Pressure Sensing Glow Plug Electrical Connector

Hmmm - I'll give them a call on Monday. I've got a quote from Optrand for a drop in sensor 1%, temp compensated, etc, just saving up my nickels and dimes. Things like new roof for the house, day care, diapers, etc keep delaying my automotive research pursuits

I have just finished reading through this thread. I like the looks of the Optrand sensor. How much does one cost?
Also I read somewhere that a way to sync the crank signal with TDC is to run the engine up to a specific (max RPM maybe?) RPM then kill the engine and measure the time between the Pressure spike (actual TDC with no combustion) and whatever you are using for a crank sensor. I guess this would need to be saved to a table or applied to a formula to generate accurate crank angles across the RPM range.
BTW, I'm interested in building a standalone datalogger for use in optimizing SOI Timing on a few different Diesel engines such as the BRM in my Jetta and a Toyota 1KZ-TE.

Yes, this is also how we do it, I think usual is 2000 rpm in overrun (engine test bench) and IIRC the offset between the measured peak and geometric TDC is around .8 degree

The Optrand sensor for an ALH (non-functioning) glowplug is ~$1200 at your door in a couple weeks.

Interesting method of determining phase lag between pressure peak and geometric TDC. I wonder how much it changes with RPM? I'd assume that it would get later and later as RPM's increase due to the speed of sound in the chamber (assuming constant VE)?

There's been a few discussions about how to build a standalone data acquisition system to do this. It all depends on how much time and $$ you have to dedicate to it.

I never did call Sensata - E-mail sent to myself to try and remember to do that tomorrow.

Ouch! That's a bit more than I was expecting. I guess I'm going to have to use one of the CR glow plugs. Some of their literature from the late '90's quotes prices for the sensors at $7-9 each. That price may not have included the electronics.

Rub87 said:

Yes, this is also how we do it, I think usual is 2000 rpm in overrun (engine test bench) and IIRC the offset between the measured peak and geometric TDC is around .8 degree

Click to expand...

One good method is to just disconnect the injector from the cylinder that you are measuring pressure in. Either fuel pipe or electrically. Then start the engine and you can find TDC from pressure trace.

Rub87 said:

Yes, this is also how we do it, I think usual is 2000 rpm in overrun (engine test bench) and IIRC the offset between the measured peak and geometric TDC is around .8 degree

Click to expand...

I take it the 0.8deg would apply to your specific sensor. This number could be quite a bit different for other sensor types, correct?

Fix_Until_Broke said:

The Optrand sensor for an ALH (non-functioning) glowplug is ~$1200 at your door in a couple weeks.

Click to expand...

I bit the bullet, and bought the Optrand sensor. VERY impressed, works great.
This is PCP (in bar) at WOT on the TT, with matching VCDS data below.

230 bar scary ..

What is good is the really low dP gradient, which makes quite silent combustion.

The data looks great! A few questions.

1) What sort of boost and IQ would this gave been at?
2) What hardware did you use to log the PCP signal and Crankshaft sensor signal?
3) What software did you use?
4) What are the units for the crankshaft data (0 to +300)?

Charlie - which Y axis is cylinder pressure? The center one or the right hand one?

I've been trying to wrap my head around this curve. Two things stand out that didn't look like I thought they would:

1) How much pressure there is on the compression stroke. I did some quick Pv=nRT math with a 3 Bar absolute starting point and got numbers similar to what's above. It makes sense when you do the math, but I just had not thought about it before - My assumption was that most of the cylinder pressure came from the combustion part of the stroke, not the compression part.

2) The pressure rise vs crank angle during combustion actually slows down compared to the pressure rise during compression (Rub87 noted this as well). That seems backwards to me?, I would think that the pressure vs crank angle would increase during combustion?

Charlie - could you fold the plot over so that TDC is on the far left and the pressure traces are on top of each other? I think this will make it easier to see what's going on as well as be able to see the net work on that cylinder (which is what we're really after anyway).

By being able to calculate the area of the shape above the compression stroke and below the power stroke, you can start getting a feel for what is making more power. Since you know the engine geometry and crank angle, it's relatively easy to perform a net work calculation with force (cylinder pressure) over a distance (piston stroke).

Demon Diesel said:

The data looks great! A few questions.

1) What sort of boost and IQ would this gave been at?
2) What hardware did you use to log the PCP signal and Crankshaft sensor signal?
3) What software did you use?
4) What are the units for the crankshaft data (0 to +300)?

Click to expand...

1) This was at full boost, approx 2.7 bar, and approx 100mg/str
2) I collected the data with a simple USB Oscilloscope, with DSO capability.
3) Then manipulated in excel by hand, time consuming and can only work on crank cycle.
4) Right hand Y-axis is just raw DSO data for the crank sensor signal

Fix_Until_Broke said:

Charlie - which Y axis is cylinder pressure? The center one or the right hand one?

Click to expand...

The center one, so 230 bar PCP as Ruben pointed out. It has run for 2000+ miles like this and seems to be holding up fine. Custom ARP head bolts, I-beam rostens, 1/2" girdle and matched billet mains, so should handle it ok.
However I suspect the rod bearings are going to look ugly quite quickly.

FUB,

Thanks for the thoughts. I am running ~2.7 bar boost, so start with 3.7 bara and you quickly get to these levels of PCP before you add fuel.
From the no-fuel data I collected, I am seeing ~60 bar PCP with 1 bara intake pressure, so dynamic CR can be back figured from that.

I will send you the data, so feel free to plot it any way you want. I have been trying to learn about calculating MBT from the PCP data, but am struggling. I am beginning to think a purchase of a TFX logger and software for $4k might be an easier answer ! Their software is pretty impressive and calculates 50% mass fraction burn, cylinder torque, etc on the fly, per cycle.

Fix_Until_Broke said:

Charlie - could you fold the plot over so that TDC is on the far left and the pressure traces are on top of each other? I think this will make it easier to see what's going on as well as be able to see the net work on that cylinder (which is what we're really after anyway).

Click to expand...

Thanks Charlie - If you were to get an analog (absolute) output for crank angle, you could plot the above curve on your scope and watch it as you drive. More for visual observation than taking actual data and calculating things though you could still record it and calculate the same things.

Maybe something like this on the cam? (I had found one that was a triangle wave instead of sawtooth, but can't find it now) That's kind of important

http://cherryswitches.com/us/wp-con...14/10/Angle-Position-Sensor-AN8-Datasheet.pdf

Once you have the crank angle and pressure, the rest is just math. An analog input for crank position would make it easier to drop the data into an excel template that automatically calculates everything you want. It's an intermediate step - not as good as the TFX or similar systems, but also not $4k either

Thx. I have it plotting live on the scope now, using a trigger off the pressure input. Works nicely and once you learn where TDC is relative to crank pulses it is pretty easy to quickly gauge where PCP is. Each crank pulse is 6 Deg, so easy to scale off.
I am going to install video capture software to capture the laptop screen so I can reference back to VCDS data which would also display live.

The hard part at the moment is trying to calculate useable info from the cylinder pressure. I have been trawling through SAE papers trying to figure it out and my head now hurts.

Charlie - Go on Amazon and buy this book. Best $20 you'll spend on understanding this. Chapters 2, 10 & 15 are most relevant.

I started to try and type it out but quickly realized that I barely understand after 15 years it and will probably do more harm than good trying to explain it.

Internal Combustion Engine Fundamentals - Heywood

OK, I think this helps some.
I did the math to figure what cylinder pressure would be with no fuel, with 3.5bara boost. There are a few assumptions, but it is close enough for now.
Now you can clearly see the additional pressure from fuel burn.

This didn't require cylinder pressure measurement (just torque and displacement), but is related and interesting to see where TDI's fall in the scheme of things.

