Hi Finger.
I would like to know how hard would it be to modify glow plugs in the duramax to be used as a pressure sensor to read/monitor cylinder pressures.
I think if we could monitor the pressures, that would really help to understand whats going on during the compression ignition cycles.:)

I'm not Fingers, but what you're talking about is done already. The probes are available, and although expensive, are a lot less expensive than the hardware required to take the measurement....very high speed acquisition that's linked to the crank position sensor.

There are a handful of people out there who know what the pressures are.

The sensors already exist. You can buy a hi speed, high end optical coupled sensor rolled into a glow plug for < $1000. The DAQ will depend on the granularity of what you want to measure, and how many channels. What is typically used on a engine dyno situation is usually an 8 channel high speed capable of outputting data in both real time and snapshot mode. They are neat in that the pressures can be shown similar to an oscilloscope's display. That will cost you >>$5000 to get set up.

What I have on the bench is a lot more conservative. I have This (http://www.measurementcomputing.com/cbicatalog/cbiproduct_new.asp?dept_id=413&pf_id=1572) on order and have been playing with it's little brother. Using these in snapshot mode with an external trigger, I have been looking at the injector pulse signal on my LLY. (the current DAQ is too slow) I should be able to do the same with a pressure sensor.

Don't ask me why I am spending $1K on this crap. It's just something I do.

Jon

What are you planing to use to pickup crankshaft angle?

I will use the DIO on the board to track the Crank Signal, thus giving me a scaling and zero reference for the data. The trigger can be most anything, even the injection pulse. Just so the period of interest falls in the scanning window. Sampling more than two strokes and tracking the CPS means the trigger timing is not critical.

The new board can do 50kS/sec per channel up t0 200kS/sec combined for all channels. The 32K FIFO limits my sample window. Roughly 4 Bytes per sample. At 5000 RPM you need 30kS/sec to get a 1* resolution and would fill the buffer in ~1/4 second. That would be about 20 revolutions at that speed. Shrinking the window increases the resolution. The sample rate is variable, and could be adjusted for RPM. I will probably use the built in counter connected to the CPS to limit the window to 4-5 revolutions.

In continuous burst sample of 4 revolutions and transfer to the PC, I could strobe refresh the display every second or two. That includes a disk write of the data for logging.

Not a perfect solution, but I am on a budget.

To bad you live so far away. I would let you borrow this (http://www.astro-med.com/products/d18.html) for the afternoon.

The Duramax crank trigger is only every 6 degree except when going by TDC as it has missing teeth then, so the resolution will not be any better than that unless I missing what you using. By injecting the Cam position sensor you can tell when your on the combustion event but I do not see where that helps any as it's not very accurate for true location. I guess if your just looking at trends and peak pressures it will do the job just fine. I know better cause your budget always seems to get bigger when needed ;)

TDon't ask me why I am spending $1K on this crap. It's just something I do.

Because "I think I can"!

I was thinking along the same lines but wasn't able to get to that level of affordability....I was in the $1600 range for the DAQ and my interest waned...

Like to hook up with you on this if at all possible. :o:

The Duramax crank trigger is only every 6 degree except when going by TDC as it has missing teeth then, so the resolution will not be any better than that unless I missing what you using. By injecting the Cam position sensor you can tell when your on the combustion event but I do not see where that helps any as it's not very accurate for true location. I guess if your just looking at trends and peak pressures it will do the job just fine. I know better cause your budget always seems to get bigger when needed ;)


Timing accuracy/resolution is a function of sample rate. The missing tooth is the marker for TDC. Even if there was only one tooth, it would be sufficient to scale the data to the crank angle. The CPS, I am finding, can vary as much as 2-3 CADegrees from engine to engine anyway. Might be sensor to sensor. I don't know. At least that is on my two engines so far. In other words, TDC is suspect anyway.


The shortcomings of this setup are many. 16bit resolutions for instance, are no good on a 3000 PSI range sensor. That's ~1 bit/psi. But more importantly is the resolution of the D/A converter. If you stack the gain and range errors on this DAQ, you are looking at up to a 10% variance. (Typically much less) That sucks. At least the results will be repeatable well within 1%. So engine to engine and run to run should be consistant.

Having said all that, this is simply a tool. It should be sufficient to give me the insight into cylinder pressures I need to tune my truck for a long(er) life. As I get closer to finishing my engine, I need to get this project off the bench and onto some real motors.

Timing accuracy/resolution is a function of sample rate. The missing tooth is the marker for TDC. Even if there was only one tooth, it would be sufficient to scale the data to the crank angle. The CPS, I am finding, can vary as much as 2-3 CADegrees from engine to engine anyway. Might be sensor to sensor. I don't know. At least that is on my two engines so far. In other words, TDC is suspect anyway.


The shortcomings of this setup are many. 16bit resolutions for instance, are no good on a 3000 PSI range sensor. That's ~1 bit/psi. But more importantly is the resolution of the D/A converter. If you stack the gain and range errors on this DAQ, you are looking at up to a 10% variance. (Typically much less) That sucks. At least the results will be repeatable well within 1%. So engine to engine and run to run should be consistant.

Having said all that, this is simply a tool. It should be sufficient to give me the insight into cylinder pressures I need to tune my truck for a long(er) life. As I get closer to finishing my engine, I need to get this project off the bench and onto some real motors.

Jon

For what your asking for I'm sure it will work but you are correct about the cam position sensor and crank sensors being wrong. I have measure over 6 degrees variation in them so far. The system we work with is much better but then it was 20 grand for the unit plus 1200 for each of the 8 cylinder pressure sensors for the Duramax. One thing you should know is the crank sensor is adjustable.

too bad all this info stays secret :(


glad to see what you're working on Jon, all the power to you! :)