MSefi.com read only forum (Jan.15,2018)

Kind of expensive, but after all it's aero stuff. The information about sensor latency is interesting, though.

http://www.jpitech.com/1body_probes_and_sensors.php4
(Link updated)

Lance,

Every now and then I'll stumble upon a site like that and bookmark it (usually I post them here, too). Here's another one I found years ago, I think I posted it on the old yahoo list, but it has been long enough that it merits re-posting. I'd encourage people to post any similar finds here, even if you can't get the parts or they are too expensive, it's still interesting to see what's in the catalogs.

Make sure to check out the spark plug gasket pressure sensor on this link:

http://www.ngkntk.co.jp/english/product ... senso.html

Eric

I've actually seen one of these spark plug pressure sensors. The product it was attached to was an electrical system for small engines like chain saws.

I would have to guess that it measures capacitance changes to determine TDC then it sparks accordingly. So folks could probably solder a wire onto the existing spark plug gasket and get the same signal. I never saw it run though. Just saw the product. I wonder if it works.

My guess would be it's a piezo-electric device. In any event, it's an interesting piece of kit, it would nice nice to have some technical info.

Anyone know the price on one of these jewels? Probably makes the old Bosch FD5POR at only $18/each pale in comparison.

http://www.intertechnology.com/Kistler/ ... _6117B.pdf

Hello cylinder pressure guys,

some time ago I found really promising pressure sensors at www.optrand.com . I also read a paper (see below) which calls them 'low cost sensors'. The reference sensor was a Kistler type.

It must be a different world, a salesman I asked for prices sent this:
[...]
We can guarantee 200 million cycles in a head mounted package such as our
M5 or 6125 style sensor. Please see attached quote. Pricing on these sensors is (for 1-20 units): AutoPSI-S (standard) $785.00 ea, AutoPSI-TC (Temperature compensated) $945.00 ea.
[...]

Ok, let's continue ion sensing project....

@Eric:
AFAIK Bosch FD5POR is just for firing plugs, haven't known that they can measure presseure - or do I missunderstanding your posting?

Not to be too off topic:
http://www.bosch-motorsport.de/en/aktuatoren/index.shtm
http://www.bosch-motorsport.de/en/sensoren/index.shtm at least the temperature table is helpful for new *.inc files.

»Horst

PS @ paper:
Conception and test of a cylinder pressure based engine management for passenger car diesel engines
http://diglib.uni-magdeburg.de/Disserta ... eschke.htm You can download as pdf (3.4MB), abstract at this site is in English, thesis itself German only.

ami8break wrote:AFAIK Bosch FD5POR is just for firing plugs, haven't known that they can measure presseure - or do I missunderstanding your posting?

Horst,

I was making a lame joke about the expensive Porsche spark plugs being cheaper than the pressure sensors.

Nice links...

Eric

efahl wrote:Anyone know the price on one of these jewels? Probably makes the old Bosch FD5POR at only $18/each pale in comparison.

http://www.intertechnology.com/Kistler/ ... _6117B.pdf

i use those exact sensors for work to measeure cyclinder press.

they retail for 1500.00... thats what our order sheet says anyway...


Derrick

this might have some affordable application if ford get them out in quantity.

http://www.sensorland.com/AppPage013.html

A spark plug with a built in pressure transducer by Ford

First link is dead.

Tuners:
There is a way we all can get around "sluggish probes" and predict what the instantaneous actual EGT is (or CHT or CLN etc.)
The method is used on dynos but is a very simple formula if you're willing to do a little testing up front.
For every brand of sensor and for each setup (1" from head in headers, 3", exhaust manifold, whatever) the net time constant will be different. If an EGT sensor takes a few seconds to react, that isn't quick enough since the engine is already enduring the actual temperatures, which responded to the change in AFR or timing or boost etc. very quickly.
Open the attached excel spreadsheet and follow along:

We know that sensors react in an exponential manner (physical fact) and we know that the time constants for heating up and for cooling down are not the same and that the time constant for the system is too slow to safely use the readings directly.
So, we can derive the time constant for our unique engine setup (inline-4, opposing-6, rotary, V-8, V-12 etc.)

With the motor running at a safe steady state set the datalog (whether its MS or an oscilliscope or whatever) to record, and make an adjustment to your engine that would change the EGT by 100 degrees and practice (or program) to be able to repeat that change several times consistently. You will eventually want to make the change suddenlty. Changing the AFR can be done within a few milliseconds so that would work fine. Cranking the distributor by hand would take over 1/2 second so that would be too slow.

Go back to the initial condition and begin recording data then make the change. Wait many seconds (15-20) then go back to the initial condition.

Now to the spreadsheet:
I put in some numbers for an example but they will need to be replaced by the measured numbers.
At the top enter the initial and final EGT temperatures for the heating phase of the test.
Then fill in the time field (I used 2 seconds with .1 second sample rate as the example)
Then read the datalog and fill in the temperature at each time.
The "tau" column is calculating the time constant for your SYSTEM.
The "Th Theory" is showing what the temperature "should" be at any given instant, disreguarding the realities of turbulence or heat dissipation etc.
The box below the chart shows what the theoretical formula for Tc is.

The cooling portion is the same as the heating. NOTICE how tau for cooling is different than tau for heating!

The "Scalar(h)" and "Scalar(c)" fields were derived from the dyno industry and there is a link to a dyno-tech forum where I found this formula. The link is there in the spreadsheet.
Basically what's going on is that you can measure any two consecutive points on the curve, determine the slope of the curve (d/dt) and then maximize it as if tau was minimized.
Play around with the Scalars (currently set to 0.800) to get the steepest curve without really overshooting the final temperature. As shown in the dotted plots on the graph.

This formula is a simple one line that could be added to your MS code.
The scalars (h and c) would then be user entered and the sensor can be used to indicate the nearly instantaneous temperatures despite the sensor time lag. With the Quick reading, the EGT sensor can be better used to run the motor as it can react to a setting BEFORE the engine actually gets to the resulting temperatures. (remember that exhaust gasses heat up almost instantly, but the engine components may take a few seconds to react to the increased thermal loads.) You could have the computer react and cool off the exhaust before the engine (CHT) can get you over heated.

What do you think, is this helpful to you?

gfolsom wrote: The method is used on dynos but is a very simple formula if you're willing to do a little testing up front.

This spreadsheet is a keeper! This is what I like to see, the application of physics to the dynamic response of a EGT sensor. An excellent method.

To produce the hot-cold cycles, you could use the new step-fuel perturbation in the MSII code (used for determining O2 sensor and x-tau response), although you may get better EGT change with a timing advance shift. The method should work very well for sure.

Excellent stuff here, must save this one away....

- Bruce

I don't get the original link (http://www.jpitech.com/1body_probes_and_sensors.php4 ) to work. Could someone update this please?

Also intresting sensors and especially injectors:
http://motorsport.magnetimarelli.com/

Full catalog:
http://motorsport.magnetimarelli.com/Ne ... 08_v01.pdf

devastator wrote:I don't get the original link (http://www.jpitech.com/1body_probes_and_sensors.php4 ) to work. Could someone update this please?

https://www.jpinstruments.com/product-c ... s-sensors/

Well, it's been about a hundred years so I don't remember exactly what that page referenced, but I think the above link is the current equivalent.