressures and design of the 'wet end' of the pump... the impeller and volute could be swapped out to change the pressure/flow ratio and within reason... startup in water would be somewhat similar... or rather 'similar enough' it seems...
The downside... many pumps have circuits like these built in, or circuits built in for other reasons like PFC (Lagunas). This complicates things ALOT. You could overload the PFC circuit, you could trigger a fuse/protection, etc. Very simple pumps are actually a bonus.
The first idea was to boost the voltage for more torque. The problems should be obvious, but if kept short enough on a 'simple' pump, it does work. If the pump has a fuse or other circuits in it... well... ouch. Without changing the resistance of the pump, the current can damage it so the startup has to be quick.
The second idea was to change the waveform (sort of like boosting the voltage, but safer). This is alot safer... the goal is to change the incoming power from a sinusoidal wave to a square wave... the main advantage of a DC pump for startup torque (yet its largest drawback as far as operating power). Still messing with that. It requires some larger caps to store power for the startup cycle. The problem is that without using lots of filters ir IC's, the pump's augmentation of the waveform makes it very hard to make a waveform that is square enough... so things get expensive fast because of the controls.
Then the other idea was to change the frequency at startup. This doesn't require power storage, but the controls can be tricky... need timing circuits... so $$$. The waveform stays sinusoidal... ist just alot slower to start. Starting with something like 5hz.... then going up to 60 and switching over to wall current... the problem is that depending on how long (with respect to rpm's-hz) it takes for air to be drawn into the pump, you might avoid sputter at startup just to have it start sputtering when it hits, say... 20hz because even though momentum is on its side, the impeller is falling behind due to drag because its not sucking in enough air yet. I did this on a Sicce PSK and beyond a certain water depth... the pump wouldn't hit enough rpm's to create enough suction to start sucking in air before it started sputtering out. I could be wrong about my conclusions based on that result though... no way to measure the drag inside the volute though...
Measuring impeller 'slip' is not easy... I would have to have some sort of sensor on the impeller so I could track its position with respect to the waveform... and that in itself is a project.
Im about to 'throw in the towel' on this... keep hitting walls because I lack the funding/resources for the equipment needed. Im starting to think that a 'double clutch' impeller head might be easier (one clutch engages the impeller up to half its radius, and then another one engages the outer half after more spinning... but that gets even more complicated).
Variable geometry impeller heads anyone? Rubber pins on the pinwheel disk could be faced inwards during startup to reduce drag (low profile) and startup torque (keep the weight closer to center) and then as the impeller starts and gains speed the rubber pins would flip out under centrifugal force, increasing their profile, drag, etc, and becoming a larger needlewheel in the process. You could also have the disk itself expand as it starts. The problem remains though that the impeller is going to try to start faster than the time it takes for the air to enter the pump... so it will likely still sputter. Although, if combined with some 'clutch' system... hmmm.... I suppose using rubber pins might still work on their own... the pins would have to be just right... strong enough to not flick out just under sputter, but actually need the flow of water (drag) to make them flip up and bend outwards. That would be a huge advance... have to call up the materials engineers for that one.
Could increase the mass of the impeller magnet somehow and add a clutch to the impeller head... the impeller would require about the same torque at startup to just start turning the magnet (the impeller head not engaged, but mass it higher), and then that heavier magnet's momentum would help start the impeller head when the clutch engages it.
I suppose the best solution is to just have a linear air pump on the skimmer pump so that the pump never has to start without air in the volute.
This all assumes that the needlewheel has TOO MUCH drag on it at startup underwater compared to theregular impeller vanes. The opposite could be true as well... the impeller could be that it has too little drag at startup until the air gets in there and thats what makes it sputter... although most of the time, we snip pins off a needlewheel if there is a startup problem... so I assume this isnt the case.
So what do you guys think? projects?
Stealing an idea from auto engines... a turbocharger/supercharger on the pump to help force air in...???
A turbo on the pump output could have be turbine that is sealed yet connected to a higher speed (gearing) blower on the other side that forces more air into the pump.
If an external pump, the shaft could be extended out the backside of the motor (more like a supercharger) and geared up to a blower that pushes more air into the pump... the motor would be pushing water and air directly.
Pumps that can run in series when starting, yet a manifold switches them to run in parallel when running. This way, they could help each other start (not likely if they are both sputtering) because two pumps in series boost each other's pressure but keep the flow), and then two pumps in parallel... well... 2x the flow. Maybe a simpler version of this would be to have a manifold on the needlewheel pump's intake so that a regular water pump could be fed into the pump at startup, then switched over to recirculating flow for normal operation... 'force feeding' (sort of priming, but you are priming the venturi more than the needlewheel pump itself) the needlewheel at startup. The pressure pump for startup could be nothing more than the system sump's return pump output being diverted with a valve actuator... enabled with some sensor relay on the needlewheel pump's power to sense when its power drops on the needlewheel but there is still power in the circuit (so a power out or some other reason to cause the sputter would trigger the valve to flip over and prime the skimmer pump but only after there is power available to the skimmer pump so you dont end up without flow to the tank for some reason).
So.... pick a project guys... Im just the idea guy (you should see the skimmer skunkworks... but thats area 51 right now... bot some stuff that will blow your gord, yet make you go 'gee, thats so simple its stupid').
Most DC pumps are the same as AC pumps, its just the controller thats added to the DC that varies the frequency and waveform(oversimplification, but the basic idea).
So how about making a 'ballast' for an AC skimmer pump?
Why?
Well... when they start, they are full of water, yet when they operate, they suck in air. With some of these 'large diameter volutes' on 'flow biased' pumps, the impeller diameter is too large to start... not just too heavy to start spinning, but their 'interface' encounters too much drag in pure water... so you get sputtering of the pumps at startup. If you make the impeller smaller for startup, then when it gets air in the volute for normal operation, the impeller 'slips' ahead and the power factor/efficiency of the pump takes a dive... possibly killing the pump, or at least making it run hotter because its 'electrical end' wants to do more work than the 'wet end' will allow.
This is why I came up with the idea of just making the volutes larger in diameter/thickness and keeping the impeller diameter small. The impeller shouldn't have more work to do at startup because its size is kept close to original, but after startup, the impeller is doing more work because you have modified the 'wet end' of the pump to move more water than when its just a water pump. So in this way, you can try to keep the power factor/efficiency of the pump closer to its original spec while running as a needlewheel, but not make it so challenged that it cant start on its own. You can kill a good pump simply by leaving it sputter for an hour or so...
But there are other ways to help the impeller start and then transition to normal operation...
You could use an impeller with a 'built in clutch'... many smaller pumps have this... the impeller is allowed to spin a certain number of degrees before it engages the impeller... the energy required to accelerate the magnet is seperate from the impeller head, and then the momentum from the mass of the magnet helps start the impeller head. The amount of rotation varies... some only spin 90 degrees, but many spin almost a full 360... say 180-270 degrees. This is something that is very important to preserve when modifying the Laguna 900 (AC110) and Aquaclear 70/802 pumps.
The other method is to increase the startup torque by using a stronger magnet, or one that is larger diameter. This decreases the space between the impeller magnet and the stators/windings in the motor... allowing the pump more torque like an external/air driven pump where the magnets can be just a hair's thickness away from the stators as they spin. This also tends to impede the flow of water around the magnet for cooling, so things like bypass lines (anti-lime loop), back vanes, and balancing holes get used to increase the flow around the magnet.
But I came up with another way... a pump 'ballast'. By changing the waveform of the electricity going to the pump from the wall current, much like a DC pump, we can give it an ideal 'startup' phase, and an ideal 'operation' phase. This is different from a PWM or DC controller though because it is only changing the incoming power during the initial startup... say... 1-5 seconds after the power comes on... then this 'power bypass' switches over to the regular wall current for normal operation. The impeller could be then designed for ideal efficiency at normal operation without being limited by its startup 'sputtering' considerations.
The downside... one is not for all. Each motor, even of similar wattage, is very different so each model of pump requires a different method of starting. So a 'starter ballast' for an eheim 1262 will most likely not work for a Laguna 1500... even if both are the same wattage.
