Injector dead time and non-linear + video

[jsmcortina]

If you have a spare similar injector that you'd be prepared to post to me I would be interested in testing it too.

James

[Perkele]

As I promised I will share my test results on the Siemens Deka 630 injectors. With the P&H board
Only problem was that I didn't get the voltage to stay high enough when the puel pump is on.
So I first did a test with one small battery charger connected to the car. This way the voltage was 12.3V@battery and 11.9V-12.0V@injector when the test was running with fuel pump. And the dead time result was 0,70ms ( MS sees the same voltage which is at the injector +-0,1V)
Next I added another charger to the car but still the fuel pump draws so much current that the voltage tropped to 12.6V@battery and 12.3V@injector when the test started. Luckily it stayed on the same level during the test.
Now the dead time was 0,63ms.

My tests are in the attachet exel -file with calculated value for the 13.2V which was DT of 0,48ms.
So it doesn't differ too much from James results.

I will try again today with another car running next to mine with cables to hold the voltage higher.

EDIT: Made the tests again with another car running next to mine with cables.
Now the voltage was 14.1V@injector and the dead time was 0,4ms so at 13.2V it would be approxmately 0,51ms.
Updated exel included.

[swnp]

Nice video, James. Very informative. I may go out and do this myself as well to see what my dead times really are. I don't trust the dead times I got from the manufacturer too much. It's always best to verify on your own, right?!

And btw, was it wrong of me to listen closely and carefully during your video in anticipation of you saying, "rubbish"? Sorry, I don't hear an english accent too often outside of watching Top Gear or Formula 1.

[j-po]

If you have a spare similar injector that you'd be prepared to post to me I would be interested in testing it too.

James

What, me? Assuming you return the injector in about a month I just might do that to help solve my problem. End result will of course be you getting a perfectly linear rersponse curve and me looking like a incompetent fool. You'd have to pm me the address I should send it to. Mail in a padded letter shouldn't be too much within the EU.

[jsmcortina]

I measured that yellow RX8 injector and found a dead time of about 0.8ms at 12.2V.

James

[redrider706]

I measured that yellow RX8 injector and found a dead time of about 0.8ms at 12.2V.

James

James,

Do you have the pulse width vs cc/pulse plot for the RX8 injector?

I am interested in the flow characteristics in the non linear region.

Thanks.
-Raj

[jsmcortina]

I am interested in the flow characteristics in the non linear region.

I haven't measured the non-linear region, but spreadsheet attached.

James

[swnp]

I tried testing yesterday and I was unable to get any results because of my current setup.

My custom fuel rail doesn't have provisions for holding the injectors in securely on it's own. It relies on the injectors being installed, and rail being bolted into the intake manifold. So when I went to remove the rail I was reminded of this fact when three out of four injectors stayed in the intake manifold when the rail was in my hand. Being hard-headed, I proceeded to test, and kept popping injectors out of the rail at pressure! My son was tickled by it; I wasn't ammused.

When I get some thin wire, maybe a coil of inductor wire, I'll wire wrap them onto the rail and continue testing. That, or grab one the simular, but not exactly the same bolt pattern, SOHC intake manifolds and secure it that way for testing.

[jsmcortina]

Please be careful. Injectors MUST be secured. It is possible to eliminate the fuel rail and use a T piece and hoses etc. to provide the injector with pressure regulated fluid.

James

[richyvrlimited]

I am interested in the flow characteristics in the non linear region.

I haven't measured the non-linear region, but spreadsheet attached.

James

Thanks James, very useful.

How does that deadtime translate for mapping onto a MSII i.e. at it's 13.2v setpoint for deadtimes

[jsmcortina]

~10% less.

James

[j-po]

I measured that yellow RX8 injector and found a dead time of about 0.8ms at 12.2V.

James

Good to know, thanks. Helps me track down my problem. By the way, do the small pulse curves apply also to injector testing mode?

[jsmcortina]

In test mode the pulsewidth you specify is the exact pulsewidth applied on the CPU pin. No adjustments of any kind are made.

James

[ashford]

I tried testing yesterday and I was unable to get any results because of my current setup.

My custom fuel rail doesn't have provisions for holding the injectors in securely on it's own. It relies on the injectors being installed, and rail being bolted into the intake manifold. So when I went to remove the rail I was reminded of this fact when three out of four injectors stayed in the intake manifold when the rail was in my hand. Being hard-headed, I proceeded to test, and kept popping injectors out of the rail at pressure! My son was tickled by it; I wasn't ammused.

When I get some thin wire, maybe a coil of inductor wire, I'll wire wrap them onto the rail and continue testing. That, or grab one the simular, but not exactly the same bolt pattern, SOHC intake manifolds and secure it that way for testing.

a long time ago i tested my injectors by using 2 tie straps, one around the injector and one from the other tie strap to the rail.

[Greg G]

Thanks James! I was installing my RX8 injectors today and Richie pointed me to this post for the dead time!

[Joe Clark]

Has anybody tried a Pico injector yet? I think I have a 460cc IWP069 one that I broke the terminal off of, it shoulf be possible to solder a couple of wires on for testing purposes, if any one is interested in testng it I'll have a look for it and post it.

[Black99rt]

I was able to get linear results for dead time testing. I used 10ms interval, and 2,4,6,8ms pulses with enough events to nearly fill my graduated cylinder

When I went to do a flow test, finding the right settings proved difficult. I arrived at using 1ms interval, 20ms pulse, 20,000 events. I suspected this would result in going wide open for 20 seconds, but on the stop watch it seemed a bit longer. Is there a better way to do this?

[Black99rt]

My set up (not as nice as James' haha)

http://www.youtube.com/watch?v=p4xEE8d6j-g

[wrenchdad]

My set up (not as nice as James' haha)

http://www.youtube.com/watch?v=p4xEE8d6j-g

You should have seen my rig for my side feed injectors!

wd

[gruff511]

I did my test on the bench using 88 rx7 460cc injectors (195500-1350) as I could not find ANY official dead times on them anywhere online and I came up with around 1.25ms Dead time. What i dont get is are all my points i plotted supposed to be Linear to get accurate dead time?
If you look at my graph on Excel you will see the points slightly off the Trend Line. Do I need to go back and retest the points?

denso460injtest.xls