Due to occasional misfire or unevenness at idle, I had reason to suspect the injectors were less than perfect. I wanted to test them rather than just a shotgun replacement approach, because if one or more were bad, I wanted to understand why. I didn't find any affordable test/clean devices that would work off the car, and I wanted to keep it cheap.

I bought a wall-mount plastic water filter casing - bottom container screws off the top, which has NPT ports for I/O. I got a pair of NPT/barb adapters for 3/8" rubber hose, stuck a short length of that on the inlet, with a valve carved off a bicycle tube crammed in. Outlet had a longer length of tubing, with another length of smaller diameter tube leading from theinside of the port to the bottom of the filter container. Into the other end of the outlet tubing I inserted the fuel distributor end of an injector line from the car. All secured with hose clamps. A bicycle pump with integral gauge used to build and monitor pressure. Total cost for hardware I had to acquire about $30. Only useful item I elected to forgo was any sort of filter.

For solvent I first used Coleman fuel (naptha), mostly to clean the lines, then test the idea. It worked, but no sign of any cleaning action. I considered 'parts cleaner', but the label warned of methylene chloride, a plastic solvent, and I decided against. I finally settled on a bottle of Techron injector cleaner. Procedure: pour solvent into container, screw on top. Connect injector to fitting, and let hang in glass jar. Pump, checking as pressure rises. Observe spray when check valve opens, and continue pumping as necessary until injector 'clean'. Disconnect pump and depressurize valve before jar overflows. Disconnect injector. I reused the solvent, checking for contamination as I went, testing in order 1 - 4.

Injectors 1 & 2 produced a lot of flow, and 'sang' loudly. I presume they are ok, but I don't know what to expect a proper spray pattern to look like. Injector #4 sang too, but less enthusiastically, and the pattern looked less even. The #3 was an issue. It seemed to have a restriction. It had way less flow than the others, did not sing at all, and the pattern was lousy. I even cocked the pintel open with a piece of toothpick and backflushed it, but it did not improve. Never in the whole process did it seem I was able to clean anything from the injectors, and I was even a bit worried I might be mucking them up inadvertently. However, I was satisfied that as a test, the system had worked.

I installed new seals (the old ones were hard - had to be cut off) and sealed everything back up. Car seems to run as before, so far, and I have ordered 4 new injectors.

Steve

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

if i remember correctly, there's a filter inside every injector that get's dirty over time, and cnnont be cleaned or changed. So better to replace the injectors

There is a filter, but I assume the debris can be cleaned or backflushed with the right solvent. Since these injectors are inexpensive and my setup not sophisticated enough for complex operations, this was indeed my choice. Nice to be able to verify a problem, though.

Our Accord uses injectors that cost each more than all 4 on the 190E, so I'd work more persuasively on those.

Steve

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

Primo hack!

How about a picture of your contraption?

__________________
Chuck Taylor
Falls Church VA
'66 200, '66 230SL, '96 SL500. Sold: '81 380SL, '86 300E, '72 250C, '95 C220, 3 '84 280SL's '90 420SEL, '72 280SE, '73 280C, '78 280SE, '70 280SL, '77 450SL, '85 380SL, '87 560SL, '85 380SL, '72 350SL, '96 S500 Coupe

Ok, here it is. Note the o-ring expanded into uselessness eventually - I will have to install a 3rd one next time I use it - if ever. Probably pick up a Viton one. Shown without the injector fuel line, since that is back on the car.

Steve

Attached Thumbnails

 

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

'Exploded' view:

Attached Thumbnails

 

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

how much pressure do you need and could you use a large syringe? say a 100ml plastic syringe hooked up to said hose? we used to give horses medicine through some whopping big syringes... but looks like your probably dealing with <50ml at a time through an injector.

cdt

__________________
2002suzukiGSF600S
1985 190e 2.3

The CIS injectors employ a check valve that opens at almost 50 psi. This allows the system to cut off fuel supply under various conditions. For testing, this requires the entire system be relatively high pressure. I chose the container I did due to its 125 psi use rating, and also needed to use hose clamps at all fittings.

One thing I verified is that even if flow was not ideal in all four injectors, the check valves worked prperly. None leaked.

Steve

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

Installed new seals when I tested the old injectors, installed the new injectors last night. Engine seems smoother overall. No more resonance on decel (might be the new seals) - less under throttle than before too. Idle still has the occasional misfire feel, which I suppose could be electrical. Maybe when it's dark this weekend in the mountains I'll look for an aurora ignitionalis.

Steve

__________________
'91 MB 190E 2.3
'08 RAV4 Ltd 3.5
'83 Lazy Daze m'home 5.7

I have been wanting to test my injectors and finally got around to it last night. The gizmo I rigged up was an air pressure regulator connected to a brass gate valve. On the exit side of the gate valve I plummed the can connector from an old R134A installation hose. The R134a can and the M103 injectors are the exact same thread size. The only readily available solvent I had was ether(starting fluid).
I filled the injector with ether and connected the R134A connector.
With the regulator valve I was able to increase the pressure until the injector "fired" which in most cases was about 45 psi.
Well, every one of the injectors has a one sided spray pattern. Some 'sang' and some didn't. Two of them leaked below 40 psi.
While my test was most certainly 'shade tree engineering', I feel that it identified a definite problem area. I'll order all new injectors. It certainly can't hurt anything except my wallet.

want to make one of these!! anybody here done this???

Interesting experiences .
The correct opening pressure of the cis injectors is 45 psi with no drips below that. A conical spray pattern is required for correct operation ,a half cone pattern is blockage (Bosch Injector Manual)
There was site which mentioned cleaning and installing new pintals for a reasonable price bit i seem to have lost the bookmark.
Ad