I want to build a tester to get an approximation of the flow rate of some injectors. I figured I would get a graduated cylinder, or make one out of pipe and mark it off so i can put a measured amount of fluid in, weld a top on the pipe and put a pressure fitting on the top, somthing I can un screw and fill the pipe with some sort of fluid, maybe gas, and then pressurize with either a hand pump, or air compressor. The pipe would have to be approximatly the same size as injector bung on the intake, so the injector can seal in the pipe. Then using an old injector plug make a circuit to open the injector for 1 sec, or whatever amount of time I need, and then measure the fluid that the injector outputs.

Does this sound reasonable, if so maybe somone can help me with some numbers, like:

- voltage to open the injector

- how long to open the injector ( I assume I could use whatever duration, calculate the flow/sec i.e. 20cc's in 1 sec, and convert that to lb/hour)

- what pressure injectors are normally tested at (44psi?)

I guess I could even do this with a postal scale attached to a computer, and have the computer do all the calculations, and heck even use the PC to control the duration of time th injecto is open.

So what do y'all think, possible to do; 1. cheaply, 2. am I missing anything

Thanks
P

 

No offense, but why?

5 gm/sec would be fairly high flow. (39.6 lb/hr)

 

You need to test the injectors under a variety of conditions. The flow rate is affected by the voltage applied because it alters the opening and closing speeds. That's a factor in developing the "offset" factors for system operating voltage. You would test the injector at the typical system voltage, say 13.0-13.5V and then vary the voltage in 1V or 1/2V steps to evaluate the affect of voltage changes on the flow rate. The ECU need to know that data so it can correct the calcualtions for the voltage.

The standard for pressure used for testing is not fixed. It is generally the result of the application. The LT1 setup is designed to operate at 3.0 bar (43.5psi). The stock RP injectors are flow rated at 43.5psi. Some late model Ford applications operate at 2.7 bar (39.15psi) so the FMS/SVO/etc injectors are flow rated at 2.7 bar.

Your scheme for pressurizing the pipe or container isn't going to work, at least not unless you have a method to hold the pressure in the vessel at a constant value. If you simply pressurize a closed container of fuel, and start to withdraw fuel via the injector, the pressure will fall. You need a regulated and controlled constant pressure.

You should also be pulsing the injector at something that simulates the real world. Running a constant stream won't give you a meaningful flow value. The injector has to be able to operate with pulse widths in the range of 1.0mSec to 20mSec. The electrical characteristics of the injector, and its "speed" will affect the flow rating when you try to operate the injector for only a few thousandths of a second. An injector "driver" is a reasonably sophisticated device.

It gets more complicated when you try and test a low impedance injector. While a high impedance injector operates at a constant voltage, the low impedance injector uses a "peak and hold" driver, which can apply a high initial current inrush to open the injector quickly, and then modulates current to a level that produces only the lower current required to hold the injector open. Setting up a 12V power supply with an on/off switch isn't going to replicate a real world injector driver.

When you start talking about a system with uncontrolled pressure and a postal scale, and pulse width in minutes, you aren't going to get meaningful results, IMHO.

 

B/C i'd like to be able to test injectors, i.e. make sure they're flowing similarly.

Injuneer; real world woul dbe nice, but I'm trying to make a setup where I can test an injector and compare it to another injector, i.e my current 84 lb/hr injectors, I'd like to compare them to each other, and then be able to compare other injectors.

Also the constant pressure thing isnt that big of a deal; air fitting to a compressor with a regulator at what ever setting I want.

I basically just need everything to be constant, except the injectors, that way I can compare injectors, correct?

P

 

Injuneer is right but, it can be done. The main thing is the pulse width and controlling it. But you are wanting to compare one to another which is how they would do a matched set for flow.

I used an Allen Bradely micro PLC to control the pulse width, you don't want to turn the injector on to long because its not designed for a constant on state. The PLC was a 120v relay output type so I used 12 volts on the relay contacts with no problem. The scan rate of the PLC is around 20 mille seconds (ms) so I turned on the relay for three scans and then off for three scans which gives me about 60 ms on and of. Rather than count time, I counted switching states but put the total time in a register for comparision.