CharlieT,

What is "raw"? Could this be volts, mv, or is it some A/D bit counts? Just trying to judge how many volts I'm to expect from the factory crank sensor? When I measured it seems to stat within 10-20V peak to peak.

As for your last graph posted in #109. I think there is some noticeable error in the no fuel plot. Everything I've read points to the fact that combustion starts before TDC so "work" starts happening before TDC. Anything before TDC is negative work that pushes the piston down as it's coming up so this needs to be subtracted from the positive work which comes after TDC. Your no fuel pressure plot shows no negative work before TDC. Am I missing something?

The crank sensor signal is inductive, so its amplitude varies with rpm.
It is about 8v P2P at idle if I recall, and increases a little at higher rpms.
On the scope plots, 32=1v, so about 10V P2P

My SOI is only 14 BTDC, so I doubt there is much burn starting before TDC. That is why I am collecting the data, to see how much more advance I dare run. From what I see we are going to advance some more, but watch PCP closely.
My injectors are pretty huge, we seem to be able to get ~100mg in 22 Deg of duration, so I don't seem to need a whole lot of advance like smaller injectors would need.

Fantastic work!

Still need some others to digest to see what issues there are and could be with data interpretation.

For reference I have pd with cam with advanced injector lobe and std pd injectors fitted at moment with Firad 120% nozzles and I run about 17 degrees soi at 2500 rpm with duration of about 19-22 degrees depending on boost as I do prefer boost to come on a little more slow at 2500rpm, targeting 2.2 bar by 2800rpm.

I have max SOI as 31 degrees at 5500 rpm and duration around 29 degrees, boost 2.4bar. (Max torque around 3200-3400rpm.)
This low rpm (Under 3500) is where I am most nervous about my soi, I may knock it back by a couple of degrees.

With your BigBore injectors, you probably have a little faster burn than the ones currently in mine, might not be noticable.

From graph on previous page, I interpret visible combustion effect starts to be seen at around -6 degrees BTDC, and looks to end around +17 ATDC.

I have always had to calculate soi using assumptions based on achieving a 50% burn by 8 degrees ATDC, what I was really after was achieving burn by about 18 degrees ATDC which equates to piston having descended about 5%, and looks like you are getting that! (All empirical data from reading around and may not be correct, but rules of thumb often prove practically accurate, and there was nothing really else to go on.)

Would be interesting to see another plot for differing start of injection same speed, but remember that will alter total injected by a little too as injector rate of displacement varies with angle of cam.
Is your cam lobe modified?

Would also be interesting to see one at higher rpm if system is capable.

Chapelhill,

Thx for the kind words, it has been a long journey.

Appears we have the same nozzles, but with different methods of customizing injectors to maintaining pressure. My BBEs seem to do an amazing job of maintaining good atomization with high flows. I am getting theoretical 100mg at 24 Deg duration, which is approx double what I can achieve with stock injectors.
My cam is a MarkoP stg2 with 10mm lift, but the injector lobes are stock. Knowing that I only need stock SOI/EOI ranges as I have much larger plungers, I felt it unnecessary to tweak it.

Here is some more data. These were taken in 5 separate runs, but DSO'd at increasing rpms as I can only grab one dataset per run. So the SOI and durations jumps around a little bit. All same tune, but over the course of 1 hour or so.

LAST picture shows WOT@ 4232 rpm with soi at 20 degrees BTDC, but other pictures have increasing soi with rpm, but this is less than shown SOI at 3683rpm (SOI 22 BTDC) and I thought tune was the same. PCP max centreline is also showing just before 8 ATDC in WOT@ 4232 rpm which I would have expected to see later PCP max because of later soi and higher rpm? Would have expected from other graphs to see around 25 degrees. If I remember correctly standard pd150 maps are around 26 degrees at this sort of speed.

I think for 4232rpm my tune has about 26.3 SOI.

