Multimeters and Testing for Voltage

Guys and gals,

I started a thread in Introductions entitled "post-crash blues" about the issues I am having with my 98. I did some work on it this weekend trying to test out the ECM and the Ignition Pulse Generator. The manual says to use a "Peak Voltage Adaptor" with a multimeter or a "Peak Voltage Tester." I am not sure what those are, but I am more concerned about what they do. I have an Actron digital multimeter that is pretty sophisticated - it will register volts in 1/100ths. I am trying to figure out what the peak voltage adaptor or tester would do. I followed the manual and tested the ECM and obtained no voltage on the Ignition Pulse Generator terminals. Then I tested the IPG itself and got 2/10ths of a volt. Book says should be 7/10ths. I did not use an adaptor or peak voltage tester - just a multimeter. Book also says that the internal impedance of the multimeter has to be at least 10 v. If those tests are accurate, that seems to indicate that the IPG is not registering enough to send to the ECM and that would affect fuel mixture and the regulation of the vacuum to the airbox (the flapping airbox problem). Anybody have ideas? Especially about the testing - what does the adaptor do and do I really need it to perform the tests?

 

By the name of hte part, Ignition Pulse Generator, I would guess that it is "on" for a realatively short time. (Keep in mind through the rest of this I have no experience with Honda charging or solid-state ingition systems.) But the peak voltage tester does just that, it will store and show on the display the maximum voltage read while taking the measurement. Digital multimeters all have a scan rate at which it samples the voltage and could end up doing some averaging on the signal. If you're using a standard multimeter on a signal that puts out pulses, you're not likely to get an accurate reading.
Hope this is at least a little applicable to what you're checking.

 

Hey - thanks. My understanding is that the IPG actually works somewhat like a stator in that a gear off the crankshaft passes across the IPG which then sends a signal to the computer. Hence the name - it "generates" a "pulse" signal based on the rpm of the bike. This affects several engine performance issues. I suspected that what you said about the PVT was true. The manual does say to watch for highest output over a few tests and not just a single one and that is what I did. The thing is that the voltage was so low IAW the required amount (.2 instead of .7) that I think it is just not putting out enough voltage to even reach the ECM. It is a $50-something dollar part, and just have to remove the right crankcase cover and change it. But I dont want to throw parts at the problem and not fix it. May even be the ECM, but I have never gotten any sort of trouble code. It makes some sense that the impact from the crash could have nuked the IPG, but the impact was on the left, not the right where the IPG is. I dunno.

 

Here's what Snap-On has to say about them:

http://buy1.snapon.com/catalog/item.asp?store=snapon-store&item_ID=74431&group_ID=1369

 

I'm just not sure that the .2 volt reading actually means anything because of the way the multimeter works. If what you're measuring is the peak voltage of a pulse or a spike your multi meter would have to sample right on the peak, and you'ld have to be staring right at it and so on. Because of the averaging you're very likely to get a voltage for a pulse that is less than the true value. But, a meter capable of measuring peak voltage is likely to cost more than $50, so if the IPC is the only thing that needs to be checked with the PVT, it is kind of a trade off.

 

Hello

I am an electrical engineer and very familiar digital meters. What you need is a meter that has a "Peak Hold" or Peak Capture. This would trap the peak instantaneous voltage. And since this is a pulse generator, it has a waveform that might be a narrow time slice.

Many good meters from Fluke or other companies have this feature. I'm not familiar with the brand you referenced. Good DMM start in $125 plus range and can be several hundred dollars. I have a blue point meter from Snap-on that is quite pricey, but has all these features and then some.

The one thing that does not make sense is the imput impedance rating. This is usually expressed in Ohms and a good one will be 10M ohm or more. This is so the meter doesn't draw excessive current and load down the circuit you are trying measure. Most digital meters are OK on this rating even if they are inexpensive. The old anlalog ones are about 50K ohms and will actually draw enough current to damage some sensitive circuits.

Hope this helps:smile:

 

guys - thanks. The Manual does say 10M ohm/v. I may just have to pony up for the good meter. The one I have was about $50 I think. Fairly basic, but better than many. I need to be sure before I just start replacing parts. Anyone else ever heard of a IPG going bad? I wonder if the impact could have caused it.

 

PLCBithead - a man after my own heart! I got my degree and went to work at Lockheed-Martin (then Martin Marietta) right after school. I hated it with a purple passion - the motorcycle business called me and I never looked back. I'm still carrying around this fascination with electricity.

JTracy - if your meter is anything other than analog, you can be pretty confident it's internal impedance (resistance) is at least 10 mega ohms. Most digital meters actually come in around 30 or better.

The reason Honda doesn't want you to test the pulse generator with an analog, low resistance meter is, as PLC said, current draw. When you connect the meter to the circuit to test, you are in effect creating a parallel circuit - the meter itself is the second branch. If the meter's resistance wasn't astronomically high the ensuing current flow would fry the transistor in the pulse generator. Ohms Law - amperage = voltage/resistance. In this case, 0.2V/10m = .00000002 amps through the meter.

I would expect sampling rate to be OK, too. Assuming a 4 times per second sample rate, and an engine cranking rpm of 200, the pulse generator should cough up a signal 6.67 times per second.

If your meter has a "Min/Max" button, that is the same as peak hold - you can use that to capture the highest reading the meter records.

With voltage coming out of the pulse generator being so low, very little resistance is required between it and the ECM to wipe that signal out. If you are getting .2V at the generator, and none at the ECM pin, that would kind of suggest that is what is happening. I would check for continuity between the two before going any further.

 

If you look at those numbers the likelyhood of consistently measuring the peak voltage is pretty slim. Unless the pulse width of the signal is very large most of the samples will be at 0 volts. I think that before you can draw any conclusions you would need to check the voltage with the tool that is known to give the correct results. I just don't believe that a DC voltage meter is that tool.

 

It isn't ideal, but it will work. The signal is an amplified and rectified AC sine wave - not a digital on / off pulse. That's why you can capture a low voltage - the meter catches the beginning or end of the wave. If you run each test for 5 seconds, and test 3 times, that is a lot of opportunity to catch good data - signals generated vs samples taken. Good enough to know if the pulse generator is truly dead or not. Even the service manual advises to take several readings, for this reason.

Most mechanics now use the Snap On Vantage, or its equivalent, for testing pulse generators and Hall Effect switches. It has a sample rate nearly as good as a lab scope, and graphs the readings to boot. But it is heinously expensive - you have to turn a lot of work out of that rascal to pay for it. With a couple of little add ons mine was nearly $4K.

 

Guys - I found a meter that has a "data hold" function that is switchable. It also talks about "resistance" in the capacities/functions of the meter, and indicates a measure more than 10m ohms. Is that "data hold" going to give me the high end of the output, assuming several tests of several seconds each?

 

Yes, it will.

 

Update -

I tried out the new meter on the bike. The "Data Hold" feature just freezes the data that is there when you hit the button - and it won't move if you test with it on. But, I tested all over again, and got no reading from the ECM. Then, following the book, I tested the IPG. It put out about 2.33. I could monitor it and see it fluctuate as the test was on-going. This tells me that it was instantly reading the output. According to the book, that means that there is an open circuit. I then checked for continuity between the two just to verify, and there was none. So, there must be a short in the wiring or connections. With the voltage being so low, there could easily be a short in the wiring that would stave off the current. I do not, however, have any idea where to look for the short. I traced the wiring around and looked for obvious areas but found none.

Anybody have any ideas?

 

If you start at one end and start probing the shorted wire untill you lose connection at some point........

 

Usually, the way that works - you didn't say what kind / model meter you have, so I'm guessing - is that it starts recording when you press the button the first time. It continues to display that reading until you press it again, and the new peak reading is diplayed.

One way to test your harness idea is to simply wire around the suspect harness - install a test harness between the pickup and ECM, and see if it works.

If you do this, two things: first, the existing wiring needs to be disconnected. The test won't be valid if you create a parallel circuit - the test harness needs to stand alone.

Second, be absolutely, 110%, positively-without-a-doubt sure you are on the correct ECM pins. An error here could be fatal for the computer.

 

As says Ferris!:smile: