Pressure Sensing Glow Plug Electrical Connector

[Fix_Until_Broke]

This might belong in the For Sale/Wanted section, but I thought it would be more beneficial in here given the usage, so moderators please move this if you necessary.

Does anyone have or know where to get the electrical connector for one of the pressure sensing glow plugs used on the new CR engines and possibly have a pin out of it?

I'm looking more seriously at adapting one of these to use in an ALH

 

[dieseleux]

I have one, i buy new glowplug with harness to do this on my ALH engine!
This is a simple sensor 0,5-4.5v for 0 to 200bar, 5khz bandwith signal.
The harness have 4 wire, one big... easy is glowplug, 3 small... Brown always ground on vw, i dont remember the last 2 wire color (i think is yellow and white), +5vdc supply and output signal.
The glowplug have same body size but much longer..., same tread size, same seal tapper angle, tip are a little bit longer and thinner.
I check and i need to do some test on old ALH head, the tread is M10x1 and is possible to do a adaptor in steel material with M12x1 outter tread to screw in modified head.
The best place is cyl 3 to match with needle lift.




Dieseleux

 

[Fix_Until_Broke]

Yes, #3 would be minimum so then you can have accurate measure of ignition delay.

I found someone who has a complete CR wiring harness with the connectors I need, but I'm not sure I want to tear up a good harness for a few plugs - seems like a bit of a waste.

 

[dieseleux]

This type of glowplug have independant harness, one per glowplug.
I new harness have very long wire, look like 5 or 6 feet long!



Dieseleux

 

[JFettig]

this is a ALH(steel autolite junk) glow plug and the CR glow plug side by side.

GP tip to 1st shoulder:
CR: 1.065"
ALH: .950

shoulder to first thread:
CR: 2.25"
ALH: .865"

Diameter below thread:
CR: .335"
ALH: .325"


CR glow plug pinout:
1. power (3v if I'm reading this bentley notations correctly)
2. signal to ecu
3. ground
4. glow from J179 automatic glow time control module

Just pinned out the harness:

I'm finding more wrong with this harness, it was given to me with this in mind, I won't ever get to doing it with my car. I will plug it into my spare ecu later and power it up to check the supply voltage to the sensor.

In an AHU head:

PD130 head:

 

[Fix_Until_Broke]

Jon - thanks for the pictures and measurements. That's quite an interesting connector they have on those. The Beru document (generic, I know), describes a 5v excitation. 3v is more common in really low power stuff (wireless sensors powered by vibrations and cell phones) where power consumption is critical. Hopefully it's 5v otherwise I'll have to build a voltage regulator and calibrate the sensor (I guess I'd want to calibrate the sensor anyway...).

I'm trying to figure out how I can make an adapter so the sensor can screw into an existing ALH (and maybe PD by the looks of it) head without having to go wild with machining a head to specifically accept this GP or an adapter.

I don't think I care if the GP tip does not protrude into the chamber, but to build an adapter that will hold the new GP in an OEM ALH GP hole will locate it back out of the hole a little over an inch and create a bit of a dead zone in front of the sensor. Not sure that it's a big deal or if it would cause any significant measurement problems or not?

I guess the other option is to machine out the M10x1 GP hole in the head to M14x something and make a separate adapter tube for the long CR GP and a simple threaded bushing for the ALH GP. That would give enough wall thickness to retain either GP and not be too difficult to machine. It's something that I'd do on my own car (with a steady pressurized compressed air source in the cylinder to blow out the chips), but probably not on someone else's.

Any other thoughts on how to use an OEM sensor?

 

[m1ketdi]

Given the "ringing"/harmonic issues that can be seen with poorly designed adpators for pressure transducers of the kistler kind I would say spacing out the glow plug any more than the stock install style would be avoided if at all possible. Although the way I see these plugs measuring pressure you may well be able to cope with that.

Maybe you just need to find a different OEM sensor, has someone got one the right length?

 

[newlitemotorist]

Could you take an original glow plug, machine the top off just above the threads and remove the guts forming a sleeve? If you can do this you might be able to do some light machining on the outer section of the smart glow plug and the inner sleeve you just made so you can create a tight tolerance between the two. Heat the sleeve up and press it onto the new glow plug. If there is enough material for machining I dont see why this couldn't be done.

 

[Fix_Until_Broke]

newlitemotorist - I understand what you're suggesting, but don't want to put any undue strain/compression on the new glow plug since the pressure sensing mechanism needs to travel through the area where this sleeve would be - see the cutaway picture in the first post and the seal near the end of the body. Too tight and the sensor won't work, too loose or different thermal expansion coefficients and I'll have a hole in my hood

I made the below adapter years ago for performing compression tests - I put a 350 bar pressure sensor on the end of it once and let the engine idle and the adapter and pressure sensor were so hot in ~5 seconds that they started smoking and it destroyed the pressure sensor. I think the compression/combustion going up through the glow plug and to the sensor caused it to heat up abnormally.

