LM3404EVAL: AN-1839 LM402/LM3404 Fast Dimming/ True Constant LED Current Eval Board issues

hi ArTzy

The only change on LM3404 EVM is replacing original LEDs with 2p3s LEDs and other parameters are kept with original values?

I will check the waveforms and see if there is a solution.  

Hello Mason,

Hope all is well.

Gregory, owner of this post is monitoring this thread, asking an update.

Any progress please? We are looking forward to your support.

73,
Archie A.

hello ArTy

I think it is quite challenging for LM3404 to stably realize 30000:1 dimming ratio with good current accuracy. Actually we have a new device that is in sample stage with ultra-deep dimming ratio. I would like to confirm the spec with you:

input 28V, output 2p3s green LED with ~9.35V.  what is the target current rating? what is the preferred dimming method(PWM dimming, analog dimming, hybrid dimming)?

I can share some information with you on this device by mail if you are interested.

Thank you for your response Mason.  I was out of town and just seeing your response today.  I would like more information on the ultra-deep dimming ratio device.  The equipment we are working on is used by navy pilots for landing on the darkest of nights and on a super bright days.  Thats why we need the large dimming ratio.  

We based the design, we are asking questions about, on the AN-1839 LM402/LM3404 Fast Dimming/ True Constant LED Current Eval Board.  Paragraph 12 of the eval board document states a contrast ratio of greater than 50K:1 is possible.  We need at least 30K:1.  We have obtained even higher ratios during some testing. However stability at the low intenstiy has been an issue.  Are you suggesting that we shouldn't use the LM3404 for this project?  We drive some of the LEDs at 350 ma and others at 700ma.  Thanks for your help. 

hello Gregory

thanks for the information. Is the project still in design stage? To realize 30k:1 dimming ratio, do you accept hybrid dimming method(combine analog dimming and PWM dimming) or only PWM dimming?

Hybrid dimming combines analog dimming and PWM dimming, for larger current range, device works in analog dimming while for lower current range it works in PWM dimming. This is beneficial to realize higher dimming ratio and maintain good current accuracy at the same time. 

Mason,

Our design, based on the LM3404 eval board, is a final prototype design.  But we haven't received the boards from the manufacturer yet for testing.  We have been working with first generation prototype boards.  So both our prototype board and when we replace our board with the TI LM3404 Eval board we have issues with low intensity stability.  We may have to make a board modification to eliminate the low intensity instability.

I will research hybrid dimming.  Anything you can send me would be much appreciated.  We did realize, during testing, that when we obtained over a 100k:1 dimming ratio that our PWM signal magnitude was being reduced as we decreased the pulse width.  Guess you could call that hybrid dimming.  

Could you also send information on the ultra-deep dimming device?  However, hopefully you can help us modify our existing design

Thank you

Mason,

This is an explanation of our prototype circuit.  This was submitted in a previous fourm but never resolved.  I include it here just for more information for you.  Using the TI eval board we have been discussing in this forum, in one case, we had a parallel string of LEDs, not a series sting of 6 LEDs as disussed below.

  Working with the following schematic I have several questions.  Thank you for your help.  The below circuit was designed using TI LM3404xx-1-A data sheets SNVS465G and the AN-1839 LM3402/3404 evaluation board users guideSNVA342E.

There are 11 of these driver circuits on each circuit card with additional circuitry.  Some boards are set to drive up to 200ma through the LEDs, other boards 800ma.  Each of the above driver circuits drive a string of 6 LEDs, either red, PC amber, or green.  The LEDs are Cree and the green P/N is XPEBGR-LL1-0000-00F02.  Our goal is a contrast ratio of 25000 to 1 or better.  The PWM signal is applied to the external MOSFET at LAMP DIM on the right end of the above circuit.  We are using a PWM base frequency of 200 to 250 Hz.  Our lowest pulse width is about 150 nano sec with the 250 Hz base rate.  There is constant random flickering especially noticeable at the low intensities.  We have 15 levels of intensity.  The lowest 4 intensity settings are generated by 2 CD 54HCT123s and intensity levels 5 to 15 are generated by 555 timers.  

hello Gregory

I will share the new device information by mail since this is NDA needed. 

for the LM3404 EVM board, has you made any modification on the parameters before obtaining the waveforms in the original attachment word document? 

from the waveform, the LED current is abnormal with quite large current ripple. I would like to know if there is any modification. thanks

Mason,

For the waveform pictures you are looking at the LM3404 Eval Board was unmodified.  Several months ago we put a larger inductor on the a LM3404 eval board but it did not change the problem of intensity flickering at low intensity.  We did not take waveform pictures during that test.

Base frequency was 250Hz with a minimum pulse width of 90 nano sec

Can you mail me the new deep dimming info to my home address or does it have to be the business address?  Home address is more reliable.

Thank you

hello Gregory

It would be better to have business address. If it is possible, pls share with me.

Mason,

My Business address is:

Naval Air Warfare Center 

Route 547

Building 678 Room 126

JBMDL, NJ 08733

Atten Greg Dyczko 732-323 1695

hi Gregory

do you mind sharing you email for information sharing?

Hello Mason,

I don't understand your question.  This post is already public information.  Just type "Hybrid dimming with the LM3404" into Google search and our post will come up.

Looking forward to the information you are sending.  With your suggestion about Hybrid dimming we are making progress with our circuitry 

hi Gregory

I would like to send you our new device information by email online since currently the information of this new device is confidential. Also I would like to confirm that customer has NDA with TI. Maybe I make you confused but I did not see any email available. 

Mason,

My email is gregory.w.dyczko.ctr@us.navy.mil

Thank you for your help

this thread will be further discussed by mail.

Mason, I have not received anything in the mail yet.  Were you able to send anything?

Thank you for your help

Hi Gregory,

I will support you further for new device information. Can you send the NDA request mail to y-tang2@ti.com with your agency name and I will forward it to your local market owner.

Hi Archie,

I sent you a mail regarding NDA for customer. Pls help on this so that we can share the information of our new device.