DIY LED driver for reef lighting

[terahz]

5) tera; I've NEVER heard of Owon. I have heard of Rigol and have been considering the Rigol DS1102E.

My problem is that I need an isolated scope for at least half of my work. I have a Tektronix that is but the numbskulls at Tek have ignored that genre of scopes now for a decade. Instead they are wasting their time on scopes that I see selling for more than half a million dollars each. :headwallblue:

Yeah, the Rigol is quite popular but the owon is not much more expensive, yet it has a huge screen+resolution and much better hardware from what I read (real 1g/s even at 10M memory). Then there are the Hantek DSO5062B/Tekway DST1062B which can be hacked into 200MHz 1Gs scopes (they run linux and I've even seen pictures of people running xfce on them ). All of these are in the $400-$500 range.

I have old tektronix 2235, or 2213, honestly can't remeber if it is 60 or 100 meg. I don't use it very often. I would love storage scope but just can't justify the cost.

Justifying 400 bucks in this hobby isn't very hard

 

[BeanAnimal]

$400 = 20 cases of cheap beer or 10 cases of craft beer....

Honestly, I don't often find use for a scope often. At one point I had visions of building high power SS and tube amps, and purchased the scope for that reason. I have piles of massive high voltage/current ONSEMI power transistors for a huge sub amp build, but never got around to doing it

My plan here is to use something like the NXP 16 channel driver to drive the PWM inputs on the LM3409 -OR- the NXP PWM output through a simple rc filter or opamp to drive the analog input of the LM3409.

 

[dannytt]

just for diy led driver not worth buy a scope.....

 

[terahz]

You guys are no fun Of course it is not worth it to buy a scope because of an led driver.

Bean, 16 lm3409s can handle e LOT of LEDs, are you changing the lights for the tank and the house?

 

[BeanAnimal]

Nope, just two smallish fixtures

It is more about building something modular and playing around with drivers than about optimizing a solution for a specific outcome. In fact, not all of the channels will be used, likely only 5 per fixture.

As you likely know, I am far from being an LED proponent. The emitters are becoming more efficient and spectral choices becoming more abundant.

I am fairly certain that a large portion of the "DIY" and many of the commercial fixtures are pretty lousy with regard to spectral output and their ability to sustain a broad range of SPS, Clams, etc.

I am however ready to give them a whirl, but only if I can control the spectral output, both to attempt to get acceptable life support AND to satisfy my eye. To date, I seen very few LED tanks that appeal to me with regard to the color of light they cast. I aim to build a fixture that pleases both myself AND the coral

 

[terahz]

I hope you get the color spectrum you're after. I'm sure with enough different LEDs you can, it is just a matter if, is the complexity worth it.

Do you judge the spectrum visually or do you have some nice toys for that?

I'm also after better colors. I've gathered quite a few color LEDs to mix with the usual NW/CW/B/RB combo to see if I can get some better colors.


EDIT: Oh, forgot to say that TI replied to me saying that I should go to the Community Support Forums, so I guess I'll be posting my question there.

 

[BeanAnimal]

I hope you get the color spectrum you're after. I'm sure with enough different LEDs you can, it is just a matter if, is the complexity worth it.

I hope as well. My first inclination was to order a pile of LEDs and experiment with combinations until I found what I wanted and then build the fixture with the ratios and drive currents found. In the end I honestly don't feel like dealing with a drawn out process so I am going to build the spectral adjustment into the fixture. Moreover, this is likely 1 time and like it or dump it adventure. I just don't have the drive (no pun) to fiddle with LEDs when there are so many more interesting things I can find to do with my time and money. Not knocking anybody who finds it interesting, but it is one rabbit hole that I am not that interested in climbing down. I have maybe a dozen (or more) unfinished micro and similar projects laying around that need attention.

Do you judge the spectrum visually or do you have some nice toys for that?

I have some toys (a fairly high end tristimulus colorimeter), but it is not well suited to the task at hand as it is more suited to relating color to the CIE colorspace as opposed to a spectrophotometer that measures true spectral output. That is, with some baseline calibration, it could give a decent idea of the relative spectral output from a known source to an unknown source. In the end, way too much work and a spectrophotometer (or spectroradiometer) is way too expensive.

