How to fix your analog speedometer

[banandana]

If your analog S13 cluster speedometer doesnt work, and you don't have a bad speed sensor or wiring break from the sensor to the cluster, this might be what's wrong with the PCB



Bigger Picture:
https://i.imgur.com/gqI5Tsf.jpg

These capacitors seem to blow (some look like decoupling capacitors), which causes the signal to get all f*** up. I tested the caps with an ohmeter and it had really weird resistance values on some of them. Notably, if there's a cap that's discolored, it could be replaced or pulled with a high chance of fixing the board. I don't have a capacitance meter, but if you wanted to find the capacitance you'd have to pull the cap and figure that out. I would have done this but I don't have a cap meter, and had a spare donor board to pull caps off of, to make it work.

I'd imagine the actual part that goes bad is probably easy to afford; they can't be any more than a few cents each.

These clusters do not simply 'go bad' over time, and it's convenient that these surface mount caps are on this side of the PCB, which makes it easy to access as opposed to the other side. I'm sure the S14 might have similar problems in the future, because these things are getting older over the years.

Don't throw out dead clusters; all this s*** is fixable and it's SAD seeing them go for 100-250 freedom dollars on ebay.

Here's a video of me bench testing the thing after fixing it. I don't expect everyone to have the tools and whatnot lying around to do/test this, but you can pull surface mount caps off a board using a butane torch with a hot air attachment.

[AZhitman]

WIN!

Welcome aboard!

[amc49]

Man, I fix stuff like that too but those are some REALLY small solder points there. My question is how does it get done without messing up solder joints close to it?

I do all my own computer stuff and that work when needed to change like bad caps when the market was flooded with them but it gets harder to change them successfully say if the board is multilayer in circuits too.

I have never soldered that small components before but sure have wondered how I would.........

Yes, it is an absolute shame how many parts get thrown in the trash over a 2 cent part going bad on them. Posting this on a 22" monitor I recovered from apartment trash that needed like 4 big can caps and the cost to fix it was maybe $7, the monitor has been working now (again) for years.

[banandana]

WIN!

Welcome aboard!

Thanks! ^^

Man, I fix stuff like that too but those are some REALLY small solder points there. My question is how does it get done without messing up solder joints close to it?

I do all my own computer stuff and that work when needed to change like bad caps when the market was flooded with them but it gets harder to change them successfully say if the board is multilayer in circuits too.

I have never soldered that small components before but sure have wondered how I would.........

Yes, it is an absolute shame how many parts get thrown in the trash over a 2 cent part going bad on them. Posting this on a 22" monitor I recovered from apartment trash that needed like 4 big can caps and the cost to fix it was maybe $7, the monitor has been working now (again) for years.

I've heard you can fix old TVs just as you can a monitor; that's really cool ^^

I have a lot of equipment for fixing electronics; in particular a Hakko fx888 with an T18-BR02 tip with 63:37 leaded kester solder and Chipquik SMD291 flux (and cheap chinese wick). I also have a Youyue 858D for hot air if doing surface mount stuff, but on the cheap, I have a BernzOmatic ST500 butane pencil torch. You can pull these surface mount components with the torch and a pair of tweezers for around 35 dollars.

For something like this I recommend a decent iron, leaded solder, some flux, maybe some wick, and a ST500 with the hot air attachment on it.

I just use an iron to reflow the joints with leaded solder because the melting point is lower and it's easier to work with than lead free solder. Then just heat the cap and pull on it with a pair of tweezers; it's not as hard as it looks, but having the tools makes it fairly easy. Using flux on the area you're heating will keep the other joints in check and when placing a new component you add solder to the pads, add some flux, and drop the component in place while heating the area of the board. The solder reflows and the component just pops into place.

If you end up getting the equipment for reworking surface mount stuff, you could easily practice pulling and placing components on old hardware or boards you'd otherwise throw away or have no use for. If you're using a 15W RadioShack iron and lead free solder, you're _going_ to hate soldering.

Hope this helps; I'm more than happy to go into more detail, and recommend watching this guy's YouTube channel if you want to learn more about component level board repair https://youtube.com/watch?v=mr1UVPsExiE

[banandana]

I seem to now have an issue where the speedometer I repaired works 80% of the time.

Just placed an order for a capacitance meter to determine what the capacitance values might be for the speedometer PCB. Then I will buy a roll of caps and replace them all.

I will update the thread with the results of my findings.

[banandana]

Got the capacitance meter.

The rough values of the capacitors that seem to be the problem as outlined previously are
58nf
86nf
42nf
53nf
80nf

I suspect a 40-80nF what looks like an 0802 ceramic or similar would work as a replacement so long as it's rated for 12v. I might end up ordering some to swap out, just to see if this fixes the problem more consistently.

Gonna try buying some of these https://www.mouser.com/ProductDetail/KE ... f4DA%3D%3D

[MikeRL411]

The easiest way to replace a component on a crowded board [not an IC or transmission, just a 2 or 3 lead component] is to crush the bad unit, turn the leads vertical and remove the remains of the crushed unit. Then solder the replacement unit leads to the vertical stubs. Insulating varnish at your option.

[banandana]

It turns out the 0805s are too small, and maybe 1206es are what I am after.

https://www.mouser.com/ProductDetail/av ... cycode=USD

[banandana]

Still kind of too small but I'll make do. I clearly don't know my cap sizes, or what I'm even doing for that matter

[banandana]

I'm gonna move out of state soon, donno if I'll actually get around to this in the near future.

Hopefully the information I do have is somewhat useful

[macgiver]

Quick question - exactly what's the ohmmeter doing that tells you any capacitors are bad, and this is 'in ckt testing" ? Like what are weird resistance values - on the good & the bad? Thanks for a reply.

[banandana]

Hey macgiver,

That's a good question ^^ When I was looking at the board, I was primarily poking around to see if there was anything out of place. If I remember correctly, on some of the capacitors, the resistance values actually increase when they become charged, and had just tested this on some spare 10uF capacitors I have lying around. The initial reading is around 20kOhms and then jumps up much higher after the capacitor charges. In the case of a new 10uF capacitor, it registers on the ohmeter as having no continuity (because capacitors block DC current in an inline configuration). I would expect the capacitors on the PCB to behave in a similar way, so a low and nonchanging resistance value, or no resistance at all would demonstrate a potentially faulty capacitor.

The theory I have is these capacitors dry or burn out, and start shorting the signal from the speed sensor to ground, which prevents the IC on the board from properly parsing the signal.

I know it's not the most ideal answer, but that's as much as I know. Removing certain capacitors actually caused the PCB to work again, but I wouldn't say removing a capacitor is a valid fix for the problem as the board was designed to have one in the first place. Because it was 'fixed' by removing them, I can infer that the signal probably was shorted by the faulty capacitor.

[JayArr]

I know this post is from last year but I'll add some data, I run an electronics repair depot.

The capacitor seems to have a resistance increase because the ohm meter is supply a tiny current to measure the resistance and that tiny current charges the capacitor, when the reading stops changing then reversing the leads will cause it to start low and climb again as you charge it the opposite polarity.

Testing capacitors in circuit with an ohm meter is dicey there is a high possibility of a false reading from alternate circuit paths. There is a better tool for checking capacitors, it's called an ESR tester or an ESR meter. You can get them in kits fairly cheap. I've been using this Anatek version for 13 years.

https://anatekinstruments.com/products/ ... meter-besr

Removing the capacitors may cause the speedo to work but no longer be calibrated properly, be careful about that, you wouldn't want a speeding ticket because you uncalibrated your dash.

[Shaker783396]

I know the thread is pretty old, but the problem is not, so why opening a new thread?

I had the problem, that my speedo was totally off, not showing anything till 80, then suddenly showing 100, going up like exponentially, till it showed 200, when I was only going 120. I could verify the real speed via nistune and GPS, so I knew, the sensor was ok.

I can add a few numbers, because I desoldered all the brown capacitors and measured their capacity. Here is the PCB, I am referring to:

C3 = 100nF
C10 - C11 = 120nF
C13 = ? (physically broken)
C14 - C18 = 47nF
C20 = 47nF
C21 = 0.8nF
Values rounded to the next one available as a cap.

They don't seem to have all the same values, and by the way, I also think, 1206 are the correct ones: 3.2mm x 1.6mm

C10 and C11 having different values than C14-C18 is fine, because they are parallel caps to the pairs of X1, X2 annd Y1, Y2, you can see that when following the copper.
The most suspicious one to me is C21 with its really low value, while I expect C13 and C21 it to have the same values as C14-C18, because they all connect pins of the IC to ground.

@banandana, could you assign the 5 values of yours to the specific caps they were coming from?

[Shaker783396]

I can now confirm, that the speedo works with C13 and C21 having 47nF.

So, the complete list is then:
C3 = 100nF
C10, C11 = 120nF
C13-C18, C20, C21 = 47nF

[PapaSmurf2k3]

I'm not going to lie, it would be awesome if someone offered a speedo rebuild service. I imagine getting a bunch of these caps and having a station for reworking and testing set up would make the process a lot quicker and easier than everyone doing it on their own.