I run my PD tune with 1 duration map, and 1 soi map at temp above 60 deg C, otherwise too many maps to change and coordinate change at crossover between maps, only loss is trip computer accuracy which is low priority, benefit is you can have confidence in which numbers are actually being used.

I did have problem with head gasket leaking but as engine had done 170k miles, I changed head gasket, reduced compression a little and put on water injection, and seems to have held, but suspect it is marginal.

Alastair,

Thx, I think that might be an error on my part, I will double check. I have a horribly manual process of doing a run, logging group4 in VCDS and clicking RUN on the laptop scope right at the rpm I want.
Then pulling it all into excel, and calculating the rpm that the scope grabbed (from the crank signal), and then graphing group4 and reading the SOI and Duration at that rpm. It sucks, but it works.

I am going to try video capturing the screen at 30fps, and see if that gives me enough PCP info just visually, I suspect it might.

Tune is definitely the same, however I do have a full set of maps. I will speak to Mike (TDTuning) and suggest the single map, I like that idea. We already discovered there is a 10th 'safe' map that it seems to revert to from time to time for no apparent reason.

I also have to play with my low IQ settings. I got idle figured out with the help of JTT and is really nice now (which is the real problem people had with these BBEs) but now I have a flat spot in IQ-to-duration between 1000 and 1500rpm under load which is a real pain.
ie 5mg is calling for almost the same duration as 10 and 15mg, so throttle pedal response is non-existent until I get to ~25mg.
More to play with.....

CharlieT said:

... We already discovered there is a 10th 'safe' map that it seems to revert to from time to time for no apparent reason.
.....

Click to expand...

Are you running and EDC16 ECU by any chance? I've dug into mine quite heavily and figured out the majority of the SOI maps and how they are used/called on. Not all ECD16s are the same and even if the structure is that still doesn't mean they work the same but I think I have enough info to figure out the "safe" map.

So, the other fun thing I was able to do on the dyno yesterday, was hook up a cylinder pressure sensor. This is the same one that CharlieT was using above.

Data sampled at 50kHz - I'm working on plotting it vs crank angle but not quite there yet.

First image of the data is really just an overview of the whole dyno run from 1000 to 4000 RPM. Since the RPM varies quite a bit with each combustion cycle (see 2nd picture) when you plot against RPM, you get "noisy" graphs, but it's still useful to show peak cylinder pressure vs RPM in relation to power and torque I think.

Looks like peak cylinder pressure is ~235bar at ~3000 RPM and a secondary peak at ~1750 RPM under 200bar.



Here's 1750 RPM zoomed in (now plotted vs time instead of RPM)
X axis is time in seconds
Y axis is kW, Nm, Bar and RPM/10

You can see the RPM variation with each compression/power cycle in phase with the cylinder pressure sensor which is pretty cool.



Zoomed in to one combustion cycle. I'm guessing that combustion starts at ~85 Bar where the sharp rise in pressure is?



Here's another zoomed in one at ~3000 RPM - Scale is different on the Cylinder Pressure (right/red Y axis). Combustion starting at ~120 Bar?



Cool pictures at this point really - can't do too much with it without knowing actual crankshaft position, injection timing, etc, but it's a start

More to come as I get bits of time to work on it (and if anyone has a cable from a crank sensor).

Bump for good info!

Anyone already found out correct linearisation of the VW PSG002? Main concern: is it powered with 3.3v or 5v? or doesnt it really matter as its ratiometric?

http://beru.federalmogul.com/sites/default/files/psg_broschuere_gb_2014.pdf here it seems M9x1 is also available. to mount them in VP head think one could just drill exitsting hole in head so the body below the threads fit. this would only require 8.5 or 9mm, should be fine meat wise. and then one could weld a rod of alu to the outside of the head. offcourse do all this with the glow plug bore dialed in on a drill press.

then after welding just drill the alu rod and tap for M9. If you dialed in the hole well this will work 100% and you can sue the GM M9x1 plug