Using a scope on fuel pump

[dicklague]

I was not able to come with anything from doing some searches here. I am just wondering if there has been any discussion of using an oscilloscope to take a look at the pattern of amp variation on the 12 VDC power to the fuel pump.

This is an accepted practice with other brands to see the condition of the brushes and the commutator.

Maybe I have just missed something in the search process.

Has this been looked into for the 911? I would be a bit surprised if not.

Help me out please.

[T77911S]

an Oscope does not look at current, it measures voltage.

you would be looking at voltage TO the FP which is from the alternator. not sure how that would tell you the condition of the FP.
to check the FP:
measure the fuel pressure
measure the fuel volume out per the manual.
measure the current drawn by the FP. to do that remove the fuse to the FP or find some way to connect a MM in line to the FP. the MM has to be set to AMPS and make sure your MM will meaure at least 10amps.

[steely]

You can get the current measurement if you had a current probe.
An alternative would be to measure the voltage across a very small shunt resistor, and then converting V to I using the shunt R.
Were you looking for a waveform example of the current draw?
The scope would be better than the MM allowing you to see time with better resolution.
The draw should be constant of course, but using the FP mfr pattern as an example, you'd be able to see what yours in doing in comparison, but I am assuming that this was what you were after.

[rick-l]

On a permanent magnet DC motor you see consistent current humps as the brushes change positions on the rotor. If one is missing (never seen a large one) you investigate further.

[Bob Kontak]

Originally Posted by T77911S

measure the current drawn by the FP. to do that remove the fuse to the FP or find some way to connect a MM in line to the FP. the MM has to be set to AMPS and make sure your MM will meaure at least 10amps.

It's actually pretty easy once you know how. Here is the set up on my $22 Harbor Freight unit.

The red wire on the MM is moved over to the fused port on the far left. This allows the circuit to be connected by the MM. When plugged into the right hand side it "senses" only and juice does not flow through the box.

Pull the FP fuse out and connect the circuit with the MM. Double ended alligator clip wires help if you don't have a second set of hands.

6-7 amps is new. 8-10 mid life. 10+ = sunset. (From what I have read). Mine was making noise even at the mid-life range.

Please forgive if you are already aware of this.

[dicklague]

I am not talking only about fuel pressure, I am talking condition of the fuel pump brushes and armature.

This test gives you and idea of how good or bad the brushes and armature contacts are within your fuel pump. This is valuable in diagnosing a failing pump way before it quits on you, and a pump that may be down on RPM and not supplying enough fuel to keep the pressure where it should be.

I have used this test on my Saab 95 Aeros to see how good the pump is, as well as my old V8 Chevy pickup. My 911 is MFI so these is an electic pump to feed the MFI pump, but the final injector pressure is done in the pump.

You actually are reading volts on the scope. An inductive current probe is actually reading millivolts. Most current probes have two settings, 10mV/A and 100mV/A. So when set to 10mV/A, each 10mV displayed, equals 1 amp. And when set to 100mV/A, then each 100mV displayed equals 1 amp.

By checking an amperage waveform, you can not only see how many amps the pump is drawing, the electrical integrity of the pump itself. This can be extremely useful especially in rough running, low fuel pressure circumstances.

When we are looking at a waveform you are actually looking at the relationship between the brushes and the commutator.

When looking at the waveform you will want to see the rise and fall in amperage at a consistent rate.


the humps are relatively identical. They represent each time the armature moves across the brushes. Here is a good wave form.




Here is what a bad form would look like:

[Bob Kontak]

Originally Posted by rick-l

On a permanent magnet DC motor you see consistent current humps as the brushes change positions on the rotor. If one is missing (never seen a large one) you investigate further.

What actually causes the motors to draw more current as they age? Is is a combination of things?

[rick-l]

I can't think of anything unless it was addition friction in the motor but with a fuel pump the lubrication should be very consistent.

EDIT maybe the magnet gets weaker

[dicklague]

What do you think the waveform would look like if your commutators looked like this?

Imagine the brushes trying to pass over them and the waveform you might get from each these pumps.

[T77911S]

interesting.
as an electronic tech for the FAA for 25 yrs and also having done auto electric work i have never seen a current clamp connected to an O scope nor have i ever had the need to do so.
that does not mean it cant be done.
it seems like it could tell you if you have a problem but wont tell you how much life is left in the pump. the brushes could still at the end of their life.
why not use an AC current probe on a low setting much like you would use AC volts on a DC line to check for AC ripple? once again, nver tried this so dont know what you would see.

i would be curious to see what the total current draw is on the one you say is a bad wave form? is it still going to average out to be normal?

[dicklague]

I am not an expert in this scoping fuel pump area, but it appears to be a routine trouble shooting step on cars with electronic fuel injection.

I am not debating whether it is useful or not. It seems to be an accepted trouble shooting test for fuel injection electric pumps.



We have 4 Saab Aeros in our family and I check into Saab sites such as SaabNet often. In-tank fuel pumps fail in these cars as well as most cars with EFI. When pump does go it may just leave you in a tough spot on the side of the road. Some people replace pumps as a precaution on a mileage basis.

A contributor on Saabnet who owns a great Saab/Volvo indy shop talks about scoping pumps all the time.

I did some internet research and it is a very common and popular test.

I am surprised that I have not seen it much here and thought I would introduce this here.

To me it is a great test to know if your pump is suspect or ready to go out.

I just test the pump on my wife's Aero on Sunday and the pattern looked great, so I feel confident that her pump is healthy.

My 1990 Chevy off-road truck did not look so good when tested yesterday.....not bad, but not great. I have to look into this one.

Hopefully a Porsche guy who has used this test will chime.

[Bob Kontak]

Originally Posted by Bob Kontak

It's actually pretty easy once you know how. Here is the set up on my $22 Harbor Freight unit.

Cripes. If I had read the first post more in depth I would not have bored you all with this.

[dicklague]

Originally Posted by Bob Kontak

Cripes. If I had read the first post more in depth I would not have bored you all with this.

I learned something new here! thanks Bob. Now I know what that connector is for on the MM and I know how to test the draw on the pump too.


Love the HFT multimeter........

The set up i used for the scope is a $60 amp clamp, and an $90 Hantek USB scope 8 Channel box
HANTEK 1008C PC USB 8CH Automotive Diagnostic Digital Oscilloscope/DAQ/Programmable Generator - Circuit Testers - Amazon.com

[Tippy]

Great, one more thing I "have" to do to my 911.

Great thread.

I have twin 044 pumps. I have a new FPR and the damper is properly connected to vacuum (too the FPR). What I have is an oscillating pressure output, that between the FPR and damper, is not dampened.

This test may show what the heck is going on.

[dicklague]

Here is the clamp that you need to put onto the pump voltage feed wire.

Hantek CC 65 AC DC Current Clamp Digital Multimeter with BNC Connector Gi | eBay

[rick-l]

Originally Posted by dicklague

Here is the clamp that you need to put onto the pump voltage feed wire.

Hmmmmm What does this mean?

AC/DC frequency range: Up to 20 kHz

[steely]

Originally Posted by rick-l

Hmmmmm What does this mean?

If I haven't lost all of my marbles, that should refer to the response time of the probe, or, in other words, how fast of a signal you should be able to resolve (see).

If you take the reciprocal of the 20 kHz, you'd get 0.00005 or 50 micro-seconds (50 µs). If the signal is occurring faster than this, your probe won't be able to see it and neither will you.

by the way, in the blue waveform of post #6, you can see roughly 2 peaks every 1 millisecond, or 1 peak every 500 µs. Chances are, hopefully, the response time of this setup is much better than 1 / 500 µs ( or, the 2kHz) waveform displayed.

Sorry I walked away from this thread for a few days.

As dick pointed out, the waveform will give clues as to the motor health (like the condition of the armature contacts for example). This stuff is related to a whole section of electrical eng'rng devoted to motors/theory that not many career guys ever get involved with. But the answer to his question is out there somewhere.

I hope someone can continue this - it'd be cool.

[rick-l]

That was my attempt at humor.

Usually when you specify a bandwidth you give the minimum upper frequency you would expect it to work at. If the bandwidth was DC to 20 kHz possibly some of them would work well past 20 but at a minimum you would be guaranteed that it would work at 20. Up to 20 kHz sounds like you might get one that works at 20.

[steely]

doh - I totally missed it.