DIY Injector Flow Bench

Forum for getting assistance with mechanical issues and custom fabrications/modifications

DIY Injector Flow Bench

Unread postby ironmanisanemic » Sat Oct 14, 2017 2:05 am

Hello,

For the last few years i have been frustrated with most injector companies only giving static flow instead of full injector calibration data. Now i do know that for most, its workable. But if your like me, it is going to drive you up the wall when your trying to dial in the MAF and your injectors. Sometimes it felt like i was pissing into the wind trying to get things right. Looking for more information i found that Injector Dynamics outlines a pretty easy to follow article on Fords injector characterization. http://injectordynamics.com/articles/ford-injector-characterization/ But alas, they dont appear to offer characterization services to the general public. Another good article i came across was from Greg Banish, https://calibratedsuccess.com/live-courses-2/fuel-injector-characterization/ but again no dice in giving any indication on where it could be done. Finally i found a place that outright will do it, http://www.deatschwerks.com/support/services/injector-services. But they are on the pricey side, costing $90 for the battery offset test, and another $90 for the "linearity" test. That was more than i spent on the injectors NEW! So i set off to design a low cost, user friendly, use what you got type of injector flow bench. Now, i know this is no easy task, but so far i have the bulk of the code written that will run on any Arduino uno or mini. It leverages the 16bit timer to control the frequency and the duty cycle, and uses an 8 bit timer to control how long to run the test. For now it uses a piece of software called MegunoLink Pro for the control interface. https://www.megunolink.com/ Unfortunetly, its not a free piece of software, but there is a trial version that i believe will work for just checking out what ive done. I plan on getting the ASE tech doc that discusses injector characterization in more detail, but i do know that most all testing is done at a simulated 6000RPM, and a pulse width from 0-10ms. This correlates to 0-100% duty cycle since 6000RPM equates to only a 100hz square wave. I plan to further expand upon the code to get away from the MegunoLink software, and move to a LCD screen with some form of menu and user interface, as well as add feedback for pump and fuel pressure control. Ive also started to rough out the hardware side of things and have made up a small PCB that currently only is controlling 1 injector, but can easily be scaled up, once the final circuit has been designed. I havent begun to draw up the enclosure, but in the grand scheme of things thats the easy part. I will attach a few pictures, screen shots as well as the source and dashboard for megunolink so others can give it a try. Im very open to suggestions, help with coding and circuitry as i am still very green in both aspects. If this becomes popular enough i could put this up on gethub, but for now posting it here will suffice.

EDIT

I forgot to add that i know there are a ton of DIY "Injector benches" on the internet. What i aim to do with this, is have a repeatable platform that can perform characterization of unknown injectors or decapped injectors. It can also serve as a cleaning station. What other solutions have failed to address is the repeatability factor. Most ive seen trolling youtube use a toggle switch and a stop watch to flow injectors.
You do not have the required permissions to view the files attached to this post.
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W

Re: DIY Injector Flow Bench

Unread postby ironmanisanemic » Sat Oct 14, 2017 2:06 am

I say simulate RPM only because it's easier to understand by most to say RPM than some arbitrary frequency. I'll show it in more detail in a video, but the frequency and pulse width are all calculated and displayed on the control interface so you still have all of your information. Right now I can select any value from 1 RPM up to say 20000rpm (haven't actually calculated out the upper limit, ive tested it to 18000rpm) 0.01% dc - 100% dc, and 0.01 Second on time to 70ish days(it's the theoretical max value in the timer that's being used for the time on). I would like to add the ability to sweep the RPM and simulate dynamic driving like what the commercial units do. But right now I need to focus on getting a reliably working schematic for the electronics, and get a few boards made up to test.
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W

Re: DIY Injector Flow Bench

Unread postby ironmanisanemic » Sat Oct 14, 2017 2:06 am

I was able to record a shortish video on how things work. Hopefully it's easy to follow. https://youtu.be/lcwjgRoaD-Q
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W

Re: DIY Injector Flow Bench

Unread postby ironmanisanemic » Sat Oct 14, 2017 2:06 am

here is the part number for the FET PSMN022-30PL i just had a bunch of these laying around. They are far from ideal with a relatively high RDS on resistance, but are still a good MOSFET for the price per unit. I attached a basic schematic of how i have it wired currently.

The pull down resistor is a 15k
the 2 diodes arent completely necessary, but are there for reverse voltage protection. They can be any diode as long as the one on the ground side of the MOSFET can handle higher currents.

The rest should be fairly straight forward.
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W

Re: DIY Injector Flow Bench

Unread postby decipha » Sat Oct 14, 2017 6:25 am

the problem with using an arduino or any other controller and driver for that matter is that your injector characteristics will change dependent upon the ecu used.