The upside... one box/pump combo could run on any skimmer without as much regard for the pressures and design of the 'wet end' of the pump... the impeller and volute could be swapped out to change the pressure/flow ratio and within reason... startup in water would be somewhat similar... or rather 'similar enough' it seems...
The downside... many pumps have circuits like these built in, or circuits built in for other reasons like PFC (Lagunas). This complicates things ALOT. You could overload the PFC circuit, you could trigger a fuse/protection, etc. Very simple pumps are actually a bonus.
The first idea was to boost the voltage for more torque. The problems should be obvious, but if kept short enough on a 'simple' pump, it does work. If the pump has a fuse or other circuits in it... well... ouch. Without changing the resistance of the pump, the current can damage it so the startup has to be quick.
The second idea was to change the waveform (sort of like boosting the voltage, but safer). This is alot safer... the goal is to change the incoming power from a sinusoidal wave to a square wave... the main advantage of a DC pump for startup torque (yet its largest drawback as far as operating power). Still messing with that. It requires some larger caps to store power for the startup cycle. The problem is that without using lots of filters ir IC's, the pump's augmentation of the waveform makes it very hard to make a waveform that is square enough... so things get expensive fast because of the controls.
Then the other idea was to change the frequency at startup. This doesn't require power storage, but the controls can be tricky... need timing circuits... so $$$. The waveform stays sinusoidal... ist just alot slower to start. Starting with something like 5hz.... then going up to 60 and switching over to wall current... the problem is that depending on how long (with respect to rpm's-hz) it takes for air to be drawn into the pump, you might avoid sputter at startup just to have it start sputtering when it hits, say... 20hz because even though momentum is on its side, the impeller is falling behind due to drag because its not sucking in enough air yet. I did this on a Sicce PSK and beyond a certain water depth... the pump wouldn't hit enough rpm's to create enough suction to start sucking in air before it started sputtering out. I could be wrong about my conclusions based on that result though... no way to measure the drag inside the volute though...
Measuring impeller 'slip' is not easy... I would have to have some sort of sensor on the impeller so I could track its position with respect to the waveform... and that in itself is a project.
Im about to 'throw in the towel' on this... keep hitting walls because I lack the funding/resources for the equipment needed. Im starting to think that a 'double clutch' impeller head might be easier (one clutch engages the impeller up to half its radius, and then another one engages the outer half after more spinning... but that gets even more complicated).
Variable geometry impeller heads anyone? Rubber pins on the pinwheel disk could be faced inwards during startup to reduce drag (low profile) and startup torque (keep the weight closer to center) and then as the impeller starts and gains speed the rubber pins would flip out under centrifugal force, increasing their profile, drag, etc, and becoming a larger needlewheel in the process. You could also have the disk itself expand as it starts. The problem remains though that the impeller is going to try to start faster than the time it takes for the air to enter the pump... so it will likely still sputter. Although, if combined with some 'clutch' system... hmmm.... I suppose using rubber pins might still work on their own... the pins would have to be just right... strong enough to not flick out just under sputter, but actually need the flow of water (drag) to make them flip up and bend outwards. That would be a huge advance... have to call up the materials engineers for that one.
Could increase the mass of the impeller magnet somehow and add a clutch to the impeller head... the impeller would require about the same torque at startup to just start turning the magnet (the impeller head not engaged, but mass it higher), and then that heavier magnet's momentum would help start the impeller head when the clutch engages it.
I suppose the best solution is to just have a linear air pump on the skimmer pump so that the pump never has to start without air in the volute.
This all assumes that the needlewheel has TOO MUCH drag on it at startup underwater compared to theregular impeller vanes. The opposite could be true as well... the impeller could be that it has too little drag at startup until the air gets in there and thats what makes it sputter... although most of the time, we snip pins off a needlewheel if there is a startup problem... so I assume this isnt the case.
So what do you guys think? projects?
Stealing an idea from auto engines... a turbocharger/supercharger on the pump to help force air in...???
A turbo on the pump output could have be turbine that is sealed yet connected to a higher speed (gearing) blower on the other side that forces more air into the pump.
If an external pump, the shaft could be extended out the backside of the motor (more like a supercharger) and geared up to a blower that pushes more air into the pump... the motor would be pushing water and air directly.
Pumps that can run in series when starting, yet a manifold switches them to run in parallel when running. This way, they could help each other start (not likely if they are both sputtering) because two pumps in series boost each other's pressure but keep the flow), and then two pumps in parallel... well... 2x the flow. Maybe a simpler version of this would be to have a manifold on the needlewheel pump's intake so that a regular water pump could be fed into the pump at startup, then switched over to recirculating flow for normal operation... 'force feeding' (sort of priming, but you are priming the venturi more than the needlewheel pump itself) the needlewheel at startup. The pressure pump for startup could be nothing more than the system sump's return pump output being diverted with a valve actuator... enabled with some sensor relay on the needlewheel pump's power to sense when its power drops on the needlewheel but there is still power in the circuit (so a power out or some other reason to cause the sputter would trigger the valve to flip over and prime the skimmer pump but only after there is power available to the skimmer pump so you dont end up without flow to the tank for some reason).
So.... pick a project guys... Im just the idea guy (you should see the skimmer skunkworks... but thats area 51 right now... bot some stuff that will blow your gord, yet make you go 'gee, thats so simple its stupid').
hahnmeister
In Memoriam
<a href=showthread.php?s=&postid=13644848#post13644848 target=_blank>Originally posted</a> by Young Frankenstein
Hahner what are your thoughts on this pin design ? it made my pump work more efficient.
Shapes and sizes... you can mix them up, play around, etc. By making the pins lower profile, you decreased the force they put on the water (which may or may not be a good thing for the pump... it just depends on where you are at). The smaller profile does 'cut' throught the air easier though... alot like needlewheels vs. meshwheels (or needlewheels with finer, yet denser pins which would be the same thing). I would hope that by using 'thinner pins' you also increase the number though. A decrease in wattage might seem like a good thing, but if you are dropping the PF at the same time you can just as easily damage the motor (heat buildup). Thats why I always tell people who intend to mod needlewheel pumps to get an air meter and kill-a-watt... keep that PF up or else you could have big problems!
Some large pin impellers dont rely on their contact with the water so much as the space between the pins to 'blend'. Rather than using pins that are smaller, finer (mesh) and denser, you can also look at the spacing between the pins as another dimension of similar importance. You could have larger pins, but with very narrow spaces between them...
I can imagine there are several possible ways to fiddle with the pins as far as spacing, thickness, as well as shape. The possibilities are endless... square pins, diamond, round, triangular, hemispherical (and then what orientation), vanes, vanes with holes drilled in them, vanes with slots cut into the leading edge (like a comb/pinwheel at the top, but regular vane at the base... I think CoralVue uses this for their Laguna 1500 impellers if I remember correctly). You could use star shapes or 'cup' shapes on the leading edge (crescent moons and horseshoes)... Hearts, stars and horseshoes! Clovers and blue moons! Pots of gold and rainbows! And me red balloons! Its me lucky charms pinwheel design!!!
hahnmeister
In Memoriam
^^^ cant link to image, so copy and paste it to view in browser.
One limitation of a CNC lathe made volute is that the ID of the volute is constant. This doesn't provide the best flow for a unidirectional pump. Of note though... having a volute that recirculates SOME might not be so bad because it allows the bubbles to spend more time along the volute wall smashing into each other... blending more than what the impeller alone does.
If you look at a stock Laguna impeller well cover, it is also shallower in thickness as well where its diameter (assume the inlet is over the impeller and the impeller axis is the center for these measurements... I know the center of curvature is actually not the same) is the smallest... and then by the time the water gets to the outlet, the thickness is increased along with the diameter. This also prevents water from just 'staying in the loop' and spinning over and over.
Now, this could work against us since these pumps are mixing as well... so maybe passing water too quickly isnt the best, but without some attention to this, you see a loss in efficiency because on the volute wall on either side of the outlet (volute throat and cutwater) there is a likelyhood for turbulence to cause seperation of flow in these places. With a constant radius volute, the water wants to go out the outlet, but this creates a drop in pressure right after the outlet, so some of the water recirculates and this causes a seperation of flow as well.
Adding stationary diffuser vanes in the volute is one way to get around this... even if you just use one to seperate the flow before and after the volute throat.
If you start to encounter vibrations (Klaus and some others have experienced this) a 'double volute' is another consideration. They are intended to reduce vibration problems... this could also help minimize seperation of flow inside the volute by spreading out the cutwater and throat areas.
But these things require an injection moulded volute (or 6 axis CNC perhaps and then some polishing afterwards).
BeanAnimal
Premium Member
No, I like to point out when people are expounding in make believe science. You are certainly correct, even by your own admission you do post a lot of useless jibber-jabber.
What burns your gord is being called out when your slinging tall scientific tales. I will very kindly say it again. Anytime somebody responds to one of your verbose and wandering scientific extrapolations, it only begets more of the same. A 2 sentance rebuttal triggers a 2 paragrah response. A 2 paragraph rebuttal triggers a 2 page response. Lets be dead honest here. I have never seen you stop and admit you are wrong, even after being kindly and clearly shown to be wrong. Instead you keep thundering away at the keyboard with more fervor and suspect explanations. The closest you ever come to admission of error is "well then we are both right" or a similar quip.
Yet you tend to toss around RPM numbers all the time. You did in your last post in this thread
I guess that depends on your understanding of DC motors and drives. For that matter, what types of AC and DC motors are you comparing here Jon? Universal? DC brushless? PSC? Induction?
Brushless DC motors can run at unity power factor, a term you love to throw around when talking about skimemr pumps. I can certianly make a side by side list between the different motor types if it will help show other pros and cons.
Sure about that, Jon? An AC VFD on an induction motor can be set to slip phase at lgiht loads (low RPM, low load, etc) to be super efficient. A lightly loaded DC motor wastes power. The exact opposite is true at higher load conditions, the AC motor suffers and the DC motor excels.
You mention torque... An AC PSC induction motor produces gobs of startup toque. A universal DC motor also produces a lot of startup torque. On the other hand a brushless DC motor has very low relative startup torque.
I think if you revisit the truth about the differences between AC and DC motors and drives, you will change your opinion. The benefits of a DC motor vs an uncontrolled AC motor are numerous. To control an AC motor to gain back those benefits is expensive and complicated compared to the drive needs of a brushless DC motor.
hahnmeister
In Memoriam
True... I suppose my 'field of vision' is limited by what I see actually out on the market... nothing too complicated. Most DC motors are going to be brushless. You are building on what I mentioned... pretty much parallel though.
It would be nice to see AC VFD's... but Im not looking to make a $1000 skimmer motor either. Such a cost undermines the potential. The VFD is pretty much what we were making anyways (well, seeing if it would work using a scope and wave generators)... or more like a V.V.F.D. (periods just so the V's dont look like a W >> VV). The PSC motor is what we were basing our idea for the Sicce on, adding a cap in series for startup. Using ACTUAL PSC motors though... less $$$ than VFD, but still $$$.
Maybe I should have mentioned it, but those ideas you are putting out there are pretty much what we were trying to do... trying to give some of those attributes to otherwise inexpensive/mag-drive motors.
True though, with a DC motor you CAN make a large one more efficient than an AC motor. With an expensive enough controller, you can start pulsing/notching the waveform... cutting the RMS power to be lower than sinusoidal. Your wallet must be pretty fat though.
Perhaps I should assign 'efficiency' a dollar value here... saving $50 in electricity a year doesn't justify a $1000 controller IMO.
As for the rpm's... its all a matter of context. There are some ideals as I pointed out later... simply increasing impeller diameter isnt always the best way to increase the velocity/pressure at the parimeter. I know you like to nitpick Bean, but in the context of each response, I qualified the conditions.
"Bean, correct me if Im wrong..."
next post: "lol, like you have to ask"
Posting jibber jabber is the point of a forum... were not posting instruction guides or something... these are blogs. If you want to post jibber jabber... be my guest. Just thinking out loud and posting it... is that also WRONG!? In which case, isnt your response nothing more than MORE USELESS JIBBER JABBER???
"I have never seen you stop and admit you are wrong, even after being kindly and clearly shown to be wrong."
Sure I have. How about you? Still not willing to admit that your 3-pipe overflow could be replaved by a 2-pipe one and perform the same? The only time I remember an 'oops, Im sorry' was when I took off the bottom plastic trim on a 40B to show you that the side glass does sit on the bottom... but I suppose you had no choice. How about that PVC pipe? Pot calling kettle black... just a little?
Your 2 sentence posts get 2 PAGE responses not because of their content, but rather their lack of it, and the fact that your wording is usually rude. You toss out very little info but seem to relish in telling others that they are wrong more than actually educating them as to why or contributing. Okay, know-it-all... you gonna help the conversation progress past that, or are you that socially challenged? Man, you just dont know how to talk to people... maybe just online... I dont know... but the tone of your writing is just... arrogant & snobbish, and dismissive... not just towards the content that others write, but to the person who posts it as well. I dont know if you are trying to deter people with a less-than-expert level of understanding from even posting, if you just dont see that, or if you just like talking down to people on the internet, but I can promise you that Im not going to take that kind of treatment. I dont care if you are right or wrong... nobody wants to say sorry to someone who continues to behave like you.
A contoured asperating venturi might allow for slightly higher water flow through the pump with less loss of pressure... even with the higher air intake undermining those gains in the volute itself. Where a fed venturi might gain alot from nice contours on the venturi, a needlewheel one just doesn't see the same gains. Sure, there might be some, but there are sooo many other things you can do that have a larger impact anyways... not to mention... the easier/quicker the air&water mix passes out the pump, the less blending time it has. So who is to say it is better? If you have equations for skimmer pump design, please share... otherwise, its still more design than engineering (art and engineering share alot).
You tend to jump on people even when they are right, just using a different means of explaining something than you would. That doesn't make them 'WRONG' or mean they are using 'pseudo science'. Your last comments on VFDs, PSC motors, etc... pretty much show that. Same thing... just different means of talking about it. You suggest that I lack understanding because I chose to use an analogy that was very simplistic... about water in a centrifuge without a means of escape along its path without moving side to side against a flat walled volute... like trying to move up and down in a cajun-cliffhanger ride. Then you blast it because I didnt mention the venturi/bernoulli effects... so someone who may not even know what that means...??? Gee... I dont know why I would use such a simple example... maybe because thats what other people do...
Its hard to explain things that use complex understandings in terms that someone without a background in the subject can understand... so excuse me for the 'jibber jabber', but Im not aiming my content at you. At least you could TRY to explain things to people rather than telling them they are wrong followed up with "If you understood _____ then you would understand why"
Gee, no... really?
It would be nice to see AC VFD's... but Im not looking to make a $1000 skimmer motor either. Such a cost undermines the potential. The VFD is pretty much what we were making anyways (well, seeing if it would work using a scope and wave generators)... or more like a V.V.F.D. (periods just so the V's dont look like a W >> VV). The PSC motor is what we were basing our idea for the Sicce on, adding a cap in series for startup. Using ACTUAL PSC motors though... less $$$ than VFD, but still $$$.
Maybe I should have mentioned it, but those ideas you are putting out there are pretty much what we were trying to do... trying to give some of those attributes to otherwise inexpensive/mag-drive motors.
True though, with a DC motor you CAN make a large one more efficient than an AC motor. With an expensive enough controller, you can start pulsing/notching the waveform... cutting the RMS power to be lower than sinusoidal. Your wallet must be pretty fat though.
Perhaps I should assign 'efficiency' a dollar value here... saving $50 in electricity a year doesn't justify a $1000 controller IMO.
As for the rpm's... its all a matter of context. There are some ideals as I pointed out later... simply increasing impeller diameter isnt always the best way to increase the velocity/pressure at the parimeter. I know you like to nitpick Bean, but in the context of each response, I qualified the conditions.
"Bean, correct me if Im wrong..."
next post: "lol, like you have to ask"
Posting jibber jabber is the point of a forum... were not posting instruction guides or something... these are blogs. If you want to post jibber jabber... be my guest. Just thinking out loud and posting it... is that also WRONG!? In which case, isnt your response nothing more than MORE USELESS JIBBER JABBER???
"I have never seen you stop and admit you are wrong, even after being kindly and clearly shown to be wrong."
Sure I have. How about you? Still not willing to admit that your 3-pipe overflow could be replaved by a 2-pipe one and perform the same? The only time I remember an 'oops, Im sorry' was when I took off the bottom plastic trim on a 40B to show you that the side glass does sit on the bottom... but I suppose you had no choice. How about that PVC pipe? Pot calling kettle black... just a little?
Your 2 sentence posts get 2 PAGE responses not because of their content, but rather their lack of it, and the fact that your wording is usually rude. You toss out very little info but seem to relish in telling others that they are wrong more than actually educating them as to why or contributing. Okay, know-it-all... you gonna help the conversation progress past that, or are you that socially challenged? Man, you just dont know how to talk to people... maybe just online... I dont know... but the tone of your writing is just... arrogant & snobbish, and dismissive... not just towards the content that others write, but to the person who posts it as well. I dont know if you are trying to deter people with a less-than-expert level of understanding from even posting, if you just dont see that, or if you just like talking down to people on the internet, but I can promise you that Im not going to take that kind of treatment. I dont care if you are right or wrong... nobody wants to say sorry to someone who continues to behave like you.
A contoured asperating venturi might allow for slightly higher water flow through the pump with less loss of pressure... even with the higher air intake undermining those gains in the volute itself. Where a fed venturi might gain alot from nice contours on the venturi, a needlewheel one just doesn't see the same gains. Sure, there might be some, but there are sooo many other things you can do that have a larger impact anyways... not to mention... the easier/quicker the air&water mix passes out the pump, the less blending time it has. So who is to say it is better? If you have equations for skimmer pump design, please share... otherwise, its still more design than engineering (art and engineering share alot).
You tend to jump on people even when they are right, just using a different means of explaining something than you would. That doesn't make them 'WRONG' or mean they are using 'pseudo science'. Your last comments on VFDs, PSC motors, etc... pretty much show that. Same thing... just different means of talking about it. You suggest that I lack understanding because I chose to use an analogy that was very simplistic... about water in a centrifuge without a means of escape along its path without moving side to side against a flat walled volute... like trying to move up and down in a cajun-cliffhanger ride. Then you blast it because I didnt mention the venturi/bernoulli effects... so someone who may not even know what that means...??? Gee... I dont know why I would use such a simple example... maybe because thats what other people do...
Its hard to explain things that use complex understandings in terms that someone without a background in the subject can understand... so excuse me for the 'jibber jabber', but Im not aiming my content at you. At least you could TRY to explain things to people rather than telling them they are wrong followed up with "If you understood _____ then you would understand why"
Gee, no... really?
FirstContact
Active member
Small question on needle wheel pumps from a novice:
Is it possible to attach a sedra nw pump to my berlin classic? Not sure if the pump will allow a hose attachment from the online pics. High end skimmer is out of my budget right now.
What size sedra nw pump would be a good candidate?
Many kind thanks in advance,
FirstContact
Is it possible to attach a sedra nw pump to my berlin classic? Not sure if the pump will allow a hose attachment from the online pics. High end skimmer is out of my budget right now.
What size sedra nw pump would be a good candidate?
Many kind thanks in advance,
FirstContact
chadfarmer
Premium Member
are all these pumps made to be used in saltwater
Young Frankenstein
New member
I agree .........Some times he behaves like that....but deep inside he is a nice person.......just stuck up and never admit he is wrong, ready to step and squash anyone that made a mistake or call them a niwik. Well nitwick you too bean, there you go. If you are that good why dont you show us a video of your nitwick skimmer....... ?<a href=showthread.php?s=&postid=13646139#post13646139 target=_blank>Originally posted</a> by hahnmeister
True... I suppose my 'field of vision' is limited by what I see actually out on the market... nothing too complicated. Most DC motors are going to be brushless. You are building on what I mentioned... pretty much parallel though.
