Injector Flow Testing Voltage - How Important?

[ErnieJones]

Hi Folks!

I'm flow testing injectors on the vehicle and was getting varied results with a graduated cylinder. I then determined that the voltage of the power going into the vehicle affects the flow rate of the injector test. At that point I connected a regulated 12v power supply. I now have a steady 12.4 volts throughout testing and consistency is spot on and repeatable.

I'm using MS Extra 2 on a Microsquirt Module. How accurate are my injector flow rates given that the vehicle runs at a higher voltage then the 12.4 put out by the power supply? I'm guessing that in a perfect world the injector flow rate would be measured at the operating voltage of the vehicle. I can't do that (at the moment). So the actual flow rate would be higher when in operation assuming the vehicle puts out more then 12.4 volts - which is does.

Am I screwing myself with this testing or does it not really matter? Is repeatability more important? Is there any way to enter into MsExtra that the flow rate is calculated @12.4 volts? Am I worrying about nothing?

[DaveEFI]

The standard voltage for most tests auto related is 13.8v. That is what things settle down to on a running engine.

[grom_e30]

that is what the 'battery voltage correction ms/v) setting is for. ideally you would test the injectors at 13.2v for the dead time.

[ErnieJones]

Thanks guys - that gives me enough info to research the issue I did find a company in the US that will clean, flow test and report the injector dead-time - that might be the way to go to nail down the injector side altogether. I suspect I should replace the battery first though as it's 8 years old and on it's last legs

[grom_e30]

the problem you may have is the dead time will be different between different injector drivers you could get them clean them ect, then make your own dead time measurements it will be more accurate then.

8 years for a bike battery thats pretty impressive lol

[ErnieJones]

Arghhh.....sounds like more work!

I was so hoping for just paying a 3rd party $20 and get back a nice clean injector with a datasheet but I agree that it would be much more accurate using my bike. Now I have to figure out how to do that

This whole thing came about because one day I saw my injector duty cycle was at 100% using the stock injector. I then checked the flow rate against a HP chart I found on the 'net and determined it was only good to 19 HP (!!) at 80% duty-cycle. So I picked up an injector made for the 150cc variant of my bike, which was better, but that one is only good to 22 HP @ 80%...

In the meantime, I installed a nitrous kit the other day.....which got me back to thinking about ensuring it would be rich enough when on the bottle....

That got me doing a ton of research to try and find a larger injector, which was really hard to do as the only bike that commonly used my injector is a....Hayabusa (!!) and they are rated at 650cc - from one end of the fire to the other....with seemingly nothing in between....

But I did find an Austrian bike (KTM) that uses the same style injector in their 250cc bikes and also the same one in one of their 350cc bikes. As luck would have it, I found a used one in the UK and the seller kindly mailed it over for the princely sum of 3P for shipping and 30P for the injector A magazine did a review on that specific KTM bike that included a dyno test an they said it made 36 HP....so my hope was that it would be in the ballpark and that KTM had chosen the correct injector for that model....

When I flow tested that one I found it really was good to 36 HP @ 80% @ 12.4 volts so now I have some overhead as this is roughly the area I want to be on the HP front with the nitrous (assuming I can incrementally boost the bike to that level without running over the crankshaft in the process) or with the turbo (I still really want to do that....)

So...it looks like my injector woes may finally be resolved and I now have some overhead and this means it should not lean out under boost (nitrous) if I do my job right. I managed to get it running last night (no small chore as that injector is BIG) and I got the idle dialed in nicely. I'm about to go out for a VEAL run and see how the VE table shapes up.

Interesting thing though....during all the idle testing and startup stuff I unplug the headlight to keep the battery from having to power it (makes a big difference in how much I can crank the thing during testing/starting). When I had the hot idle/VE dialed in I was about to call it a night when I plugged the headlight back in...instantly the idle changed and I had to go back in and change the VE a little to clean it up. (this always amazes me - how the smallest thing can alter the tune and how easily it is to adjust for it). After I did that, I looked at the battery voltage - it was between 12.4 and 12.8 - and 12.4 is what I was testing the injector flow rate at

I think the reason the stock battery lasted 8 years is because the bike has no parasitic draw and it's a Honda. So the chances of it being shut down with a full battery are high and with no parasitic draw....but it is now shot - all the starting/cranking/flooding/testing has taken it's toll. I knew it would but I do wonder if it's 'last leg' status is affecting my tuning. I suspect it is and probably on levels I don't understand (yet).

What I want to do is get a EarthX battery and be done with it, I can see a lot of tuning advantages to a battery that more/less locks it's voltage level all the way through 'til it's dead but they aren't cheap. I'm in the process of getting my muster up to just pull the trigger on it...

As always, your help is appreciated!

[LAV1000]

If you get the battery voltage correction right.
Than there's no need to buy an expensive exotic battery.

If you are flowtesting injectors it is easy to run those injectors on a lower voltage.
You need 1 or 2 diodes to get voltage drop (0.7Volt for each diode) and run the fuelpump on a separate (car) battery.
Put those diode(s) in the feeding line of the ms unit and injector this way you can check the voltage in Tuner Studio.

[ErnieJones]

Hi Lav;

The battery is really on it's last legs - the other night it wouldn't even start the bike properly with a charger connected I had been planning to get the lithium battery for some time and so I just did it - it hurt But I'm quite certain it's a lot better then the Shorai so....anyway, it's done now and will show up in probably a week or so.

I haven't had a lot of time to research 'battery voltage correction ms/v' yet - it's been busy. What I thought I would do is go this route...

Install new battery, let it stabilize (go for a long ride) and then see what voltage the bike idles at with the regular electrical load. Send out the injector and have them clean & flow test it and also do a dead-time test on it at the voltage that the bike normally idles at.

My thinking is this would be the most accurate way other then what Grom was talking about - the 3rd party testing outfit may have a different injector driver (?) At this point I'm not even sure how to get the battery voltage correction set, but I haven't looked into it yet. I did check the on-line help in Tuner Studio but it was pretty useless - all it said, effectively, was not to change it as it's usually right with the default settings

[SvenCS]

Send out the injector and have them clean & flow test it and also do a dead-time test on it at the voltage that the bike normally idles at.

Hi,

Do you have the contact info for this service. Thanks.

[ErnieJones]

Do you have the contact info for this service. Thanks.

Sure, it's www.fuelinjectorclinic.com

[LAV1000]

My thinking is this would be the most accurate way other then what Grom was talking about - the 3rd party testing outfit may have a different injector driver (?) At this point I'm not even sure how to get the battery voltage correction set, but I haven't looked into it yet. I did check the on-line help in Tuner Studio but it was pretty useless - all it said, effectively, was not to change it as it's usually right with the default settings

Grom has a point.
The injector is switch on and off with electronics and all elektronics have a little delay in switching on and of.
And than there's the way how the elektronics are driven could also be a timing issue.
That's why its better to measure the deadtime whit the system wich drives the injector.

And an other point, be awere of the altenator output, if it is up to system needs.

[ErnieJones]

I get that it may not be perfectly 100% accurate but it would be a lot closer then just guessing - which is what I'm doing now. As I recall, every time I looked into measuring injector dead time it looked pretty complex Certainly I'm not sure I could get it any closer

[LAV1000]

Yep youre right, should be close enough to get it running.

But thats the fun whit a DIY system put effort and time to it and youre spot on.

[sedd]

I hope I am giving good advise. I don't feel I am expert in this. I have done a fair amount of testing on the vehicle. I might have a few suggestions for others to comment on. It is difficult to sum this up quickly.

You need the flow at various voltages and injection rates at those voltages to create a curve. That curve sets the dead time. There are several posts on this with spreadsheets to help out. Megasquirt uses this voltage curve and the measured voltage to correct for the amount of milli seconds of injection.

I found that injector flow testing was difficult as the voltage variation, the voltage source, also the temperature of the gas seemed to make a difference in the readings. I also found that the injection frequency set by the tuner studio drop down affects the flow rate. I could not get consistent readings from one test to another.

I used a battery with variable voltage supply hooked up. It appears you need to use a battery so that there is sufficient amp flow. I ran my gas pump from another vehicle.

I used higher times between injections = low frequency, I used numbers of 100 and higher. I also used quite a number of injections to ensure a good average.

I found it was important to measure all injectors at the same time so that variations in temps and voltage supply were equal. I only had a rig to test one at a time and that was not optimal.

I feel that you will need to figure out the difference between injectors and not just get a dead time for one of them. I found that there was a variation in the slope of the curves, the dead times, and the flow. Once you know the % difference between the injectors, and the flow of the base injector, the tuning system should figure out the rest by setting the VE correctly. I used VE trim. I have assumed equal flow of air to each cylinder which is unlikely, so I have pulled the plugs to read them and see if they are all close.

[ErnieJones]

Don't think twice about it - I appreciate the help In my case it might be easier as it's a single-cylinder engine so only one injector

I have to look into the issue some more at some point because I'm not ever sure where the place is to put in the curve in Tuner Studio. I'm running MSExtra 2 and all I'm seeing are these two options;

InjectorDeadTime.gif

calibratebatteryvoltage.gif

You raised a bunch of good points though

[Yves]

I hope I am giving good advise. I don't feel I am expert in this. I have done a fair amount of testing on the vehicle. I might have a few suggestions for others to comment on. It is difficult to sum this up quickly.

You need the flow at various voltages and injection rates at those voltages to create a curve. That curve sets the dead time. There are several posts on this with spreadsheets to help out. Megasquirt uses this voltage curve and the measured voltage to correct for the amount of milli seconds of injection.

I found that injector flow testing was difficult as the voltage variation, the voltage source, also the temperature of the gas seemed to make a difference in the readings. I also found that the injection frequency set by the tuner studio drop down affects the flow rate. I could not get consistent readings from one test to another.

I used a battery with variable voltage supply hooked up. It appears you need to use a battery so that there is sufficient amp flow. I ran my gas pump from another vehicle.

I used higher times between injections = low frequency, I used numbers of 100 and higher. I also used quite a number of injections to ensure a good average.

I found it was important to measure all injectors at the same time so that variations in temps and voltage supply were equal. I only had a rig to test one at a time and that was not optimal.

I feel that you will need to figure out the difference between injectors and not just get a dead time for one of them. I found that there was a variation in the slope of the curves, the dead times, and the flow. Once you know the % difference between the injectors, and the flow of the base injector, the tuning system should figure out the rest by setting the VE correctly. I used VE trim. I have assumed equal flow of air to each cylinder which is unlikely, so I have pulled the plugs to read them and see if they are all close.

I had to test them one by one as well.

I found that the difference from one test to the other with the same injector gave comparable results.
The injectors seem to vary to some degree. As I do not use sequential, I will not be able to account for that.

Not sure if a balanced set of injectors would be better.

[sedd]

ErnieJones: I have MS3. I can guess a bit on your input values. I assume you have to test at various voltages and at the same ms value. Plot those on a graph and you will get the slope of that curve for the ms/volt value. I found that the curve was not liner. If you might run at lower voltages you might want to check out MS3 as it allows a better curve fit. If you run around 13.2 volts, plus or minus 1 volt, the curve fit will be pretty linear and you will be close enough.

The dead time is calculated at 13.2 volts and based on measuring flow at various ms injection values. Similar thing, plot them out and you will get a curve. there is information on MS site and quite a few posts to help you. I will convert a file and attach it later for results of what I have done.

[sedd]

this spreadsheet is what I used. there are a number of cells you may not need to use, for instance the temperature adjustments. I tried to do all my testing in the winter months and I think temperature of the gas and injectors makes results vary. I used voltage and small pulse width at the cold temp testing. I retested later in August all at the same temps and at 13.2 volts to reduce that variation potential. I used that info to compare injectors and to get the deadtime.

I measured my flow by grams as I found the volume measurement to be much less precise. I had access to a lab scale so I went that way. The spreadsheet allows for using volume as well. Check that your running above the non linear portion otherwise you need to do much more testing to get info on that. see tab for that. mine was around 2 ms. I likely will never run at that level.

[ErnieJones]

Sedd, thanks for doing the legwork for me - much appreciated!

As always, I'm fighting a war on several fronts. My exhaust valve is getting mashed so a top-end rebuild is in order and I also swapped out the OE injector (120cc flow rate) for the smallest size up I could find which turned out to be 240cc (!!) My new battery showed up just the other day so that's been installed and it's been a real treat to finally see a stable voltage!

I really appreciate you putting that post together for me as I'll be coming back to injector dead time in short order - this time armed with all the info I need.

[LAV1000]

Want to ad something to this.
I'am also running EFI on motorcycle.
Bike would not idle well on the injectors when using the spec data for these injectors.
Engine did quit at an AFR 13
After flowtesting and getting deadtime spot on it quits at an AFR 17.

Like Ernie said use a kichen scale to measure in grams in stead of cc , don't tell the misses.
Make sure measurements are 100 grams or more to rule out scale inaccuracy.

And little injectors try weber/magneti marelli iwp injectors.