Meanwell LDD driver: for those who want to dim to 0 using Arduino

[O2Surplus]

You can drop it to 30Hz if you (for some reason ) want to. There are possible values from 30Hz to 62KHz:
http://playground.arduino.cc/Main/TimerPWMCheatsheet

Tim

I've used 10bit @ 150Hz and was able to dim my A6211 and CAT4101 based drivers to 1/1023. The low PWM frequency helps the low duty cycle (1/1023) signal overcome the built in capacitance of the wiring between the controller and the driver.IIRC- it's the capacitance of the wire that tends to "dampen" signals using high frequency PWM. LOL- sort of the same reason that I2C communication is only good for a few feet.

 

[R_Mc]

Agreed there are ways to change it (at least increase the frequency not reduce) but there are other issues with doing that and it probably doesn't work well with Storm or Blue controllers.

Also using the 10bit resolution reduces the PWM output pins from 6 to 4. Again may be fine if you are doing your own controller but it's not going to work with a pre-made controller.

And again the frequency is just the number of times the signal changes state per second. It does not affect the duty cycle of the on/off duration. Nor does the current output effect frequency or duty cycle.

I'm really confused by R_Mc's comment about having a dual dimming driver that does both PWM and current regulated dimming. If the issue is flicker at low duty cycle PWM then his idea would not solve that at all. If you are using PWM to dim at the low end of the dimming then it doesn't matter if the current is 10-15% the flicker is still going to happen. It might be dimmer while it flickers but it's still going to flicker. This is more of an either or thing and really does not suit itself to a combination. If you wanted a driver that had an adjustable current max but was still dimmed viw PWM then that is acceptable but having multiple types of dimming is a bit absurd. Krazie:jester:

It would make the issue much less noticeable as you could be at a higher duty cycle with a lower output current.

So instead of running a 1% duty cycle at 1A, you might run a 5% (or higher) duty cycle at 200ma... I'm not sure how much of a perceptible difference this may make, or if the ratios are linear (likely not) but it was a thought.

Also, how noticeable or bothersome the flicker is is definitely tied to perceived brightness. Imagine a 1,000 watt LED flickering at 200HZ with a very steep on/off curve. Now imagine a 1 watt led flickering. It's not as problematic, and this would be the effect of lowering the current. It would still flicker, but the flicker would be less prominent at the same brightness.

 

[Krazie4Acans]

It would make the issue much less noticeable as you could be at a higher duty cycle with a lower output current.

So instead of running a 1% duty cycle at 1A, you might run a 5% (or higher) duty cycle at 200ma... I'm not sure how much of a perceptible difference this may make, or if the ratios are linear (likely not) but it was a thought.

Also, how noticeable or bothersome the flicker is is definitely tied to perceived brightness. Imagine a 1,000 watt LED flickering at 200HZ with a very steep on/off curve. Now imagine a 1 watt led flickering. It's not as problematic, and this would be the effect of lowering the current. It would still flicker, but the flicker would be less prominent at the same brightness.

I'm not sure how to tell you to fix what you are trying to fix. The LDD (what this thread is about) can't do what you are asking. It can run at higher PWM frequencies that what you state you are running so maybe that is the solution for you? There may also be another thread that has a driver like the one you are thinking would be better. In either case it doesn't sound like the solution you need is in this thread. Krazie:jester:

 

[R_Mc]

I think you are getting a couple of things mixed together here. the bit resolution/ number of values has nothing to do with the PWM frequency. If the PWM frequency is 1khz then it is 1khz. It's not dependant on the PWM resolution. I have never seen an Arduino with a PWM frequency of 240hz (or ish).

Quoted from Arduino web site: "The frequency of the PWM signal on most pins is approximately 490 Hz. On the Uno and similar boards, pins 5 and 6 have a frequency of approximately 980 Hz. Pins 3 and 11 on the Leonardo also run at 980 Hz."

This would mean that the minimum frequency of the PWM from an Arduino would be 480hz. While the human eye may be able to see 60hz-120hz flicker on a florescent bulb and maybe even 240hz at a 10% or less duty cycle, I find it very hard to believe that the human eye can see 490hz at any duty cycle. Nor would the human eye be able to perceive 980hz at all.

Now if you personally don't like the effect of the light that you see, then that is a personal opinion and I would say "then don't use it". However, if you don't like LEDs at those frequencies then I'm not sure how you could possibly like florescents at all. Krazie :jester:

Two different issues

PWM Freqency = flicker
PWM bit resolution = smooth dimming

A higher frequency would help with flicker (given there are capacitance limits / turn on times for the led etc). 10 hz would be more noticeable than 30 hz which is more noticeable than 100hz / 1khz / 10khz

I don't want to derail this from the LDD - so my present question is whether a higher frequency than 1Khz has been tried. With the arduino one click can output ~480hz or ~3700hz (nothing in between). I'm curious if the 3700hz frequency would work with the LDD (also note that I do not have an O-scope... which is why I'm trying to get as much of this hammered out in theory as I can).

 

[MagicJ]

I'm curious if the 3700hz frequency would work with the LDD (also note that I do not have an O-scope... which is why I'm trying to get as much of this hammered out in theory as I can).

According to the spec sheet, the PWM frequency range for PWM dimming is 100 to 1KHz.

 

[wilkerschnepf]

O2, did you ever post the files for the SCW-LDD adapter board? I just finished re-reading all 126 pages and that little adapter would help me get my fixture one step closer to complete.

Also i picked up some fans that claim to be 12V .08A, would you suggest running both from a single SCW05C-12 or should i step up to the 08, or run each fan from its own SCW for redundancy?

 

[Krazie4Acans]

O2, did you ever post the files for the SCW-LDD adapter board? I just finished re-reading all 126 pages and that little adapter would help me get my fixture one step closer to complete.

Also i picked up some fans that claim to be 12V .08A, would you suggest running both from a single SCW05C-12 or should i step up to the 08, or run each fan from its own SCW for redundancy?

I don't remember seeing those files published but they may be on there with his other files to git.

I would use an SCW08C-12. Remember that you want a little over head so that the SCW runs cooler and lasts longer. They get pretty warm running at 100%

 

[zachts]

What is the lowest duty cycle you can get them to?

Off at 0/4095, ever so dim at 1/4095.

 

[wilkerschnepf]

Thanks Krazie, Ill wait to hear from O2, if he is going to share the board design or if maybe i can just purchase one from him. I enjoy building the boards, but don't have the knowledge to design one myself. Once i get confirmation i will get some SCW08C-12 ordered for my build.

 

[perkint]

Surely two fans at 0.08A is only 0.16A and well under what a single SCW05C-12 can give

Tim

 

[wilkerschnepf]

I would think so too, maybe i will break out the multimeter and test what amperage two together are really using. I just hooked it up to see how loud this silent fan was, and i am pretty impressed, i had to hold my ear close enough that the fan hit my cheek before i could really hear it. Of course these china box leds are running and it makes it hard to hear anything in the house.

 

[O2Surplus]

O2, did you ever post the files for the SCW-LDD adapter board? I just finished re-reading all 126 pages and that little adapter would help me get my fixture one step closer to complete.

Also i picked up some fans that claim to be 12V .08A, would you suggest running both from a single SCW05C-12 or should i step up to the 08, or run each fan from its own SCW for redundancy?

I'm posting the Gerber files for my SCW to LDD-H adapter PcB. I spent some time redesigning it a bit to make it easier to configure for different power needs. This version will allow the SCW to provide power output "full time" (assuming the SCW has power on it's input) or "part time" by triggering a transistor that takes the SCW negative output to ground. All that's needed to change it's behavior is blob of solder on the SMD jumper's solder pads. No solder= Full PWM or On/Off control using the PWM pin. Solder= "Full time" power.

This design is panelized to include two adapters per 45mm X 32mm PcB. Use the 50mm X 50mm size option when ordering and you'll get two adapters for the price of one. I recommend using Seeedstudios, since their PcB ordering system is very user friendly.http://www.seeedstudio.com/service/index.php?r=pcb
:beer:

 

[wilkerschnepf]

Awesome! Thanks again O2!

 

[Krazie4Acans]

Surely two fans at 0.08A is only 0.16A and well under what a single SCW05C-12 can give

Tim

You would think so but don't forget that the 0.08ma is the runtime rating and not the startup spike current. Then add two of them together on the same SCW and your probably pushing it a little for the 05. Yes that is a startup spike only but if you can't provide what the fan needs to start up then you don't have to worry about what it takes to run it because it won't run. Going with the 08 will cover the startup spike and during the running time will be much lower than the rated power so it will be nice and cool.

The 05 will probably work, but the 08 will work better and probably for a longer time.

 

[wilkerschnepf]

You would think so but don't forget that the 0.08ma is the runtime rating and not the startup spike current. Then add two of them together on the same SCW and your probably pushing it a little for the 05. Yes that is a startup spike only but if you can't provide what the fan needs to start up then you don't have to worry about what it takes to run it because it won't run. Going with the 08 will cover the startup spike and during the running time will be much lower than the rated power so it will be nice and cool.

The 05 will probably work, but the 08 will work better and probably for a longer time.

I had a chance to do a little test on the two fans. The .08A rating on the package was actually spot on. Both together pull .160-.170A

When first connected the highest in rush amperage i saw was .223A. It was hard to take a picture at the exact instance i connected the power but this was pretty close. I don't think my multimeter has a peak hold option.

I will still order the swc 08 version just in case the fans ever get changed in the future and these aren't available. Looks like mouser is the only place that has the 08 versions available right now.

 

[Hamato]

I've been using a 10 LED strand (cree XT-Es) with a LDD-700HW for years now with no issues. I added a second LED strand today and now neither power up I tested all the LEDs with batteries and they all power up and all my connections are good so I fear I have somehow fried both LDDs. I have tried swapping the LDDs around and still nothing from either strand. No shorts or anything and I measure 48V across the input terminals and about 50mv across the output terminals of the LDD.

I'm using a 48V 1.8A PS and was using PWM but now for testing I have the PWM line unconnected and only powering the original strand of LEDS. Here is what I had wired originally (only 2 strands) but now I have the white line unconnected.

Any help would be appreciated, if I have somehow killed the LDDs I don't want to order more and kill those too!

 

[jedimasterben]

I've been using a 10 LED strand (cree XT-Es) with a LDD-700HW for years now with no issues. I added a second LED strand today and now neither power up I tested all the LEDs with batteries and they all power up and all my connections are good so I fear I have somehow fried both LDDs. I have tried swapping the LDDs around and still nothing from either strand. No shorts or anything and I measure 48V across the input terminals and about 50mv across the output terminals of the LDD.

I'm using a 48V 1.8A PS and was using PWM but now for testing I have the PWM line unconnected and only powering the original strand of LEDS. Here is what I had wired originally (only 2 strands) but now I have the white line unconnected.

Any help would be appreciated, if I have somehow killed the LDDs I don't want to order more and kill those too!

If you have them mounted to an LDD board with pull-down resistors, the drivers shut off with no PWM input.

 

[Hamato]

If you have them mounted to an LDD board with pull-down resistors, the drivers shut off with no PWM input.

Nah they are just the wired versions. I basically wired the second LDD VIN +/- pins in series with the existing LDD and then connected the VOUT +/- contacts to the second string of LEDs.

Since my last post I tried a different power supply and connected them to only 1 LED to eliminate any variables but still no go. I'm probably going to go with the LDD boards to replace these but I really don't know how I killed 2 working LDDs Besides doing something really stupid like wiring them backwards what else can kill these things?

 

[pwreef]

LDD VIN should be in parallel. Both (+) of LDDs to PS (+). Both (-) of LDDs to PS (-).

 

[Hamato]

LDD VIN should be in parallel. Both (+) of LDDs to PS (+). Both (-) of LDDs to PS (-).

Yeah you are correct, I mistyped. They are in parallel, but for some reason I couldn't edit my post here.

Oh well hopefully I'll figure out the problem before the replacements arrive :worried: