combination meter repair - DIY

Not sure how you value a person's time on eBay. He may be happy making $10 an hour, which means it's potentially a 10 hr repair

 

Can someone elaborate on this "key FOB trick"?

My combination meter seems like it's starting to go bad, but it's only happened twice this year where it would start without dash lights. So it's not quite an urgent replacement for me yet.

 

You are paying for his expertise and his resources. He does electronics testing/calibration/firmware loading for one of the manufacturers it seems. It is his part tike niche, and $100 is dirt cheap when the other option is a messed up CAN network, setting codes that can mess up your car. Mine stopped turning the reverse lights on, and after a while, I couldn't shut the car off. I payed his fee, and had my cluster back within a week, or you can buy one he has for sale. As he explained it to me, (without giving away too much info) one of the chips starts to glitch out, sending stray codes into the network, and there are some loose solder joints, and bad caps. I don't know how much of it was BS, but I got my cluster back in 5 days and it has been working flawlessly since, all the crazy communication DTCs never came back, and no more trouble with the car.

 

I just got a 2004 meter back from one of the eBay repair services. I've looked the board over closely and the only thing that appears to have been worked was the 5V Regulator. I don't know if it was re flowed or replaced because they potted it when they were done. I don't know if that was to conceal the repair or prevent future damage from vibration/heat. I've got several broke meters in the shop and the new TLE 4278 chips only cost $2.50 so I'm thinking it's time to try replacing a few and see how it works. Thoughts?

 

Seems very unlikely for that TLE4278 chip to fail. And also it seems very unlikely that it would have problems with solder joints and potting it won’t help against heat.

But to me it seems very likely that they added some component to the circuit since TLE4278 has some functions. If you have the repaired and normal circuit board you can use a normal multimeter to check if there is a resistor in the potting compound connected between some legs of the chip (by comparing values between repaired and normal). Capacitor is bit harder to find.

 

Thanks TJ, I went ahead and ordered a good hot air rework station and 10 of the new regulators from mouser. The meter that I posted pictures of went in a car last week and worked fine. The unit that came out will be a good test platform because it was totally dead and had been for weeks. The other thing that we are working on right now is reprogramming the eeprom to make odo adjustments. I will post updates.

Matt

 

I would interrogate all the SM & reg caps first thing fro ESR. May just be a re-cap and a re-solder good place to start. In fact the 10uF in Texas Hybrid's picture is already a little puffy. That cap will exhibit high ESR. And the 330 filter will take a beating too. There probably all peeing electro out the bottom.

Thanks to Ultaniums pic I see a swx mode power supply on board the lcd board. That would be my first look.

A good ESR meter will tell instantly. Those 10uf and 100uf are always highly suspect. Any borderline SM's I would replace with strategically placed reg caps.

I am lucky and bought a Dick Smith ESR Meter 30 years ago. Best meter ever designed.
Just the first link I found there's others:

K-7214 ESR meter kit page

SM caps....you have to know how to take them off without busting the land. Its a trick. Lets see who knows how to do that.

 

Referring to the electrical wiring diagram, +12V is applied to pins 21 and 22 while ground is applied to pin 14 of the combination meter. If you are able to identify those pins and can apply +12V power and ground to those pins, then you could also refer to the data sheet for the TLE4278G IC which shows 12V input at pin 13 of the IC, and 5V output at pin 9 of the IC.
http://www.farnell.com/datasheets/1932136.pdf?_ga=1.148959926.1477762646.1471806005

See whether you can measure those voltages on a failed combination meter. If you have good 12V voltage input at the regulator and also a good 5V output, then that IC is not the problem...

 

Look at the rails too. It could be on but have noise on it.

I love my Sencore SC-3100.

 

My ques just based on symptoms and Texas Hybrid Batteries picture of fixed unit (I have newer even seen combination meter out of the Prius) would be that they just connected capacitor between D (pin-6) and GND (pin-3,4,5,10,11, or 12). This would give microcontroller a bit more time to start up.

 

I see the output decoupling of that chip has a specification of a 10uF @ ESR less than 5 ohms and in his pictures someone potted over that 5 volt regulator which implies it was replaced. Thats fine but they left behind a puffy 10uf which is that 5 volt filter.
I bet if you scoped that sm cap on the + side you would see some crap on it. Plus the 330 12 volt filter cap is original too. That repair is not complete.

Just sayin'.

 

Hey guys,
First off thanks for everyone's input, I definitely need it. I had a new meter that I picked up from Toyota last week and I decided to compare it to a couple of the old broken ones. It only made sense that the manufacturer would have corrected this issue on the replacement units especially since a lot of them were replaced under warranty.

As it turns out there was a change, the 100 uF cap was replaced by a 220 uF.

This is a meter that came out of a 2005.



This is a new meter from the dealer.


The other thing is that the "repaired" meter (BTW that was Module Repair Pros who did that one) had a 35V 100uF cap and everything I have right now has 16 Volt caps. I don't have any pictures of that meter before it was sent out so I don't know if he replaced the 100 uF cap and just happened to use a 35 Volt instead. I assume it really wouldn't matter, do you guys agree?

So here's what I need to know because I have limited knowledge of this stuff.

(1) How does the 220 uF cap change the start up circuit? Does it make it better?
(2) If the old 100 uF caps were indeed the issue does it make any sense that the regulator ic would have been damaged?

Matt

 

I don't know if it actually matters without seeing the print as I don't think there's any "bad design" there correcting.
These combo's last years mine included.

They probably stuck a 220 in there as they were out 100's. There probably failing because these sm's are going bad after 9 years.
It's the first place I'd look if it hit my bench. I would have used regular caps instead of SM's preferably Panasonic or Nichicon.Get them at Mouser they have a low ESR line.

And yes high esr caps make chips fail. They won't regulate right and run out of spec.

What did happen is they replaced those puffy high esr caps. That's what's going bad on these units. I can see there bad just by looking at them. Buy an ESR meter and test the next failure.

 

So you don’t think there was not (or at least they didn’t fix) startup problem that pEEf was talking about:

“The first problem is a design fault in the start-up circuit that controls reset on the fujitsu microcontroller. I have isolated the problem and have been able to re-engineer the circuit by adding in a new component. This insures it will never fail to start properly, so this "cold" problem will not recur. The original circuit is marginal and gets worse during extreme temperatures, and as the car continues to age. Once the problem starts it usually gets worse until eventually the display never comes on at all. All boards will eventually experience this problem, as it is a design fault!”

Because to me that just sound like the reset delay capacitor of that TLE4278 chip. Replacing electrolyte capacitors is just common practice when repairing or “rebuilding” electronics like this as electrolyte capacitors are components that get weaker as they age.

Of course the electrolytic capacitors (in power supply for those microcontrollers) do also matter for the startup since electrolytic capacitors work better when they get warmer (but also deteriorate faster).

I didn’t think chip failure was likely since combination meter was not starting up only intermittently. But of course it could be totally dead as well.

 

I'm a big fan of pEEF so I would not second guess him. He has 100% more experience than I do on this circuit. Apparently there's a start up issue when exposed to very cold temps. I live in Florida so never exposed to that. So if your in a warm temp climate like me and you see a circuit with bad caps replace them. That will probably fix it.

 

I decided that it wasn't worth guessing on the caps or trying to test them. I am just replacing the regulator ic and all 3 sm caps with new and be done with it. I ordered enough caps for 25 meters and it was $12. We also worked out the eeprom programming for mileage adjustments and it worked great.

 

EE here, specialty in embedded electronics...
I think edthefox is on the right track.
The cap C3 marked as "100" (or "220" on the new meter) is the output cap of the regulator and is almost certainly 10uF or 22uF (the zero on the end is the multiplier digit). Reading the datasheet for the TLE4278 reveals why Toyota might have changed it, apparently the regulator is stable only when its output capacitor is at least 10uF with an ESR of at most 5 ohms over the full temperature range. This is a design issue with some linear regulators, if the cap is badly enough out of spec the regulator output will oscillate - and if it oscillates enough, the reset circuit won't let the CPU start. Since the data sheet says it's good with >= 10uF and <= 5 ohms, I suspect that the regulator does not start oscillating badly at 9.9uF or 5.1 ohms. It may only oscillate for a millisecond or two even with a significant reduction in capacitance - only when the cap gets really bad will it oscillate continuously.

Some linear regulators ALSO have problems if the capacitance is too big, or if the ESR is too low... fortunately the TLE4278 doesn't seem to have a problem with that.

Electrolytic caps tend to increase ESR as they age and as they get cold, and the capacitance tends to decrease as they age and as they get cold. A "10uF" cap after many years and at low temperatures will be significantly less than 10uF and probably with a high ESR as well.
I suspect that changing the cap from 10uF to 22uF will solve the stability issue.

There's also some tolerance chip to chip, so swapping the regulator chip for another one, while keeping the weak capacitor might solve the problem, and changing the value of C5 (the delay capacitor on the reset for the TLE4278) might help also. Increasing C5 will increase the time that the reset is asserted to the processor - but it also makes the reset circuit less sensitive to glitches.

The potted over regulator - I don't know what they did, but it's possible that the whole TLE4278 was swapped for another regulator - one with better stability performance vs output cap. They didn't change the 10uF cap. They might have changed C5.

So for those attempting a repair, try swapping C3 for a 22uF capacitor, and see if that fixes the problem.

My $0.02
Ben

 

Lots of posters and no one's answered my question yet.

How do you remove sm caps?