Restoring Air Core Temp guages

[posplayr]

My GS 1100EZ/ED temp gauge tends to bounce, I have 4 of them and they do it to varying degrees. Fortunately I have one that is nice but was looking to see if I could fix an otherwise good guage.

What happens is that at idle the guage is relatively stable, but as the engine revs up the guage starts to bounce. I put a scope on the sensor and found that the ignition noise causes bounce and the guage is very responsive to the noise. From the factory the gages had oil in them but they seem to loose that and so you get bounce. Except for the noise they are still accurate so if you can average through the noise no issues

I did check to see if it was vibration induced and it is not. You do this by simply grounding the temp sensor wire and the bounce mostly goes away. But at low temp where there is highest resistance at the sensor the bounce is high.

I have considered electronic filtering and that may be an alternative if I cant get any oil into it. So any body done this successfully? The only opening for getting oil back in is the small space between the shaft and the plastic housing the rotor spins inside. I'm worried about filling with l as if it leaks out it will make a mess.

Please note this is an Air Core gauge and not the Thermal Bimetalic spring type that came on earlier models.

I did find this link using google where an old Corvette guage is improved using oil injection. Unfortunately you can see the back of the guage it is completely covered in wire windings

http://www.lbfun.com/warehouse/tech_info/gauges & instruments/C2_fuel_gauge_and_sending_unit.pdf

http://www.lbfun.com/warehouse/tech_info/gauges%20&%20instruments/C2_fuel_gauge_and_sending_unit.pdfhttp://www.lbfun.com/warehouse/tech_info/gauges & instruments/C2_fuel_gauge_and_sending_unit.pdf

Some points on ?damping? in the gauges - Jason Oliver
The way the dampening is done inside the gauges, fuel and
battery/ammeter especially, is in the air core movement itself.
The way I ?fix? this is to inject a non-hardening grease/gel into the air
core movement and it slows the needle ?bounce? down quite a bit
without messing with calibration.
There are 2 problems and the 63/64s are worse than others. One is
that the sending units are "bouncy?. Surely someone has a picture of
one taken apart to show how they work. Then, the 63-64 gauges
(fuel and battery) are touchy. The gel I use is injected into the
needle shaft and it helps a lot on the jumpy gauges. This is a GM fix
that has been around for awhile; just not well advertised...same
principle as liquid filled gauges (helps dampen needle bounce).

http://www.lbfun.com/warehouse/tech_info/gauges & instruments/C2_fuel_gauge_and_sending_unit.pdfhttp://www.lbfun.com/warehouse/tech_info/gauges & instruments/C2_fuel_gauge_and_sending_unit.pdf

 

[Guest]

Good luck
We usually leave this kind of thinking to you

I know Ed has messed with guages.

 

[rapidray]

Holy CRAP Jim, get a life!!! LOL! Ray.

 

[Billy Ricks]

Holy CRAP Jim, get a life!!! LOL! Ray.

No schitt. I do admire your skills though Jim.

 

[posplayr]

No schitt. I do admire your skills though Jim.

Talked to a guy at work; I will be able to squirt some electronic oil into it.

 

[Billy Ricks]

Talked to a guy at work; I will be able to squirt some electronic oil into it.

You'll have to excuse my ignorance, is the electronic oil non-conductive or what? Maybe even conductive?

 

[posplayr]

You'll have to excuse my ignorance, is the electronic oil non-conductive or what? Maybe even conductive?

Sorry Billy: I was using "electronic oil" as an oxymoron for electronic damping (i.e. capacitive). Just got a selection of electrolytic from radio shack cause I was too lazy to compute the response even though I solved the voltage divider equations

I'm hoping to be able to mount one on the back of the gauge and get all the required damping.

 

[Billy Ricks]

Sorry Billy: I was using "electronic oil" as an oxymoron for electronic damping (i.e. capacitive). Just got a selection of electrolytic from radio shack cause I was too lazy to compute the response even though I solved the voltage divider equations

I'm hoping to be able to mount one on the back of the gauge and get all the required damping.

Thanks for the explanation, I should have known something was up the way you had electronic in bold face. I even googled electronic oil thinking it was something I never heard of. So you're trying to cut off high frequency at some level or have a slight amount of reserve power? Sorry for the dumb questions, just trying to learn something here.

 

[posplayr]

Thanks for the explanation, I should have known something was up the way you had electronic in bold face. I even googled electronic oil thinking it was something I never heard of. So you're trying to cut off high frequency at some level or have a slight amount of reserve power? Sorry for the dumb questions, just trying to learn something here.

Billy

It is basically a form of filtering. The voltage across a capacitor is related to how much charge in on the plates. That rate of change in voltage is proportional to how much current can be pushed into it compared to it's capacity (a small cap will fill faster han a big cap), What that means is that the speed of change of the sensor voltage can be controlled by limiting the rate of current to it. The current is limited with resistance.

It is mathematically equivalent to the hydraulic accumulator I made to stop the mechanical oil pressure guage from pulsing due to engine pressure pulses.

I just did a test a few minutes ago and a 1000 uF cap completely eliminated the voltage bounce 400 mV peak to peak (I have an oscilloscope) by using that cap. There is still bounce in the rest of the coils so I need to put a cap across the input as well. That requires pulling my fairing back off but since the voltage improved so well at the sensor I'm pretty sure it will work.

more reading on RC (resistance and capacitance filtering).

http://en.wikipedia.org/wiki/Low-pass_filter

More on RLC (resistance , inductance and capacitance)

http://en.wikipedia.org/wiki/RLC_circuit

I derived the filter equations but since i have no way to directly measure the inductance, I'm doing by trial and error. In noise reduction it is pretty standard to put a capacitor between a noise point in a circuit to ground.
It works because of the RC filtering concepts.

RLC generates a 2nd order system otherwise know as a "tank circuit" (like ringing a tank and it wants to oscillate at a certain frequency depending upon the RLC values. It has a preference to accentuate or retard certain frequencies. This is analogous to our motor cycle suspensions if the shocks are bad (i.e. when you have bounce).

In 2nd year electrical engineering you learn to write the "mesh equations" for any combination (parallel/ serial and compound) of RLC's and derive the transfer functions from inputs to various outputs. So there are plenty of variations which those two links just touch on. Now adays circuits are so complex you can't real always derive the equations and so a computer will be used to numerically solve for the circuit conditions as a function of different initial conditions and driving functions. So suffice to say after the Junior year in EE you have worked this stuff backwards and forwards and in various ways (at least that is the was it was 30 years ago).

Jim

 

[Guest]

Billy

It is basically a form of filtering. The voltage across a capacitor is related to how much charge in on the plates. That rate of change in voltage is proportional to how much current can be pushed into it compared to it's capacity (a small cap will fill faster han a big cap), What that means is that the speed of change of the sensor voltage can be controlled by limiting the rate of current to it. The current is limited with resistance.

It is mathematically equivalent to the hydraulic accumulator I made to stop the mechanical oil pressure guage from pulsing due to engine pressure pulses.

I just did a test a few minutes ago and a 1000 uF cap completely eliminated the voltage bounce 400 mV peak to peak (I have an oscilloscope) by using that cap. There is still bounce in the rest of the coils so I need to put a cap across the input as well. That requires pulling my fairing back off but since the voltage improved so well at the sensor I'm pretty sure it will work.

more reading on RC (resistance and capacitance filtering).

http://en.wikipedia.org/wiki/Low-pass_filter

More on RLC (resistance , inductance and capacitance)

http://en.wikipedia.org/wiki/RLC_circuit

I derived the filter equations but since i have no way to directly measure the inductance, I'm doing by trial and error. In noise reduction it is pretty standard to put a capacitor between a noise point in a circuit to ground.
It works because of the RC filtering concepts.

RLC generates a 2nd order system otherwise know as a "tank circuit" (like ringing a tank and it wants to oscillate at a certain frequency depending upon the RLC values. It has a preference to accentuate or retard certain frequencies. This is analogous to our motor cycle suspensions if the shocks are bad (i.e. when you have bounce).

In 2nd year electrical engineering you learn to write the "mesh equations" for any combination (parallel/ serial and compound) of RLC's and derive the transfer functions from inputs to various outputs. So there are plenty of variations which those two links just touch on. Now adays circuits are so complex you can't real always derive the equations and so a computer will be used to numerically solve for the circuit conditions as a function of different initial conditions and driving functions. So suffice to say after the Junior year in EE you have worked this stuff backwards and forwards and in various ways (at least that is the was it was 30 years ago).

Jim

That's a thread killer
:hand:

 

[posplayr]

That's a thread killer
:hand:

For future reference only :-k