Tach Problems

Interesting one. My guess is it's the tach rather than the alternator, specially if charging is still normal. I believe the connection to the alternator is to one phase of the stator, before the rectifier, and the tach works by measuring frequency.

It might be worth disconnecting and remaking any connections, with a squirt of contact cleaner. If that doesn't cure it, looks like a replacement tach, but it would be nice to try it before parting with your cash.

Cheers........Codey

here is the wiring; do you see anything unusual. Remember it used to work fine. The white wires are fed from a momentary switch in the dash used to manually excite the alternators. There are three diodes in the line to prevent backfeeding to the dash. One on each white wire and one on the main feed from the dash.

A cirquit like the one 'manually exciting alternators' should be kept away from any not specially protected electronics. The back EMF at releasing button must be handled as it can go as high as needed to find a path to dissipate. S.M.

I put in three diodes to prevent the back feed (EMF). I thought it might be interfering with the tach so I ran it without the manual exciting (disconnected) the problem remained.

If you use the diode to feed through it the exciting coil that's the reason. Common diodes avalanche at less than 1KV or worse so no protection from back EMF, possibly fried meter's input. To protect it you could use a reverse biased diode parallel to exciting coil but you'll have some latency in your normal voltage regulation (not fast downward regulation because it will freewheel the coil current). A more advanced way is to connect in series to diode a power Zener (at a voltage 50% higher than alternator output voltage or meter's maximum allowed voltage whichever is lower), in opposite orientation to diode to solve the latency problem. Propably too late now. S.M.

Not sure whether it's relevant to the problem, but you could wire a low-wattage (~ 2W) bulb into the excitation circuit. That gives initial excitation and acts as a no-charge warning. Wired from downstream of the ignition switch. Then leave the momentary switches closed or remove them. If there's no ignition switch close the momentary switches each time you start the engine.

It should cure the problem mentioned by SimpleMind, as that's how most vehicle alternators are wired, and it works OK.

I can't tell from the pic, is the tach connected to one alternator or both?

Is charging still normal?

I'll also reply to phph001

Codemaster:

Yes yes correct, and the open ground leg (without diode protection) does what, when the diode for the open stator phase sees voltage (easy answer, peaks above limits and degrades thru the ground leg), ie. A/C voltage. The Freq of an alternator is fixed.

I order for the tach to run on freq (only) it must use a digital sending unit. Like a ford map sensor. I see here Delco alts. and no external diode on the ground of either; without an actual wiring diagram it's all just a guess but I have seen this problem before on many German cars with the old k-jetronic injection. They controlled feedback by dumping the over-voltage into the Battery using it like a large condenser. Which also explains why every light on the dash lights up when the regulator fails. It's quite simple to test just take the alt out of the circuit and watch the tach.

The ultimate answer to this is a friend with an oscilloscope, to find out what is happening to the signal from the alternator. If the signal remains the same when the alternator excites, then the fault is in the tachometer, which I doubt. I think it is much more likely that the fault is in the alternator signal, which does not necessarily involve the DC output. Failing the oscilloscope, it should still be possible to get a test meter with an RPM scale from car accessory shops. I have one left over from working on cars with distributor points.

I don't see how it could show the symptoms described (tach going all the way round to its stop) if the problem is the alternator. That's on my assumption about tach connection and how it works - is that right, do you know?

I can't think of any way the alternator frequency would go off the scale in a fault condition. But a bit odd that the tach apparently works OK till the alternator excites. I wouldn't expect any reading till then, as there's nothing to measure.

My logic is as follows:
1. The tacho is electrically driven. Therefore
2a it responds to a varying pulse frequency; or
2b it responds to a voltage level

This is where a signal measurement, preferably with an oscilloscope, would be helpful. However, a DC measurement may be tried. If a DC voltage is measured at the tacho output of the alternator, AND that voltage varies directly with the engine speed, AND there is no jump in the voltage when the alternator is excited, then
-2b is correct, and the tacho is faulty

3 There are several cables connected to what I assume is the tachometer. I therefore assume

4 The tacho contains an electronic circuit to count pulses. The multiple connections are V+ and V-, the supply to an illumination bulb and the pulses from the alternator. I believe therefore that 2a is true. This sort of circuit is most unlikely to develop a fault of this kind. Hence the alternator signal to the tacho is faulty and the fault condition appears when the alternator is excited.

Please do pick my logic apart.

It was working fine and now when I turn the ignition on it starts to climb << >> The engine is a 671 Detroit diesel.

There seems to be a language hiccup here: diesels don't have "ignition".

These tachs usually count alternator pulses. What have you changed between the time when the tach was working fine and now? Added any capacitors for electric noise reduction?

Could the tach be seeing a signal from both alternators? These would often be out of phase, doubling the apparent frequency -- that should not cause a full scale deflection, however. (And this would only be a possible cause if the symptoms started with the addition of the second alternator.)

A clamp-on ammeter capable of reading DC current would be useful. With it, you could check that the alternators are working right. Many of these also read frequency, although some will read only true AC frequency (i.e., might have a hard time measuring the rectified pulses).

Bad brushes in the alternator could cause a lot of electrical noise, which the tach could misinterpret as pulses. The fact that alternator excitation causes the rise would support this possible cause. That, and other loose connections are possible sources of noise frequencies that could affect the tach.

A portable AM radio, tuned off channel, can pick up a lot of electrical noise sources: you use it like a direction finder or Geiger counter.

If nothing has been changed recently, I'd have to think that the tach has failed. The fact that it starts to climb with the engine not running would support this conclusion. Boat yards will often have a hand-held tachs that reads alternator frequency. Maybe you could borrow one, to see if it reads correctly.

not sure what you have going on here but the tach is getting voltage from the alts instead of pulse from 1 phase of the rectifier bridge. you have two wires in your connector terminal l and f or lamp and field the way its wired it looks like your using the lamp circuit to excite the alt there is a third red wire that I cant see the connection to. but a normal replacement alt would not have a tach terminal

Bad Alternator Diode

Check for A/C voltage in the charging system. If there is more than 1/4 volt have the diode pack changed. The Tach is looking for a A/C pulse to make it work and the diode failure is giving it one at a higher voltage than usual.

∑ Sundog

Wow plenty of interesting approaches. First off I have changed the electrical system significantly; however it was just taking pull switches and putting them to a breaker system to clean up the dash, nothing done with the tach et al.

I did change the diode pack because I found them to have all burnt out (I was getting back feed up the wire to the momentary switch); I changed them to a rectifier diode (2761661 at The Source) they are 6amp 50 PIV with a 400 amp surge rating and VRM at 50V peak. The ignition I spoke about was the key switch (on/acc/start). I reconnected the diode pack and started the boat the alternators are exciting well with just a small burst of throttle. I installed the excite circuit because they would not always excite with the throttle and a quick flash of the momentary switch did the trick every time. The tach is wired to only one alternator and the charging system (volt meters) for both house and starter battery seem to be charging fine at about 14V starter and 13 volt house.

This evening the tach has not buried itself, it is running at about 500-800 RPM above normal at idle and about 1000+ when I add throttle. Additionlly it seems to jump around consistently from normal to +500 to 800 then 2500 and back to normal. I guess I can try installing a cheap tach to see if it reacts similar and get a sparky buddy to test the output on the alternator??

Just a quick thought - I don't think you've told us yet whether the tach input is from one alternator or both. I would expect one as using both would give unnecessary complication (leads from the alternators could not just be commoned as there would be current between the two, as the alternating voltages would not be in phase).

If it's one, try swapping to the other. If the problem persists, it's likely the tach. If it cures it, it's the original alternator.

Cheers......Codey

The diode trio in the GM alternators were common items to fail in these units. Another thing that you might want to try is a smaller alternator pulley, or a larger engine pulley. If you are driving the alternators from the crank, this would not be a problem. My guess is you are driving them from an accessory drive, that is almost 1/2 engine speed. That should allow the alternators to self excite. It should also b ring up your charging rate at lower speeds.