It would be nice to see AC VFD's... but Im not looking to make a $1000 skimmer motor either. Such a cost undermines the potential. The VFD is pretty much what we were making anyways (well, seeing if it would work using a scope and wave generators)... or more like a V.V.F.D. (periods just so the V's dont look like a W >> VV). The PSC motor is what we were basing our idea for the Sicce on, adding a cap in series for startup. Using ACTUAL PSC motors though... less $$$ than VFD, but still $$$.
Maybe I should have mentioned it, but those ideas you are putting out there are pretty much what we were trying to do... trying to give some of those attributes to otherwise inexpensive/mag-drive motors.
True though, with a DC motor you CAN make a large one more efficient than an AC motor. With an expensive enough controller, you can start pulsing/notching the waveform... cutting the RMS power to be lower than sinusoidal. Your wallet must be pretty fat though.
Perhaps I should assign 'efficiency' a dollar value here... saving $50 in electricity a year doesn't justify a $1000 controller IMO.
As for the rpm's... its all a matter of context. There are some ideals as I pointed out later... simply increasing impeller diameter isnt always the best way to increase the velocity/pressure at the parimeter. I know you like to nitpick Bean, but in the context of each response, I qualified the conditions.
"Bean, correct me if Im wrong..."
next post: "lol, like you have to ask"
Posting jibber jabber is the point of a forum... were not posting instruction guides or something... these are blogs. If you want to post jibber jabber... be my guest. Just thinking out loud and posting it... is that also WRONG!? In which case, isnt your response nothing more than MORE USELESS JIBBER JABBER???
"I have never seen you stop and admit you are wrong, even after being kindly and clearly shown to be wrong."
Sure I have. How about you? Still not willing to admit that your 3-pipe overflow could be replaved by a 2-pipe one and perform the same? The only time I remember an 'oops, Im sorry' was when I took off the bottom plastic trim on a 40B to show you that the side glass does sit on the bottom... but I suppose you had no choice. How about that PVC pipe? Pot calling kettle black... just a little?
Your 2 sentence posts get 2 PAGE responses not because of their content, but rather their lack of it, and the fact that your wording is usually rude. You toss out very little info but seem to relish in telling others that they are wrong more than actually educating them as to why or contributing. Okay, know-it-all... you gonna help the conversation progress past that, or are you that socially challenged? Man, you just dont know how to talk to people... maybe just online... I dont know... but the tone of your writing is just... arrogant & snobbish, and dismissive... not just towards the content that others write, but to the person who posts it as well. I dont know if you are trying to deter people with a less-than-expert level of understanding from even posting, if you just dont see that, or if you just like talking down to people on the internet, but I can promise you that Im not going to take that kind of treatment. I dont care if you are right or wrong... nobody wants to say sorry to someone who continues to behave like you.
A contoured asperating venturi might allow for slightly higher water flow through the pump with less loss of pressure... even with the higher air intake undermining those gains in the volute itself. Where a fed venturi might gain alot from nice contours on the venturi, a needlewheel one just doesn't see the same gains. Sure, there might be some, but there are sooo many other things you can do that have a larger impact anyways... not to mention... the easier/quicker the air&water mix passes out the pump, the less blending time it has. So who is to say it is better? If you have equations for skimmer pump design, please share... otherwise, its still more design than engineering (art and engineering share alot).
You tend to jump on people even when they are right, just using a different means of explaining something than you would. That doesn't make them 'WRONG' or mean they are using 'pseudo science'. Your last comments on VFDs, PSC motors, etc... pretty much show that. Same thing... just different means of talking about it. You suggest that I lack understanding because I chose to use an analogy that was very simplistic... about water in a centrifuge without a means of escape along its path without moving side to side against a flat walled volute... like trying to move up and down in a cajun-cliffhanger ride. Then you blast it because I didnt mention the venturi/bernoulli effects... so someone who may not even know what that means...??? Gee... I dont know why I would use such a simple example... maybe because thats what other people do...
Its hard to explain things that use complex understandings in terms that someone without a background in the subject can understand... so excuse me for the 'jibber jabber', but Im not aiming my content at you. At least you could TRY to explain things to people rather than telling them they are wrong followed up with "If you understood _____ then you would understand why"
Gee, no... really?
