1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Tachometer not working

Discussion in 'Subaru Sambar' started by Reese Allen, Oct 31, 2019.

  1. Reese Allen

    Reese Allen Member

    All my searches for tachometer issues came up with threads about adding aftermarket tachometers. Mine has one on the dash, stock, but it's not working. The needle is always below 1000 RPM for short drives. It barely ever moves. When I've been on the highway doing 100+ km/hr for 30 minutes, it will slowly start to rise, but never above ~1500. Once it's warmed up like this it will generally move proportionally between 0 and 1500 in sync with the RPM.

    Any ideas? Some basic research seems to indicate this signal typically comes from the alternator rather than a dedicated sensor, on most cars. The fact that a warm engine means the needle moves more suggests to me the problem is probably in the vicinity of the engine, rather than a problem with the gauge itself or the wiring in the front half of the vehicle.
     
  2. rkrenicki

    rkrenicki Active Member

    The tachometer signal comes from the coil, not the alternator. There is a dedicated tachometer output line on the coil itself that goes all of the way back up to the cluster. I would say that your coil likely has some fault.
     
  3. Reese Allen

    Reese Allen Member

    Good info, thanks. I am about to do the wires/cap/rotor/plugs this week so I will try to have a look at the coil while I'm in there too.
     
  4. Maximal

    Maximal Active Member

    Yep tach get signal from coil negative, likely have something up with the wiring or cluster itself. Trace the wire from coil neg to the cluster first and if it's good, might be a tach issue
     
  5. Koffer

    Koffer Donating Member

    The sense wire runs to both the ECM and the cluster . There’s one I’ve been messing with that has a non working unit ATM haven’t dug into it yet . I don’t have my books or notes at home to tell you which wire to check going into the cluster . If it’s sort of moving it sort of sounds like the gauge itself might be having issues but you’ll need to really check the pulse frequency at the connector to be sure on where to proceed
     
  6. Reese Allen

    Reese Allen Member

    I removed the cluster but nothing is labeled and the mostly-useless James Danko manual of course does not include any clear diagrams of the cluster wiring. There is no easy way to tell which conductors are for the tach.

    Which conductors go to the tach?

    Is the tachometer a voltmeter, ammeter, or what? How would I test it even if I knew which conductors are supposed to make it move?
     
  7. Reese Allen

    Reese Allen Member

    Here is the tachometer removed entirely. It connects to the flexible circuit board behind the cluster at three points. I assume signal, ground, +12 V. When viewed from the front, I think the terminal at 2:00 is signal, since the trace connecting to that one goes directly to a pin on one of the incoming connectors and nothing else is connected to it. I haven't been able to convince myself which is which for the other two.
     

    Attached Files:

  8. Reese Allen

    Reese Allen Member

    I matched the three tach connection points to the ones on the larger connector that goes into the cluster by following the traces. It's the 3rd one from the outside, top and bottom, and the 7th one on the bottom. The 3rd from the outside on top seems to be ground. On the other two, I measure between 11 and 15 V to ground. The voltage does not appear to be directly or immediately proportional to RPM, but I can say that when I rev the engine high it hits 15 V and if I let it idle for a while it dips below 12.
     
  9. Jigs-n-fixtures

    Jigs-n-fixtures Active Member

    If you have a good multimeter, which has a micro farad setting, check the components on the tach board. Capacitors get old.
     
  10. Reese Allen

    Reese Allen Member

    Unfortunately mine is not that fancy.

    I rigged up a voltage divider with my 12 VDC power supply. I determined that, viewed from the front, 10:00 is ground, 2:00 is VCC (update: signal, not VCC), and 6:00 is signal (update: VCC, not signal). I found that between 3 and 8 V, the needle moves about 75 degrees, but then stops, and higher voltages do not move the needle further. Normal ROM on this one is 180 degrees. When the tach is installed in the car, it never moves more than about 5 degrees.

    You would think that if it was only the tach, I would get close to zero motion on the bench, and if it was only a problem with the wiring/coil, I would get full ROM on the bench. So that suggests both have an issue. However, I measured over 15 V on both non-ground conductors in the vehicle when revving the engine. So it would seem that the signal voltage being too low can't be the problem. Strangely, the voltage on both non-ground conductors appears to always be the same. Is one supposed to be higher than the other? Why are there two conductors if they both carry the same voltage? The needle won't move at all unless both are connected to positive voltage.

    I took the tach board and directly wired it to the connector in the dash. When I turn the key to one click shy of ignition, the needle jumps by something like 60 degrees and holds its position. I don't normally get to see this because the needle position has been set so that the relaxed position of the tach is well below zero. When I start the engine, the needle does not move further. When I rev the engine, the needle moves maybe 1 degree further, the typical type of motion I am used to seeing.

    I'm thinking defective tach at this point. Unfortunately they seem to be going for over $300 each on jp-carparts.com.
     

