The most under utilized M800 install ever?

[Ben-S]

Hi guys,

Are you doing this to test the accuracy of the ECU with different injectors (inc dead time), check the flow rates of injectors or to set the ECU up to monitor fuel usage for circuit racing?

All of the above.... also it's a good starting place for individual cylinder compensation.

I initially used only dead time in my table in the ecu but after some experimenting common sense came into play and I realized that the closing time has an effect of making the flow vs pulse width non linear as well... Maybe it's just these huge injectors I'm using but I have found that doing test of the lowest linear pulse width, the highest used pulse width, and one half way in between results in a non linear flow vs pulse width graph so I am instead using a dead time number that makes those three tests as close to linear as possible. I have found this to be slightly different than actual dead time which I believe normally only takes into account opening time. This may not have a huge effect in practice but we will see.

[adam7]

Ive been doing a similar set of testing and also found the "deadtime" is not just function of supply voltage, eg my hi-impendance injectors have no flow for a drive pulse width <1ms, but once your into the 1.5ms + pulse widths the "deadtime" is stable (for a given supply voltage). Some interesting reading on the subject here:
http://www.yawpower.com

Re how to get a 0-5V for the weight ... I simply measure the weight using a cheap electronic kitchen scale (+-0.5g readout) and also note the volume using a graduated cylinder tube, so cant help with any tried and tested solution.

However, one idea might be to open up a digital kitchen scale and find the ADC component to tab into. I think they are based on a strain gauge in wheatstone bridge. Another idea is to use an old fashion scale (the balance type) with say a 100g counter weight, then run the injector test until the scale trips a micro switch or better still an optical switch...

Adam

[Ben-S]

or a potentiometer you could accurately graph flow vs time... best idea i've heard or found yet.

[Scott@FP]

You can get analog or pulse output industrial flow sensors, how about a McMillan 102?

http://www.mcmflow.com/displayproduct.asp?PRODUCT=102

[Pascal]

Mark, HA, i send u one M400? And you ship it NOW that M800. Hows that sound?

[ejenner]

I think my underutilisation of an M800 really goes a bit further!

I bought the M800 at the beginning of a long-term project in around 2004. For the next couple of years it sat on the shelf doing nothing.

Then I got a car for free:




I decided the best thing to do with a free car was to install a DIY nitrous injection kit. See the joke boot badging? Was this the only Saab 90,000 ever made?

So off I go down to the local air-fittings shop to get some industrial solenoid valves and some air fittings to hook it all up. I used an old co2 fire extinguisher as the bottle for the nitrous.

This is where the M800 comes in. I knew to get a good response from the system I would want to retard the ignition timing. I decided to use the M800 to control timing. I used a 2-slot crank mounted trigger wheel from an old Saab 9000 for engine position and then kept the distributor for sending the spark to the appropriate cylinder. I mounted a throttle pot for the ECU load info. The only other thing I did was to program the M800 to open the solenoid valves at a particular throttle position and RPM.

The guys at the rolling road thought this was hilarious and called it the biggest case of ‘overkill’ they’d ever seen. But just for the pure entertainment value they were happy to put it on the rolling road anyway:




There's a baseline run and a run with nitrous. You can see the nitrous cutting in! Also, check the air/fuel ratio. = gone in a cloud of black smoke! Look at the way the torque climbs, this is pure acceleration and even though the total power figure isn't that high it's obvious from the torque curve that this car really would leave most things standing when the boost started. This was a totally excellent project all things considered. I blew the engine once with nitrous and once mapping the static rpm ignition points. But new engines are cheap and easy to change so it was only a slight inconvenience in return for a lot of fun

[MarkMc]

Hi Aaron,
We do the test to work out the dead time of the injector over a range of voltages and fuel pressures. We need to manually enter this into the ECU so that our tuning is always compensating for dead time. This information also means our fuel usage calculation is more accurate.

[AnthonyK]

Just to follow up on Ross's quote ealier, the ADL or should I say ADLs on the MoTeC injector test bench logs the complete test also....

many hours and hours...

Cheers

[Holmz]

Two questions - and yes showing my ignorance:

1) Is there some detail on how to do this measure each injector's dead time as well as the flow rate?
Does one test at a few discrete injector open times and back out the dead time mathematically?

2) I am assuming that each injector can have a separate dead time and %enrichment set in the ECU to have them calibrated before actually running the engine?