Short answer, If the coral DOES somehow grow, then tweaking so that my eyes are happy is all that counts

EDIT: Oh, forgot to say that TI replied to me saying that I should go to the Community Support Forums, so I guess I'll be posting my question there.

Typical, instead of getting an answer from an engineer, they pawn you off on a group of nitwits who want to be engineers but can't (Microsoft MVPs and the MS community forums anybody?).

I will call the local Allied rep tomorrow and get him to get us an answer. The guy knows I am a hobbyist, but treats me no different than a volume manufacturing customer. I am using them more and more due to the wonderful customer service.

 

[kcress]

Tera; Definitely let us know which scope you get and what you think of it.

What scope do you have now?

And I can't remember and can't find what question you are positing with TI. What was it?

 

[BeanAnimal]

I think the question was asking for technical clarification of the dimming resolution between PWM and ANALOG... something an engineer should answer, not a self educated nitwit (like most of us) in a community forum.

I do like this LM3409, kinda magical... not bacon magical, but close.

 

[terahz]

Not knocking anybody who finds it interesting, but it is one rabbit hole that I am not that interested in climbing down. I have maybe a dozen (or more) unfinished micro and similar projects laying around that need attention.

I have a beaglebone with 7" touch LCD sitting on my desk, waiting to become my next controller... now that I got my openbeam parts to build my fixture I should be able to finish it and move to one of the many things on my desk .

Short answer, If the coral DOES somehow grow, then tweaking so that my eyes are happy is all that counts

I think you'd have a hard time finding HP LEDs that don't grow coral. Making the eyes happy is a different story

I will call the local Allied rep tomorrow and get him to get us an answer. The guy knows I am a hobbyist, but treats me no different than a volume manufacturing customer. I am using them more and more due to the wonderful customer service.

Great, let us know. I'm also interested to know if I wasted my time with a 12bit dac.

I will try to go to a friend this weekend with a scope an maybe get some photos of the output, just for fun

 

[terahz]

Tera; Definitely let us know which scope you get and what you think of it.

What scope do you have now?

And I can't remember and can't find what question you are positing with TI. What was it?

Sure. Reading about cheap-o DSOs in the last few days still points me to the OWON sds7102 so IF I get one, it is going to be the owon.

It doesn't really qualify as one, but I have a DSO nano .

I think the question was asking for technical clarification of the dimming resolution between PWM and ANALOG... something an engineer should answer, not a self educated nitwit (like most of us) in a community forum.

Yep, I asked them if they can clarify how the analog dimming works and how they got the 250:1 number in the datasheet, and they sent me to the community forums .

 

[AHutch]

Terahz - I should be using the following components:

Roff 7.32K
Coff 470pF
Q1 FQP27P06

The excel spreadsheet i posted in my original post has all of the components that i selected.

EE is not my forte but i borrowed a scope from a friend anyway to play around a bit.

If I'm using it correctly, PGATE/Coff frequency is ~2.5MHz with the DAC at the lowest setting and it looks like a 50% duty cycle. That is, Ton and Toff appear to be at 200ns each.

From the lm3409 spec: "In practice, switching frequencies higher than 1MHz may be difficult to obtain due to gate drive limitations, high input voltage, and thermal considerations."

Is Toff to short requiring a higher switching frequency? Either that or maybe i did a poor job soldering the back side of the lm3409 and it can't dump to the heat into the board.

 

[terahz]

AHutch, the .xls you linked says your max current is 2A, but the XTEs' max current is 1.5A. You might want to put a .151 Rsns instead.

Also, the 8 LEDs at max current will probably put the driver in dropout mode, because they have a Vfw of 27.2V and your psu is 27.65V? At lower currents you should be good though.

If we plug your numbers in the datasheet calculation, you should be at about 450kHz. Try going to 11K resistor (if you don't have one, maybe stack two R7s(30K) with one R8(100K)). That should lower the Fsw to about 300kHz(250kHz) and hopefully reduce the load on the chip.

I still can't explain why it would get hotter at lower currents though.

Also, I think you want to measure Coff and LED+ (Vo) to get the proper Toff.

 

[AHutch]

Terahz, i failed to fill in the average LED current in cell B7 but it doesn't look like it's used in an calculations. The calculated 1058mA which is close to what i am able to hit with 7 XT-E LEDs. Running 8 or 9 does limit the total current to under 1A due to the Vin of 27.65v.

One question: the switching frequency is expected to change as you change the output of the DAC, correct? I assumed the switch frequency in the spreadsheet would be the switching frequency required to limit the total LED current to the desired value. In my case, this is 1A.

Lowering the DAC output voltage to further reduce the LED current will require an increase in the switching frequency, correct?

Thanks again!

 

[terahz]

One question: the switching frequency is expected to change as you change the output of the DAC, correct? I assumed the switch frequency in the spreadsheet would be the switching frequency required to limit the total LED current to the desired value. In my case, this is 1A.

Lowering the DAC output voltage to further reduce the LED current will require an increase in the switching frequency, correct?

(one of the EEs feel free to slap me if what I'm about to say is garbage)
While this is correct, it is a not directly because of the change in current, but the change of forward voltage of the LEDs as a result of the changed current. The switching frequency depends on the input and output voltages of the driver (apart from the roff/coff combo).

In your case, of 7 LEDs, at 100mA, they should have ~ 18.55Vfw. With 27.8Vin you need to have a 3.5K Roff and 470pF Coff to get to 2.5MHz. And my guess would be, at that frequency it is likely that the chip gets on the hot side .

With the same 7 LEDs at 1A, Vfw is ~ 22.05V and that should give you a Fsw of 760kHz with your 7400 Roff.

So again, my recommendation is to get a higher value Roff. A 16.5K Roff, which is what I use for my 1A drivers, should give you a Fsw range of 340kHz-540kHz (high current - low current). That should make the driver happy.

 

[shirley386]

Terahz - thanks for the tip on putting the boards on the stove to remove the chip; worked like a charm.

I've got two "working" drivers now. They drive a string of 14 XTEs that I have been burning as a test. I have them on a 12 hours cycle with a 2 hour ramp up and ramp down. The chip gets very hot, to hot to hold your finger on. I actually thought it would burn up or unsolder itself before it got through a whole cycle. It gets hottest at the lower settings. I set it ramp from 1 to 4095 and same coming down. The lights never do get very dim. Even at the lowest setting that will light them they appear to be about 50 - 60%. Also, at the lowest settings they will begin to flicker. I can hold a pencil on the chip and the flicker stops.

I originally thought I had a short somewhere on the chip but both boards do the same thing. I'm thinking now that I must have a wrong component(s) somewhere. I went by your 1 amp BOM and the power supply is right at 48V. I tried to check the current but I can't get my meter to work on amps.

I guess I'll double check the components and if I can't find the problem I don't know what else to do.

 

[kcress]

Switching loses are the major loss in these setups. The higher the frequency the greater the loses, the hotter the chips get. Anything over 1Mhz should be avoided - like the plague.

 

[terahz]

shirley, have you soldered the bottom of the chip (I assume by chip you mean U1)? It sounds like you maybe haven't, and by pressing with the pencil, you ground the chip properly?

Also, does the chip heat up right away or does it take a while? Can you also take a temperature reading somehow? If it peaks at something less than, say 150F, it is fine to leave alone. 200F is probably fine too, though pushing it. The IC has thermal protection, so it will shut itself off when it hits 320F and go back on at around 290F. I surely hope you're not getting that high!

I'm not certain why this is happening, I haven't put mine in "production" yet, as I'm working on the actual LEDs array, but from the various several minutes of running while I've done testing, I never noticed the lm3409 getting too hot to touch. Admittedly, I've never done 'load' testing on the lowest stetting...

I'll try to replicate what you guys are getting and see what I can find. Manwhile, shirley, you can read up on measuring current with a big resistor with your multimeter in voltage mode. I think even here on reef central kcress has explained it a few times,

 

[terahz]

Switching loses are the major loss in these setups. The higher the frequency the greater the loses, the hotter the chips get. Anything over 1Mhz should be avoided - like the plague.

Hmm, I think this will be a problem for people with long arrays, because the LEDs we use can go down to 2.5Vfw and in shirley's case that would make the total Vo only 35V. On a 48V psu that will be a 800kHz frequency. On the same setup at 1A, the total Vo is 44V and Fsw is only 145kHz.

 

[shirley386]

Yes, I soldered the back of the chip. I measured it with a digital thermometer @ 141. In the process, I allowed the thermometer to short something out and it all went dark. Now I'm down to one driver.