For fluid, I used rubbing alcohol. Although the specific gravity is not the same as gas, I wasn't worried about this because I'm just compare overall flow rates. I used a piece of 1" pipe and reduced it down and used rubber hose to connect to the injector. I put a tube fitting on the top an pressurized it with the air from my compressor set at 40#s.

I sprayed into a buret for my readings. I had to adjust the counts to get a set fluid level, then compared around 20 different injectors and took the eight that were the closest together. Neat thing about the whole experiment was that I could also adjust the pulse with long enough to see the spray pattern OK. I was looking for something soluble in rubbing alcohol to color the fluid to get a better look. I could also see how well the injectors were closing. Some leaked a tad at the end, just the pressure from the system was all that was needed here.

 

The point I was trying to make is that opening an injector and letting it flow for a given period of time.... e.g. one second.... will not produce meaningful results. The opening and closing of the injector has to be factored into the equation. Without that, the results will be meaningless.

From what you described, you are basically measuring the (equivalent) orifice size of the injector. But that's only 1/2 of what determines the flow capacity of the injectors.

Help me understand what you are trying to accomplish... are you trying to find a "set" of injectors that is closely balanced? Or are you trying to somehow evaluate the flow capacity of an 84# Bocsh vs. an 84# Lucas? Or both? Or something else?

RC Engineering used to have a real good paper on injectors, including the issue of flow testing. Unfortunately that paper was posted on a Supra site, and appears to have been removed. And Russ doesn't have it on his own website. You might want to find a TPIS catalog.... they also have a section on how they do their injector testing and balancing.

 

http://www.cruzinperformance.com/injsteps.html
You might find that interesting. Would use alcohol or something instead of gas, don't want to see anybody get hurt.

 

Basically I want something I can compare multiple injectors, and match them. I have several sets of stock injectors, I want to be able to find the ones that flow similarly, and those that dont leak, and those that actually still work.

What I DON'T need to do is compare Lucas, to MSD, to whoever might be out there.

The other thing I'm wondering is how could I open and close low impedance injectors, if I'm correct I can just apply voltage to high impedance, but for low I have use "peak and hold" tech, so bascically unless I adapted one of LJs Versafuelers to power a single injector, or made my own peak hold device, I wouldnt be able to test my low impedance injectors?

More or less I want this to just be a tester, so I can test my own injectors, and make sure they are working, flow evenly, and dont leak.

Thanks for all the advice

P

 

Coil design is the main factor for the different types of injectors. If you apply a voltage and current flow to a wire, it will create a flux field around the wire. Now if you wind this same wire into a coil, then as the current moves down the wire it creates a flux field that is imposed on the wires next to it. This flux field acts as a resistence to the flow of current. Voltage lags current in an inductive field (coil) and current leads voltage in a capacitive field ELI the ICE man. The difference between the way the injector works is how it is wound, the higher the impeadance, the more restrictive it is to current and voltage. A low impeadance injector will allow the current to flow quicker, thus moving the internals faster. The voltage will rise quicker and stay there longer. High impeadance will resist the flow. Peak and hold is a fundemental difference in the way the coil reacts.
There are also factors that involve the pintle, after all it is the device that is being moved and it is being moved by the flux fields that are being created from the coil. The design of the pintle and the material can play a major factor. Installing dissimular metals in the pintle can make it move at different rates for applied voltages.
Then there is strange things that happen when the voltage is removed. The fields in the coil colapse and try to move the current back down the line. This can be helpful when designing the pintle.

 

http://www.bgsoflex.com/fitester.html

MegaSquirt creators started on this some time ago. Plans and software are available. You can find many graduated cylinders on ebay. Stoddard solvent/mineral spirits is usually around a sg of .76. This is what we flow all of our shipping fuel pumps with. A 4.6 ford fuel rail has perpendicular injectors and would make for a good test setup. There are diagnostic injector drivers out there that could also be used.