Author Topic: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!  (Read 10235 times)

0 Members and 1 Guest are viewing this topic.

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Hi all.

I was given by my uncle he DB9 rear light and asked if I can repair it. Given I have repaired a few automotive board, one being my body control ecu on my Alfa, I thought yea ok no worries. It’s only an LED driver board, how hard can it be..!!
My bad..!!!
Pictures are below for reference.
I will try and describe the best I can to help understand the issue.
The top part of the light has two sets of leds. One long set and one short set. The short set AND long set come on when the lights are on. But only had say 50% brightness.
You can see that’s it’s being pulse width modulated when you record it and play it back. Also given the ic’s being used inside but I will go more into that later.
Then when the brake light it pushed while the lights are on the short strip lights up full brightness while leaving the long strip at 50% brightness.

I know this from watching the other working light on his car.

Now there are 5 wires on the back of the controller.
1. Neg
2. Brake
3. Lights
4. Indicators
5. Fog light.

Now I’ve checked the voltage going to the connector when plugged into the car and each input is just seeing a strait 12v. And nothing comes up on the dash as a fault light to say the lights are out. Eliminating the possibility that it’s canbus controlled.

The issue with these lights are when you put 12v to the light on brake nothing happens. Just draws 67ma. If I put 12v on the lights input the brake part lights up dimly and flickers badly. But the other part does not light up as it should.

I have replaced the 4x ti ucc2813d-3

Also replaced 2x st mc33079dt

I have reflowed all components, checked all
Mosfets and diodes, capacitors.

But can not find any dodgy components anywhere. Used the microscope to check the tracks. Even tested most of them for continuity.

Spent over 10 hours on this now. And begining to runout of ideas on what to test not.

I’ve read the data sheet on the ucc2813d-3 which is very full on. And understand it to a degree. Checked all supporting components and their values and all ok. But the ucc2813d-3 will not put out and pwm signal to the mosfets that drive the leds. The data sheet is limited on what to check. This is where I’m up to. Trying to find out why it’s not outputting a pwm to the driver mosfets. Is the ucc2813d-3 in a fault condition? How do I tell?

Any help would be much appreciated.

Cheers. Steve






« Last Edit: July 24, 2019, 07:52:52 pm by cpposteve »
 

Offline tunk

  • Super Contributor
  • ***
  • Posts: 1099
  • Country: no
Re: Aston Martin DB9 rear light LED driver repair
« Reply #1 on: July 16, 2019, 12:50:54 pm »
I have no idea what the problem is, but here's a
couple of things you may already have tried:
- switch the clusters to rule out cable problems.
- switch the PCBs to confirm that it's a PCB problem.
The PCB may also possibly be used in other (cheaper) cars.
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #2 on: July 16, 2019, 10:07:04 pm »
Cheers

Haven’t tried what’s on your list. Swapping the boards would be a right pig of a job. As the lights are glued shut and the other one is working. So not really a good idea but may have to if all else fails.

I will swap the lights over to check cables though.

Cheers
 

Online fzabkar

  • Super Contributor
  • ***
  • Posts: 2805
  • Country: au
Re: Aston Martin DB9 rear light LED driver repair
« Reply #3 on: July 17, 2019, 12:59:10 am »
ISTM that the UCC2813 PWM controllers are operating in open loop mode with the FB pin as the control pin. That is, the chip outputs a PWM signal with a duty cycle determined by the DC voltage on its FB pin. If the voltage at FB is less than REF/2 (Vref = 4V), then the output is switched off.

I would determine which chip is driving the FB pin and work backwards from there.

http://www.ti.com/lit/ds/symlink/ucc2813-0.pdf
https://www.st.com/resource/en/datasheet/mc33079.pdf

I'm assuming that you have used an LM324 as a substitute for the MC33079.
 

Offline Zucca

  • Supporter
  • ****
  • Posts: 4611
  • Country: it
  • EE meid in Itali
Re: Aston Martin DB9 rear light LED driver repair
« Reply #4 on: July 17, 2019, 01:42:13 pm »
your last picture shows a corroded PCB, the one you attached shows an immacutale board, I'm confused.

Did the light suffered any water ingress?

You need to get your hands on the known working board and with a scope point to point check back and forth.

Now I’ve checked the voltage going to the connector when plugged into the car and each input is just seeing a strait 12v.

Did you check with a scope or a DMM in AC?
Can't know what you don't love. St. Augustine
Can't love what you don't know. Zucca
 
The following users thanked this post: Johnny10

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #5 on: July 17, 2019, 06:26:50 pm »
ISTM that the UCC2813 PWM controllers are operating in open loop mode with the FB pin as the control pin. That is, the chip outputs a PWM signal with a duty cycle determined by the DC voltage on its FB pin. If the voltage at FB is less than REF/2 (Vref = 4V), then the output is switched off.

I would determine which chip is driving the FB pin and work backwards from there.

http://www.ti.com/lit/ds/symlink/ucc2813-0.pdf
https://www.st.com/resource/en/datasheet/mc33079.pdf

I'm assuming that you have used an LM324 as a substitute for the MC33079.

hi fzabkar.

thanks for your reply. ive read the data sheet and  you are correct abount the feedback pin. however, ive checked the voltage on that pin and it is at the voltage required. (4v) im going to check tonight with my scope wether there is even a hint of a pwm signal before possibly shutting down. as it states in the data sheet that it can entor into an error state and not put out a pwm signal, but cant work out what can cause this error state apart from there being no feedback voltage.
i used the lm324 as a temp substitute just to see if it would have any effect showing the mc33079 was faulty, but was not the case. although i did order some new ones and replaced them.


your last picture shows a corroded PCB, the one you attached shows an immacutale board, I'm confused.

Did the light suffered any water ingress?

You need to get your hands on the known working board and with a scope point to point check back and forth.

Now I’ve checked the voltage going to the connector when plugged into the car and each input is just seeing a strait 12v.

Did you check with a scope or a DMM in AC?

hi zucca.

where is it corrodded? this board was conformaly coated, so no water has been on it, although the connector was slightly corroded where the leds plug onto the board, only because this was the only part of the board not covered in conformal coating.
If you mean those red looking splodges thats the glue used to hold the components inplace before being wave soldered.

thanks all
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #6 on: July 17, 2019, 06:58:23 pm »
forgot to mention that i didnt check it with my scope no. but i did check it on ac and dc and only dc was present
 

Offline Zucca

  • Supporter
  • ****
  • Posts: 4611
  • Country: it
  • EE meid in Itali
Re: Aston Martin DB9 rear light LED driver repair
« Reply #7 on: July 17, 2019, 08:40:41 pm »
where is it corrodded? this board was conformaly coated, so no water has been on it, although the connector was slightly corroded where the leds plug onto the board, only because this was the only part of the board not covered in conformal coating.
If you mean those red looking splodges thats the glue used to hold the components inplace before being wave soldered.

Ah ok, got it. Well, comparing with the good board it's the easiest thing.
« Last Edit: July 17, 2019, 08:42:12 pm by zucca »
Can't know what you don't love. St. Augustine
Can't love what you don't know. Zucca
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #8 on: July 17, 2019, 09:52:51 pm »
yes agree with you there on testing a board thats working and do a side by side comparison. but its unfortunatly not really an option at the moment. i think my main task is to work out how the 2813d-3 works so i can figure out why its not chucking out a pwm signal to the leds.

steve
 

Online fzabkar

  • Super Contributor
  • ***
  • Posts: 2805
  • Country: au
Re: Aston Martin DB9 rear light LED driver repair
« Reply #9 on: July 18, 2019, 12:04:13 am »
If you are measuring 4V at the FB pin, then I must be wrong about my open loop hypothesis. The 4 transformers (chokes) must then be the inductive elements in 4 switchmode regulators.
 

Offline jh15

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: us
Re: Aston Martin DB9 rear light LED driver repair
« Reply #10 on: July 18, 2019, 02:24:01 am »
Check that all commons, grounds, earth is fully intact.  very important. Many check for power supply voltage, ripple, probe on some ground, but the pcb may have corrosion/opens.
Tek 575 curve trcr top shape, Tek 535, Tek 465. Tek 545 Hickok clone, Tesla Model S,  Ohio Scientific c24P SBC, c-64's from club days, Giant electric bicycle, Rigol stuff, Heathkit AR-15's. Heathkit ET- 3400a trainer&interface. Starlink pizza.
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #11 on: July 18, 2019, 07:06:49 am »
Hi all.

Thanks for all the help.
I scoped the inputs to the mosfets last night to see if I could see switching on briefly then shutting down. Possibly to an error. As I was right. I did a single capture of it and could see it kicking in with a pwm signal but then shuts down. Was measuring 10.8khz but the period it was on for varied. Found it hard to measure as my memory depth on my crappy owon usb scoop is just rubbish.
More poking and probing needed tonight.

As for the grounding points, this was next on my list. Even though I have check many things multiple times. I’ve not gone and checked all ground points yet.

If I could understand how the 2813d-3 works better I think it would help me fault find.

Cheers all
 

Online fzabkar

  • Super Contributor
  • ***
  • Posts: 2805
  • Country: au
Re: Aston Martin DB9 rear light LED driver repair
« Reply #12 on: July 18, 2019, 11:14:20 pm »
ISTM that it would be worth the time to trace the circuit around one of the four outputs.
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #13 on: July 19, 2019, 09:32:34 pm »
well after 6 hours on it tonight i can finally say we have success. luckily there are 4 2803d-3's in there, so checking voltages on each pins is what helped me here to compare. Pin 1 was getting a 2 higher voltage at its pin. tracing it back it is fed from the out put of the quad op-amp. the two inputs are fed from else where, but passing via a resistor, of which one of them was open circuit causing the op-amp to go higher then it  should, fitted a new 0805 resistor and hey presto its fixed. i have spent almost 30 hours on this, and did not want to give up untill i found the issue, and im so glad i didnt stop until i did.

would just like to say many thanks to all that have helped and given advice.

cheers

steve
 
The following users thanked this post: sokoloff, Mario87

Offline BillyD

  • Regular Contributor
  • *
  • Posts: 218
  • Country: ie
Re: Aston Martin DB9 rear light LED driver repair
« Reply #14 on: July 19, 2019, 11:24:10 pm »
It might be a fair bit of your time, but I bet it's still way cheaper than having to buy a new unit!
Well done, was following this one with interest and no doubt someone will google it in future and find it useful.


 
The following users thanked this post: cpposteve

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #15 on: July 20, 2019, 01:52:39 pm »
Yes. But it was sure worth it. It’s not the money that I do it for. It’s the learning along the way that comes with this sort of thing. I’ve learnt a lot about the 2803d-3 and the 33079 op amp. And what voltages should be where. Having these faults are very time consuming and tedious. Especially when your not 100% up on how this stuff works. But it’s sure worth it in the long run.

Cheers.
 

Offline Mario87

  • Regular Contributor
  • *
  • Posts: 249
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #16 on: July 20, 2019, 02:02:08 pm »
Congrats on fixing the fault! I too was also watching this topic with interest.
 
The following users thanked this post: cpposteve

Offline Zucca

  • Supporter
  • ****
  • Posts: 4611
  • Country: it
  • EE meid in Itali
Re: Aston Martin DB9 rear light LED driver repair
« Reply #17 on: July 22, 2019, 01:25:58 pm »
the satisfaction when you tell your buddy:"I got it fixed" is something just reserved for us EE.
And the 30 hours suddely disappears.
Congrats.
« Last Edit: July 22, 2019, 01:42:42 pm by zucca »
Can't know what you don't love. St. Augustine
Can't love what you don't know. Zucca
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair
« Reply #18 on: July 24, 2019, 05:57:33 pm »
the satisfaction when you tell your buddy:"I got it fixed" is something just reserved for us EE.
And the 30 hours suddely disappears.
Congrats.

Couldn’t agree more. It was so worth the long nights I spent on it.

Cheers
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 4136
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!
« Reply #19 on: July 24, 2019, 08:23:39 pm »
It's an awesome feeling and one of the reasons I prefer buying broken stuff.
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!
« Reply #20 on: July 24, 2019, 08:32:01 pm »
that and keeps me getting under the misses feet  :-DD
 

Offline PaulAm

  • Frequent Contributor
  • **
  • Posts: 939
  • Country: us
Re: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!
« Reply #21 on: July 25, 2019, 12:04:07 pm »
Given those lights go for $5-800 USD on ebay (what's dealer cost, over $1K?), even at 30 hours it's not minimum wage  ;D

Nice job
 
The following users thanked this post: cpposteve

Offline AndyC_772

  • Super Contributor
  • ***
  • Posts: 4315
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!
« Reply #22 on: July 25, 2019, 12:27:05 pm »
Don't ask... even dealers try to fix them if water gets in. There's a service bulletin out there which describes where they tended to leak, and the official procedure to vent them and blow any water out.

There's also a company which can completely replace the electronics, which is a much cheaper option than new lights:

http://www.fisherperformance.co.uk/aston-martin-rear-lights/

Offline Mr. Scram

  • Super Contributor
  • ***
  • Posts: 9821
  • Country: 00
  • Display aficionado
Re: Aston Martin DB9 rear light LED driver repair
« Reply #23 on: July 25, 2019, 12:27:31 pm »
the satisfaction when you tell your buddy:"I got it fixed" is something just reserved for us EE.
And the 30 hours suddely disappears.
Congrats.
Not just EE, but pretty awesome nonetheless.
 

Offline cpposteveTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gb
Re: Aston Martin DB9 rear light LED driver repair.......... FIXED!!!
« Reply #24 on: July 25, 2019, 12:37:37 pm »
The lights have a one way vent on them. Alowing air to go out but if you try and put air in it locks off. But clearly doesn’t do it’s job properly as there was signs of condensation in the light unit.

Anyhow, he doesn’t drive it in the rain nor use it in the winter so minimising the risk of another failure. I think most of the damage was done by previous owners either keeping it out side in the rain or using it in all weather.

That said though. I’m adamant this wasn’t a water damage issue. It’s comformaly Coated and no sighn of corrosion anywhere but the plugs where the flat flex plugs into the board. I think this was solely a component failure. Time will tell.

Steve
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf