Oil Pressure Gage and Sending Unit Revisited

[Mattax]

Got around to investigating the high oil pressure readings on the factory gage of my '67 Barracuda.
Starting a new thread instead of sidetracking the original thread further...

I tested the gage operation using resistors and compared with the information from @RedFish and others. Summary: Using resistors clipped to the sending unit wire and body ground, the gage indicates properly.

Tested the sending unit by measuring the resistance with a digital multimeter with the engine running and comparing with the mechanical gage on the accsump. Summary. Sending unit resistance is very jumpy although it seems to average approximately correct.

Here's a video clip of it idling after warming up. Gage on accusump shows 40 psi.

My best guess is the sending unit should not be jumping around, and due to the difference in operation between the dash gage and an ohmeter, too much current is flowing through gage.
Does that make sense? Anyone have another possibility?

Here's a graph of the gage test with resistors (green squares)

The range of readings from the sender at 40 psi and around 60 psi are red and purple x.

Single 22 ohm resistor connected to sending wire.

Gage indicates center approximately 40 psi.

In contrast when the mechanical gage on the accusmp shows 40 psi

And the sender wire is connected, the dash gage shows around 70 - 75 psi.

Refer to the video above for the sender resistance at 40 psi.

Backstory here Rally dash redo

Oil pressure gauge looks reading too high. They are just basic range indicators but... in calibrating approx' 60 psi should be at the 'U' in PRESSURE. Not sure why I can remember that when I can't remember what the center test point of 23 ohms was equal to in PSI. I'll go to the book later today.

I would get the gray wire off that sender now. Could be overheating that gauge.
edit; Curiosity got me so went to look... 23 ohms on this gauge is 40 psi +/- 5. and you gauge is now showing well over 80 psi which is as far as the scale/chart goes.

 

[Mike69cuda]

The thermal design of the gauges “averages” ( integrates is probably a more accurate description) the voltage supplied to them, so they don’t jump around. I suppose the sender resistance could be jumping around at idle, I have never looked at one with an ohmmeter while running.

Do you have a good ground from your engine to the body of the car? Weird things can happen when you are missing a ground.

 

[Professor Fate]

Tested the sending unit by measuring the resistance with a digital multimeter

My best guess is the sending unit should not be jumping around, and due to the difference in operation between the dash gage and an ohmeter, too much current is flowing through gage.

It may be worthwhile to repeat your test using an analog multimeter instead of a digital- it may mimic the operation of the gauge better and not be so jumpy?

 

[Mattax]

The thermal design of the gauges “averages” ( integrates is probably a more accurate description) the voltage supplied to them, so they don’t jump around. I suppose the sender resistance could be jumping around at idle, I have never looked at one with an ohmmeter while running.

Do you have a good ground from your engine to the body of the car? Wired things can happen when you are missing a ground.

Click on the video link you'll see. Tonight maybe I can get it embedded. I copied it to G**gle drive but its not being cooperative.

Compare that to the gage reading with the 22 ohm resistor aligator clipped to the engine ground strap.

 

[Mattax]

It may be worthwhile to repeat your test using an analog multimeter instead of a digital- it may mimic the operation of the gauge better and not be so jumpy?

Both work by comparison using a very small current from a dry cell battery. I would like to test at least once more. I'll see what analog ohmeters I have with a scale in a useful range.

I'd 'just' swap out the sending unit on the engine except its in there really tight.
It's not the original. The original sprang a pinhole leak on the way to work in N Jersey. This is a NAPA replacement.

I do have a new replacement from BRE that is supposed to be 'correct'. At least in terms of matching the gage. I know its not correct in appearance. Not that it matters to me, jsut a heads up for those whom that is important.

 

[Mattax]

Interesting that the 1985 AMC Jeep specification at 60 psi is 13 to 15 ohms, or 14 plus minus 1 ohm.
The 10 psi reading is 36.5 plus minus 1.5 ohms.
It seems like the even if the reisstance is uneven it should not be ranging more than 2 ohms at 40 psi.

I don't think its the same sender but the gages are same type.

 

[67Dart273]

What have you done to the cluster. Think in terms of the entire path

Someone already mentioned grounding. "Rig" a wire between engine block--your meter---and the cluster ground. Should be zero "drop" between the two

On the hot side suspect anything. How tight and clean is the sender to wire connector? Possibility of poor connection in bulkhead connector? Or did you do the test resistors right at the engine sender connector?

Another difference:

You tested gauge, was it running or stopped? "It might be" that with system running at charge IE 14v instead of 12....something., that the limiter output is going down instead of up, as it may be cycling differently

 

[Mattax]

Gauge test was with the engine running.
So yes the IVR was also supplying the fuel and temperature gages.

Block has a ground strap to the firewall.

More resistance should result in the pressure gage reading lower. Got the opposite problem here.

 

[Mike69cuda]

Gauge test was with the engine running.
So yes the IVR was also supplying the fuel and temperature gages.

Block has a ground strap to the firewall.

More resistance should result in the pressure gage reading lower. Got the opposite problem here.

I would agree with the more resistance not causing your problem except in the case of a ground loop. If something else is pulling a lot of current, the ground difference can turn into a voltage source for the sender loop. Not highly likely, but all the unlikely stuff happens to me.....

 

[67Dart273]

Mattox, You don't want to make assumptions. Voltage "loops" can do funny things.

EDIT LOLOL Pretty funny ....we thinking alike...and at same time

 

[Mattax]

Thanks guys.
I can certainly measure Voltage differnce between the instrument panel ground and the block.
I can also test the gage operation with the 22 ohm resistor grounded to the block rather than the firewall.

I'm not too hopeful of a ground loop being the issue because the termperature gage also is in hte block, but as you've both said ground loops will do weird thing...

In the meantime I'm coming up with pieces to test the sender from Bill Rolick (short of installing it)

 

[Mike69cuda]

Yep, them oil senders are a pain for access.

 

[Mike69cuda]

It seems like when I can’t figure something out, it is often a ground. I can’t think in negative logic....

 

[rustycowll69]

I had a 79 300, which read low oil pressure on dash gauge. It kinda scared me, so I thought I'd screw in a mechanical fitting and check it. Never got that far, because when I pulled the electric sender out of the block, I could see a hanging booger of Teflon tape was covering the hole in the sender like a little flappy check valve. If you insist on using Teflon tape, hold it back a few threads from the bottom of the fitting, so it doesn't get in the way.

 

[Dana67Dart]

Ok matt this is going to sound crazy but...

Run a jumper from your fuel tank sender to the terminal under the hood for the oil pressure sender, your tank is showing about 1/2 tank, oil pressure gauge should show about the same if he gauge is functioning correctly.

Also unless I did my math wrong you are dissapating about a watt in that 1/4 watt resister.

5 volts / 22 ohms = .23 amps

.23A * 5V = 1.15 watts

Here is what I don't understand...

For the gauge to read high it would need lower numerical restance. Like 10 to 15 ohms.

The only way to lower the resistance value of the sender would be to put some resistance in parallel to the sender, like a partial short to ground.

Parallel restors create a restance lower numerically than the lowest resistor. So if you had a 22 ohm sender and a 50 ohm short to ground your resultant resistance would be in the range of 15 ohms

 

[RedFish]

There is another possible cause. You would have to pull that gauge out of the panel to see it. In the beginning, all 3 of these thermal gauges were different. Main difference was in the slot where needle is attached to bi-metal beam. There was a design flaw in the oil gauge that had to do with needle weight and gravity. My guess is Stuart Warner never tested them positioned same as they are in the car. How they corrected this? The temperature gauge design became the oil gauge too. They simply put the oil gauge screen on it, adjusted it differently and tweaked the chit out of the needle stops to make it work right. I have no idea exactly when the change happened or how many 67 model rally panels got the flawed design oil gauge.
By the way, top end oil pressure does fluctuate. Just right low oil pressure would make a oil warning lamp flicker with their i/o sender. Thermal gauges can't respond to resistance changes so quick.

 

[Mattax]

@RedFish I have a fair number of photos of the gages when they were out last year. If you recall at that time I tested them mounted on the board with a power to the IVR from the car battery and resistors. Would there be any clues from last year work?

@Dana67Dart I think the easier check of the wiring may be to swap the temperature and oil pressure sender hookups. Yes?

For the gauge to read high it would need lower numerical resistance. Like 10 to 15 ohms.

The only way to lower the resistance value of the sender would be to put some resistance in parallel to the sender, like a partial short to ground.

Bet it is possible. There seems to be a few different ways pressure senders have been made. Really can't be sure what the Echlin design has inside. Some are pretty complex and if damaged inside it wouldn't take much to find a parallel path. I *thnk* the path would be more likely to show up at 14 Volts than whatever reference the ohm meter is using - probably a 1.5 volt AA battery.

FWIW one sold by BRE is stamped Pat 4079351
That's a 1978 patent that if I understand it correctly allows the manufacturer to use adjust the final build for the application. patent US4079351A - Pressure responsive sender - Google Patents
An much older design was patented in 1958 by King Seely. That one is mentioned in the unit section of 1970 Chilton's, but doesn't mention which car companies used it when. A bit less complicated.
US2846549A - Pressure indicating device - Google Patents

Chrysler's 1960s sender might have been yet another variation.

edit. FWIW. Scott Smith Harms post some photos of original on a '70 in this thread
Anyone have a picture of a correct 340 oil pressure sender?

 

[Dana67Dart]

@Dana67Dart I think the easier check of the wiring may be to swap the temperature and oil pressure sender hookups. Yes

Well there is that!

 

[Mattax]

Chrysler's 1960s sender might have been yet another variation.

I did some digging into the Chrysler parts books.
The oil pressure senders were changed March 1971.
Gauges installed up to March 1971 use sender 2495 222.
Gauges after that use sender 3488 609.

edit: I was prompted to check the gauge section, and the oil gage part numbers for the A and E bodies are the same throughout the '70 and '71 model year.
So really have no idea why the book calls out a different gauge part number for service replacement for cars made before and after March

 

[Mattax]

Set up a pressure equals resistance test for the sender I bought from BRE.
I haven't crunched the numbers but they are all over the place.
So either there is flaw(s) in the test, something I don't understand, or the sender is no good for this application..

One of the few times the resistance was in the ballpark.

 

[Mattax]

Test results graphed.
Even ignoring the readings outside the center of the range, it seems way off of spec.
For reference AMC '85 FSM allowed plus minus 4 ohms at 0 psi, and 0.5 ohms at 80 psi. Assume Chrysler's allowance was similar.
Bad unit? Wrong sender calibration for this application?

This might be useful too. A common replacement according to the cross reference charts is Standard Motor Products PS-59.
As luck would have it there was a snip from the catalog on the web showing ohm range of the sender at two pressures.

Blue Dashed lines connecting the 8 psi to the 90 psi spec. show where the in-between resistances might be.