For example, you cannot put the same injector offset in a ford ecu that you can in a big stuff, holley, or GM ecu. Even though on paper it should be identical the controls algorithms (timers) and hardware delays will skew it quite a bit.

best bet IMO if you intend on using the injectors on fords is to rig up either a crank trigger on a wheel or a distributor stand and wire up the ecu. Let it actually fire the injectors and use the ecu for the testing.


btw, ive used lm1949 drivers with great success
User avatar
decipha
Tooner
 
Posts: 19118
Joined: Mon Jul 15, 2013 5:29 pm
Location: New Orleans, LA
Name: Michael Decipha Ponto
Vehicle Information: Supercoupin' x10
90 (4x 5spds) - Dante, Ruby, Daja, Ava
91 4r70w - Skarlett
92 (2x) 5spd & auto - Bianqa, Andrea
93 auto - Danika
94 5spd Rionda
95 auto Aisha
Vehicle 2 Information: Others:
00 Lincoln LS - Luanda
98 Camaro SS - Bounquisha
02 Harley F-150 - Sasasha
03 Marauder - DyShyKy
00 Explorer 5L - Bernyce
07 GMC 2500HD 6L - Veranafer
06 Chrysler 300 AWD

Re: DIY Injector Flow Bench

Unread postby ironmanisanemic » Sat Oct 14, 2017 8:38 am

I plan on implementing the same testing procedure as outlined in the sae document J1832. The only difference there should be is how each oem does their specific injector characterization. i would like to get this to the point where there are specific tests that can be run that will go through the process of testing parameters for each major oem. As far as using the arduino im not using any delay functions, it's all software interrupt leveraging the hardware timers. The timers will run no matter what the code is doing until one of the timer registers are changed. Right now my code is pretty basic, it will turn on timer 1 at whatever frequency and duty cycle you plan to test at, and timer 0 will shut it off after the determined amount of time. The rest of the microcontroller is free to do whatever it wants at that point. So it should be just as accurate if not more as using a trigger wheel and a ecu.
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W

Re: DIY Injector Flow Bench

Unread postby jsa » Sat Oct 14, 2017 4:09 pm

Great little project Ironman.

Further to decipha's post, your transister and a ford transistor could have entirely different switching times.

In your video you say the injector becomes unresponsive at 92%, is injector switching time involved?
Cheers
John
jsa
Power Poster
 
Posts: 479
Joined: Thu Jan 16, 2014 1:44 am
Location: In the shed or On the Computer, 'straya
Name: John
Vehicle Information: Escort RS Cosworth
EEC-IV GHAJ0
ANTI, COSY, QUIK or CARD
SMD-490, SMD-190

Re: DIY Injector Flow Bench

Unread postby FancieFink » Thu Mar 05, 2020 11:20 am

Hi...testing injectors at 100% duty is not correct, they should be tested at various duty cycles to get a picture of how they will flow under real conditions.
Spray pattern will undoubtedly change a bit at various pulse widths too. But there's definitely no way a mechanical relay would be able to mimmick real open/close times
If you could change that for a large transistor or solid state relay, then you could definitely re-create real world operation.

pcb quote
FancieFink
User
 
Posts: 1
Joined: Thu Mar 05, 2020 11:17 am
Name: Fancie Fink
Vehicle Information: FORD MUSTANG

Re: DIY Injector Flow Bench

Unread postby ironmanisanemic » Mon Mar 09, 2020 2:26 pm

FancieFink wrote:Hi...testing injectors at 100% duty is not correct, they should be tested at various duty cycles to get a picture of how they will flow under real conditions.
Spray pattern will undoubtedly change a bit at various pulse widths too. But there's definitely no way a mechanical relay would be able to mimmick real open/close times
If you could change that for a large transistor or solid state relay, then you could definitely re-create real world operation.


I dont know if you're trolling or are serious. It's apparent that you didn't even read my post. I never said anything about testing at 100% DC, or using a mechanical relay.
ironmanisanemic
A+ Contributor
 
Posts: 186
Joined: Sat Nov 09, 2013 10:50 am
Location: Vancouver WA
Name: Miles
Vehicle Information: 89 Ford Bronco - 393W, Professional Products 190cc heads, Custom Comp Cams hyd roller cam, Edelbrock EFI Mustang intake, 10:1 compression, FRPP LU34 34LB injectors, ProM 80mm Mass Air Meter, 2-1/2" into 3" y pipe, no cat, and hooker aerochamber muffler. U4P0 computer with Quarterhorse and BE, EA. 4R70W


Return to Mechanics & Fabrication

Who is online

Users browsing this forum: No registered users and 1 guest