Measure injector flow with ms2

masterx81 · Post by **masterx81** » Sat May 27, 2006 4:27 am

Hi to all...
I'm starting to msns a opel/vauxhall corsa gsi, but i've already buyied the ms2 chip (for the ms2-extra version that james is working on).
Befor install the ms-ecu in my car, i need to measure the flow of my injectors (i wasn't able to find any info!!!).
Is this possible with the ms2 chip? How accurate is?

Thanks!

Bernard Fife · Post by **Bernard Fife** » Sat May 27, 2006 5:39 am

masterx81,

In the release of the V2.6x code, Bruce said,

Bruce wrote:MS-II (has a) mode which can be used as an injector flow bench. We get several inquiries a week on the JectorRate FI driver setup, enough that we think that adding this mode will be useful. The plan is to set up a specific injector pulse width, duty cycle, and an adjustable commanded number of pulses. So, an injector an be plumbed up on a test stand with test fluid, and the injector exercised for a specific number of pulses/widths. And the injector open time and PWM current limit will still be there, so these can be characterized in the flow. The test fluid volume can be measured, alternatively the mass change introduced by the fluid can be measured on a sensitive scale. If people really like this mode, then we can also fire up one of the ADC inputs and allow one to add an accelerometer input. What you can do is take one of the low-G accelerometers available from Freescale, ST, or ADI, physically clamp it to the injector, and rapidly sample the analog voltage. The accelerometer will pick up the motion of the injector pintle, and the open and close positions (the point where the pintle slams onto a stop) will really show up on the time-sampled spectrum. So injector open and close times can be directly measured - pretty cool indeed!

and,

there is a fuel injector test mode which allows you to set up n squirts of any input PW and off time. It then sits there waiting for either a reset command from MT or change the test option and reset to resume normal mode. This could be used on the bench for commercially testing/cleaning injectors or to clean/test them in your car by pulling out the injectors from their pockets and letting them squirt into a can. The cost of a complete MS-II fully assembled is still way cheaper than a commercial tester/cleaner. (Bruce and Al saw one for $8000 and all it had besides a controller were 8 burettes.)

I will attach a rough INI file that exposes the test mode.

Lance.

Ludotech · Post by **Ludotech** » Sat May 27, 2006 7:23 am

Lance,

I'm reading this message and this is what I want!!

I was thinking about a way to measure the opening time of my injectors and i saw the jectorrate FI but now I know that the 2.6x code can do this!

I don't understand what you have written: So, I have to download the 2.6x code (which?) and then select the INI file you have attached in megaconfigurator and that is all? In this case, does this mode need the RPM or others informations to feed the injectors or is there a a specific menu in MT?

For the opening time, if I buy an accelerometer(without using the 2.6 code), all I have to do is to fix it to the injector and measure with a scope the voltage while the injector is opening?
Did you ever do this?

I know that there are a lot of questions, thanks for your reply

Note: where can I find the software and files for the jectorrate Fi?

Ludovic

Post by **grippo** » Sat May 27, 2006 7:50 am

There is no Jectorate harware unless you build it yourself. But you can get a good estimate of opening time from the MS2 injector test mode in two ways. The most accurate would be to build the accelerometer circuit, and display the accel output on a 2 channel scope triggered by the injector line from MS2. A less accurate way would be to set into MS2 a .1 ms Topen time and then increase the PW in .1 ms steps until you can see a fuel dribble. That tells you how long it takes for the current to overcome the inertia of the pintle and just start opening it. Then you can nmeasure flow rate with say a 50 or 100 ms pw and then fit the results to a model with 4 pieces: the just barely open time you measured in step1, the unknown time from there to fully open, the time from fully open to the command to stop, and then from there to the time at which all fuel flow ceases. The reason for using the 50-100 ms pulse is to get a very good estimate of the flow rate when fully open. With a very large pw you can ignore the open / close times. I would use an even larger pw - I just don't have a feel for how long you can hold one open without damaging it or the inj driver. With all of the measured data you can then do a series of measurements of flow rate vs pw using the 4 time model and do a least squares data fit for the 2 unknowns( tstart fuel to full fuel flow and tstop fuel to fuel stopped). If you would rather tinker with electronics use method 1. If you like math use method 2.

Ludotech · Post by **Ludotech** » Sat May 27, 2006 8:09 am

Thank for your reply Al.

If I build the jectorate by myself, i think that there is a software for this, Am I wrong??

The most accurate would be to build the accelerometer circuit, and display the accel output on a 2 channel scope triggered by the injector line from MS2

Whith this way, I'll know the opening time of my injector but during this opening time, the fuel is flowing until the injector is fully opened.

So, Is there a way to compute or measure the flow during the opening and open time?? i'm curious.

I just don't have a feel for how long you can hold one open without damaging it or the inj driver

A student of my school fed an injector (high impedance) during 30 seconds to measure his flowrate without any damage but I think that it's huge!!

Ludovic

Mike_Robert · Post by **Mike_Robert** » Sat May 27, 2006 10:09 am

Ludotech, you wrote "Whith this way, I'll know the opening time of my injector but during this opening time, the fuel is flowing until the injector is fully opened. " I'm told by reliable people that the injector doesn't gradually open. The coil flux apparently builds until it's able to overcome the force keeping the injector shut then the mechanical parts open very rapidly, I think they mentioned microseconds. This leads to the conclusion that during the opening time, no fuel is injected at all. It's probably more accurately called "delay until open time".

-Mike

Post by **grippo** » Sat May 27, 2006 7:45 pm

Mike,

I have heard the same thing, but I have also read papers where they model it as an average of 2/3 full flow during opening and 1/3 full flow when closing. I have derived this based on the equation for charging/ discharging a coil plus factor in the spring. Both can be correct statements if you consider 2/3 flow for a couple of us to be negligible - which I do. But still, the spring is what makes me think its not quite just a few microsecs. Would certainly be an interesting measurement.

teg · Post by **teg** » Mon May 29, 2006 12:32 am

Ludotech wrote: For the opening time, if I buy an accelerometer(without using the 2.6 code), all I have to do is to fix it to the injector and measure with a scope the voltage while the injector is opening?
Did you ever do this?

I have been measured opening times from a REALLY big injectors (from a diesel engines with several thousand kilowatts). In those injectors it is possible to see a change in current when it has been fully opened. I don't know if it possible in these small injectors - I have to try...

Injector current will suddenly sink and then continue to increase. So there is a small valley in current when it has been fully opened.

Mike_Robert · Post by **Mike_Robert** » Mon May 29, 2006 5:14 am

Al, that's quite interesting. I guess the real life best way would be to determine it empirically. ??

teg, yes, I can clearly see that behavior with the limited number of small, Japanese/Bosch style injectors I've bench experimented with. The current dip does appear to correlate well with real life experimentation on a running install.

-Mike

teg · Post by **teg** » Mon May 29, 2006 9:30 am

Mike_Robert wrote:The current dip does appear to correlate well with real life experimentation on a running install.

Are you sure? I'm pretty sure that many automotive (or ECU) manufacturers uses this current dip to detect valve (or injector) opening in realtime. Are you sure that this doesn't correlate real life? This current dip occurs when magnetic core stops moving...

Of course when using electro-hydraulic injectors, it doesn't correlate real life.