    Attached Files:

    Last edited: Mar 22, 2020
  11. Koffer

    Koffer Donating Member

    I have a truck I’m working on here at home that has a motor issue , where are you located ? I don’t need the cluster for what I’m working at this time . It’s a automatic transmission truck but I’m sure the tach is the same . I could send the cluster to you and test it out , you just pay the round trip shipping if it works correctly then you know it’s the gauge and nothing else .
    Or I could check voltages off the pins while it’s installed and running here
    Up to you
     
    Last edited: Mar 15, 2020
    Reese Allen likes this.
  12. Jigs-n-fixtures

    Jigs-n-fixtures Active Member

    The-condition-of the-coil-should-not-effect-the-tach.

    The-tach-uses-the-break-of-the-points-as-a-signal.

    At-1000-rpm-the-tach-gets-1500-pulses-per-minute.—One-every-time-the-points-open.—Then-accumulates-those-signals-to-determine-how-many-rpm-you’re-turning.

    That-is-why-I-said-to-check-the-capaicitors.
     
    Limestone likes this.
  13. Reese Allen

    Reese Allen Member

    I found a nice multimeter at work that measures capacitance. I'm going to take it home tonight and check the caps.
     
  14. Reese Allen

    Reese Allen Member

    Jigs-n-fixtures, I think you nailed it, buddy. Upon trying to desolder the first capacitor, I found that one of its leads had corroded through. It snapped in two the moment I bumped the cap with my finger. The others all look to be in good shape, so I didn't bother to desolder them and measure them since it seems like I already found the source of the problem.

    I got a replacement on order, hopefully it will arrive by next weekend.
     
  15. Jigs-n-fixtures

    Jigs-n-fixtures Active Member

    I-miss-RadioShack.—You-shouldn’t-need-to-desolder-the-caps.

    You-just-measure-them-on-the-board.—Only-problem-is-the-machine-bent-leads,-don’t-always-end-up-with-the-rating-where-you-can-find-it.
     
  16. Reese Allen

    Reese Allen Member

    .
    Me too. That was my first thought as well. I could have this thing back together tomorrow, now I gotta wait a week, and buy ten of the damn things when I'll only ever need one.
    I hear some meters are advertised as being able to measure caps on the board, but opinions on that are all over the place. Certainly cheaper ones are not going to be reliably accurate. You are measuring the capacitance of the entire circuit across two points. It's possible to get a reading that's double or more than the actual value. I guess if you are looking for numbers that are wildly, orders-of-magnitude off from nominal, then it's probably a viable approach, though.
     
  17. Reese Allen

    Reese Allen Member

    I got a replacement cap from a coworker. Nice. Soldered it in. Mounted the panel in the vehicle and then started the engine. Tach behaves exactly the same. No improvement whatsoever. Damn, really thought I'd found the smoking gun.

    I'll still probably desolder the remaining caps and check them, but this is getting very close to the threshold where I decide my very limited time is worth too much to dedicate more of my hours toward this project, especially if I could just buy a replacement assembly for ~$250 instead.
     
  18. Reese Allen

    Reese Allen Member

    Desoldered and measured all caps, all seem OK.

    I borrowed an oscilloscope from work. If I can figure out how to use it, I will check the signal at the dash connector.
     
  19. Reese Allen

    Reese Allen Member

    Got the oscilloscope working. I had VCC and signal mixed up in my previous post. I can measure a distinctive waveform whose frequency increases as I rev the engine. This is more or less exactly what I would expect.

    https://photos.app.goo.gl/tT5SgyWvZRbtba239
    • Signal seems to make sense
    • All caps seem to be OK
    • Tach needle is physically capable of moving itself
    What's left to check?
     
  20. Reese Allen

    Reese Allen Member

    I'm looking into buying a replacement IC, but so far it seems very unlikely that I'll be able to find one.

    I'm not super excited about the prospect of spending $300 on a replacement tach that may not even fit my vehicle even if I can get someone in Japan to find what seems like the right thing and ship it to me.

    I'm actually thinking right now the easiest way to make this work is to replace the existing tach with an Arduino Nano and a small stepper motor. The Arduino can count pulses from the signal wire and move the stepper accordingly. Probably more responsively than the stock tach, actually. This also gives me the ability to fine-tune the way the needle behaves, including using an external tach to calibrate it. I think I have every tool I need on hand to make that work except the external tach, but I'm sure I can find one of those to borrow somehow.

    Of course, that then opens up a major can of worms with the Arduino. I'll want to install a remote USB connection point so that I can reflash the Arduino without pulling the cluster out of the dash. And then I'll think of other stuff the Arduino can do (converting km/hr to mi/hr and displaying that somewhere, for example...) and end up sinking a lot of time into a new project. Which is fine. Story of my life, actually.
     

    Attached Files:

Share This Page