[alfaparticle]

This is a spill-over from the fuel gauge anti-slosh thread.
Rich - gprocket posted

Finally, i know (i think) that with oil pressure some sending units work in reverse.

I didn't mean to hijack this thread - would it be better to start a new one? For that matter Ed, is this something you have the time and interest in doing. I'm happy to share the cost...

Rich and I had a conversation yesterday evening. He will send me Spider and GTV oil pressure gauges and I will determine what electrical signals they need from the sender.

In the meantime I dug out 3 old senders and I put together a test rig to check the calibration of them. Two of the senders were junk so I cut them up to see how they worked. Oil pressure moves the copper diaphragm which acts on a button with a linkage which swings a wiper arm over rheostat type arrangement. The pictures show the parts for the sender. In the out of focus shot I am pushing on the button to move the wiper to the mid position. The last shot is of the test rig that uses parts from a fuel injection test kit. The pressure source is my compressor.

I will post information about the gauges when I have tested them.

 

[m3_tom]

I suspect that OE senders have a wide variance in resistance value with respect to any given pressure, and the gauges may have similar issues as far as reading displayed for any given resistance value at the sender, thus rendering the whole system a vague approximation of actual oil pressure at best. Has anyone ever gutted a stock gauge and retrofitted a mechanical gauge movement?

 

[alfaparticle]

suspect that OE senders have a wide variance in resistance value with respect to any given pressure

The measurements that I have made over the years support your suspicion. It is possible that they are reasonably accurate when new but lose accuracy during service. I can think of several likely causes of this.

and the gauges may have similar issues as far as reading displayed for any given resistance value at the sender

That is possible but less likely. The gauges live in a much nicer environment and electrical gauges are usually reasonably stable.

I will soon be in a position to make measurements that will replace suspicion with data.

 

[tdskip]

Subscribed - thanks for doing this.

 

[gprocket]

... Has anyone ever gutted a stock gauge and retrofitted a mechanical gauge movement?

Tried. I got kinda close but it was a real pain. It then occurred to me that electric oil pressure systems have worked just fine for many applications for many years. There is no reason this can't be the case for our cars. Hence Ed and my conversation to try to further the knowledge base.

 

[Gubi]

The main failure mode of the senders seems to be that oil can leak around the diaphragm. Then you get pressure in the "bell" portion and the readings get thrown off.

I did the trick several years ago where you drill a tiny hole in the casing to relieve pressure. The senders on the Spider and Milano have been fine since.

http://www.alfabb.com/bb/forums/164-168-1991-1995/52549-bad-oil-pressure-sender.html

 

[alfaparticle]

The main failure mode of the senders seems to be that oil can leak around the diaphragm. Then you get pressure in the "bell" portion and the readings get thrown off.

One of the two that I opened up was contaminated with oil. It was on the resistive windings and had caused the erratic behaviour. I tried drilling a vent hole before I opened it up and it did not help. The other one was dry and clean inside. The wiper had pushed and pulled the fine resistive wire leaving gaps and dead spots.
There appear to be several failure modes and I would have to open up a lot of failed ones to say which is the dominant one. Drilling a vent hole might help and it definitely should not hurt.

 

[65Sprint]

thanks for doing this.

+1

The donation of knowledge and time, as well as the idea and support to expand the sender/gauge testing effort are extremely generous gestures. Thank you both. If you need addtional senders or gauges (fuel and oil), let me know.

 

[alfaparticle]

A failure mode

Here is a picture of the resistive winding from the clean but malfunctioning sensor. This is a clear failure mode.

 

[Gubi]

Interesting...looks like that may be one of the older models. The newer ones I've seen use a carbon resistive pad like in the AFM. Not sure if both types are still being made.

 

[alfaparticle]

Mine are made by Veglia. The youngest Spider or GTV that I have parted was an '82 so my spare senders are most likely 70 - 82 vintage.
What is the end to end resistance of your new style?

 

[Gubi]

Mongo dunno...that pic is from the thread I linked to above.

The part numbers for the ones I've seen are valid for '69-'94, so my assumption is that the resistance range should be pretty much the same on all of them. But don't quote me on that.

In any case, accuracy isn't really important from the gauge (provided you have a good, working low pressure switch & light). You mostly just want something repeatable enough that you can note any changes or trends.

 

[gprocket]

From Classic Alfa:

"Oil pressure sender veglia early":


"Oil pressure sender to fit late Series 3 and all Series 4 Spiders with the 'pod' style instruments":


"Oil pressure sender to fit all 105 Series cars with Jaeger gauges and all cars with Veglia gauges from 1968 onwards. Also fits Montreals.":



From Centerline Alfa:

"Oil Pressure Sending Unit for Veglia Gauges 1965-68":


"Oil Pressure Sending Unit for Gauge 1971-95, All":


Clear as mud, don't you think...

 

[m3_tom]

Tried. I got kinda close but it was a real pain. It then occurred to me that electric oil pressure systems have worked just fine for many applications for many years. There is no reason this can't be the case for our cars. Hence Ed and my conversation to try to further the knowledge base.

I agree with your position, but with one proviso- some brands of senders do not have a good reputation for consistency (I used to make a living selling parts for vintage Jaguars, oil senders were not so good). I suspect that Veglia is not the most respected producer of these things. If VDO made something with a similar resistance to pressure curve, I would be on board with fitting one. If all you want is relative consistency to spot trends, stock Alfa is probably all you need. Who knows where they come from now, could be Chinese knock offs.

 

[alfaparticle]

Rich sent me oil pressure gauges from a Spider and a GTV. They clearly require different senders.
Spider
ohms 8; 130; 340
psi 114; 57; 0

GTV
ohms 0; 44; 88
Bar 0; 4 ; 8

The GTV sender has a lower resistance and it is inverted wrt the spider sender. That is, the Spider sender has low resistance = high pressure and the GTV sender has high resistance = high pressure.

I hooked up Rich's Spider gauge to my spare sender and at mid scale 61 psi injected into the sender gave 57 psi indicated on the gauge.

 

[gprocket]

Excellent!

The gauges that I sent were from a 1968 Euro GTV (large, built into the Tach) and a 70s Spider (small round gauge).

Here are the data charted (I converted bar to PSI). Interesting that the Spider gauge is close to linear and the GTV gauge is spot on...






Also very interesting that the Spider gauge is as accurate as it is...

 

[Del]

A few of us with 164's are using the inexpensive SunPro CP7577 oil pressure sender, recalibrated with a 22 ohm resister wired in line. Suppose to be spot on at 3000 rpm (generally cruise rpm). Costs about $15 off eBay or whatever. Using the standard washer, it screws in nicely enough (slightly different threads) without any damage to anything and seals well. No leaks.

 

[Gubi]

A few of us with 164's are using the inexpensive SunPro CP7577 oil pressure sender, recalibrated with a 22 ohm resister wired in line. Suppose to be spot on at 3000 rpm (generally cruise rpm). Costs about $15 off eBay or whatever. Using the standard washer, it screws in nicely enough (slightly different threads) without any damage to anything and seals well. No leaks.

Just as a note, the Sunpro is 1/8" 27 TPI tapered thread and the original sender is M10x1.0 straight thread. They're close enough that some folks have gotten it to work with some teflon tape, but personally I would NEVER EVER recommend this, especially since you're screwing the sender into aluminum.

It's your car, though.

 

[gprocket]

Del - great tip! Roadtrip has done his homework as usual. http://www.alfabb.com/bb/forums/164...re-gauge-sunpro-sending-unit.html#post1161390

Taking the data from his post I charted the Sunpro Output with the Spider gauge input:



The trend lines are sort of close. By my reckoning you would need a 60 Ohm resistor to match the sender to the gauge at 60 PSI. But you would be off the further you moved away from that point. So the question is: Can one create a circuit that would match the rate (slope) of the gauge to the sending unit? Now you'd have something.

Dr. P, the ball is in your court...

 

[alfaparticle]

Interesting that the Spider gauge is close to linear and the GTV gauge is spot on..

There is an uncertainty when you read a small gauge with a fat pointer. The Spider gauge may well be linear.