That's my main concern with using a long adapter with the CR glow plug in the ALH head...all the compressible flow/combustion occurring up in the dead volume in the GP hole.

 

[dieseleux]

The pressure wave of gaz combustion travel in the tube!
In CR glowplug the gaz keep in cylinder and the pressure are communicate by the move of internal heat element and keep the high temperature in head by contact of 45degree taper seal.
This glowplug are very sensitive at the top, a light lateral pressure on the part over copper ring read by sensing element easily.



Dieseleux

 

[newlitemotorist]

newlitemotorist - I understand what you're suggesting, but don't want to put any undue strain/compression on the new glow plug since the pressure sensing mechanism needs to travel through the area where this sleeve would be - see the cutaway picture in the first post and the seal near the end of the body. Too tight and the sensor won't work, too loose or different thermal expansion coefficients and I'll have a hole in my hood

I did not consider the pressure sensing mechanism's travel inside the tube so yes that would probably not work.

What about cutting threads down the shaft and then threading on just a threaded sleeve? Sorta like a heli coil or a quick-sert? The sleeve would only need to be as big as the number of threads you would need to secure it within the head which would solve the issue's mentioned above and the problems you are concerned about.

Just thinking out loud here since you seem to have the resources to make something like this happen.

 

[DPM]

^ that might work, you'd have conical seals each end of the bush then. I'd consider a support tube/collar above it in this case tho, there's a lot of GP body swaying around in the wind up above...

only issue might be unscrewing the bush if you ever needed it removed- permanent Locktite onto the GP perhaps?

 

[newlitemotorist]

only issue might be unscrewing the bush if you ever needed it removed- permanent Locktite onto the GP perhaps?

I thought about loctite as well but one of the ways you get permanent loctite to release is to hit it with heat so that may not work in this case. How about a skim coat of hard setting epoxy such as jb weld??? Only downside to this is that it would be a "permanent" solution.

 

[Kiggly]

From the original post, has anybody found a source for the OEM connectors? I'm looking to adapt these to another project and I'd rather buy OEM connectors than try to hunt down somebody parting out a harness or something from a scrapyard.

Great information here!

Thanks,
Kevin

 

[Fix_Until_Broke]

I did not consider the pressure sensing mechanism's travel inside the tube so yes that would probably not work.
What about cutting threads down the shaft and then threading on just a threaded sleeve? Sorta like a heli coil or a quick-sert? The sleeve would only need to be as big as the number of threads you would need to secure it within the head which would solve the issue's mentioned above and the problems you are concerned about.
Just thinking out loud here since you seem to have the resources to make something like this happen.

I like this idea, but am not sure there's enough wall thickness in the CR GP body to accommodate...
M8x1.0 external thread
Major diameter = 0.307"-0.314"
Minor diameter = 0.260"-0.271"

The diameter of this area on the CR plug is 0.335" and I'm just not sure there's enough wall thickness left after cutting a ~0.035" deep thread in there. Maybe?

A M8x1.0 Helicoil insert has a max OD of 9.70-10.25mm and by definition the pitch has to be the same so it might just thread right in to the existing GP hole. The tap major diameter maximum is 9.42mm though which seems a little small for at M10x1.0 so it might be a little sloppy?

Handy technical info from HeliCoil

So - maybe if I can cut the M8x1.0 threads on the large side (like an M8.5x1.0 - OMG it actually exists as a tap!) to minimize the amount of material removed from the glow plug, the M8x1.0 helicoil will grow enough to fit snug in the existing M10x1.0 thread in the cylinder head.

I know a guy with a plunge EDM that might be able to take an existing M8x1.0 die and turn it into a M8.5 pretty easily...I don't think that I could spread the die enough (1.5mm) on the circumference without breaking it.

 

[Fix_Until_Broke]

From the original post, has anybody found a source for the OEM connectors? I'm looking to adapt these to another project and I'd rather buy OEM connectors than try to hunt down somebody parting out a harness or something from a scrapyard.

Great information here!

Thanks,
Kevin

I have not - I'd start with getting a connector from the scrapyard and seeing what markings are on it to maybe help you find the company who makes it for Beru. If you find it, please let us know!

 

[Fix_Until_Broke]

It looks like they're made by Hirschmann Automotive which has a typically useless website for the average consumer...

Below is what I found with a quick Google search...

 

[Fix_Until_Broke]

and one zoomed in on the connector

 

[Fix_Until_Broke]

The highlighted part makes me a little uncertain - does anyone know what kind of phase offset is applied in the ECU? Or at least approximately how much it is?

 

[flee]

It looks like they're made by Hirschmann Automotive which has a typically useless website for the average consumer...

Maybe one of our Austrian members could call or visit and get some samples...

 

[dieseleux]

For basic test i do and you do, dont butter with that!

In case of fast pressure wave (like pulse or square wave) the 5khz filter react like time delay line.
At 100hz the delay are insignifiant
At 1khz some delay appear
At 2.5khz delay are significant
And at 5khz the delay are 1/2 wave.

You have also some attenuation of signal versus frequency to take in the calcul.

All that to make good 0-5khz bandwith sensor, if you just read without compensate, you have good sensor for 0-2.5khz and less significant signal for 2.5 to 5khz!

At 6000rpm you have 50 pressure wave per seconde on the sensor and each pressure wave have 1/200sec width with rise time of ingnition pressure aprox 1/1000 to 1/4000sec (1 to 4khz), the sensor are very good bandwidth to see the pressur wave.
At 6000rpm, 5khz resolution is 7.2degree, this is at this speed you need phase compensation to mesure timming but i think you dont work at this speed.




Dieseleux

 

[Kiggly]

Ouch, unfortunately the center of that notch filter is right at the primary knock frequency for a low 70mm bore engine. More common 8Xmm range bore engines will still see significant knock signal attenuation as they're in the 6kHz range. This is, of course, talking about using the sensor on a gasoline engine.

Looks like I'm going scrapyard hunting.

 

[Fix_Until_Broke]

Dieseleux - I think the notch filter is generally always active due to the natural frequencies of the heating rod, bellows and strain diaphragm, hence the implementation of the notch filter to attenuate those frequencies so the signal isn't noisy. That digital filter takes time thus imparting a "constant phase displacement" (shift) that I highlighted.

At 1000 RPM you have 0.060 sec/rev (16.66Hz) of which you have maybe 1/3 of that which is useful data (~180 crank degrees, 0.030 sec, 33.33Hz) and of that to catch PCP's with any kind of accuracy you'd want at least ~2 degree resolution (let's call it 1.8 degree for easy math) which is 0.0003 sec, 3.3kHz. Cut all the times in half/double the frequencies for 2000 RPM, etc.

The below image has the phase shift correction applied and it looks very good. What I don't know is if this phase shift is 1 degree or 10 degrees. I guess as long as the peak values are still accurate (and they should be) the phase is not as important.

 

[dieseleux]

The phase is important if you need to found the exact moment of ignition and ajust the timming at 1 or 2 degree of precision on high rpm range of TDI engine.
The phase is time delay on one frequency.
If you know the delay, is easy to apply by hand a correction factor in degree for each rpm you take.

15 degree after the top dead is sweet spot for PCP, is complexe to explaine but is a good number remember.




Dieseleux

 

[Fix_Until_Broke]

Yes - it's easy to correct for if I know what it is. With the needle lift sensor I was hoping to see what the ignition delay is - I'll be able to get a measurement (I hope), but the actual number will be less than the straight time difference between the needle lift sensor and the pressure sensor due to whatever this lag is.

Kiggly - these are 79.5mm bores so closer to 80 than 70

 

[dieseleux]

The ignition delay are mostly stable on rpm and boost range, the sensor delay are also stable.
Is easy to see the delay at idle to 5000rpm you have between needle lift and ouput signal of pressure sensor.
Is more difficule to see what the delay of pressure sensor only in real for ajust the timming in ecu at the right place.



Dieseleux

 

[Fix_Until_Broke]

Sorry - not meaning to make a big deal out of this. Most important is knowing that there is a phase shift.

If anyone knows what the ECU is doing with signal, I'd appreciate any info.

 

[Digital Corpus]

I have a solid 200 MHz scope so if you guys wanted to visit sunny SoCal you could stop guessing on how to characterize the signal and we could just measure it . Anyhow, I like the look of the progress guys, keep it up .

 

[Fix_Until_Broke]

So do I, however unless you can get to the signal right off the strain gauge before the ASIC with the notch filter, it won't do any good.

The company that has the M8.5x1.0 tap does not have any dies - I'll try and hit up my friend with the EDM machine tomorrow about making a M8.5x1.0 die from a M8x1.0 die

 

[Digital Corpus]

I wouldn't mind grabbing it after the ASIC unless you're worried about the processing delay, if I read the above post correctly yesterday. It'd be a *much* cleaner signal that can be utilized. In order to do what the ASIC does, we'd have to either develop an FPGA (best choice) or run it through other hardware (RC or LC, 2nd best choice) or software filters (worst choice).