Originally Posted by mach4 (Post 3003978)

I've been working through a problem with a rough idle that has just recently become what feels like a gasser running with one or two spark plugs unplugged.

The first step was to pop test all the injectors and see if I could find a suspect injector. The results on my new DIY tester were as follows

1- 1850# good spray
2- 1800# good spray
3- 1800# good spray
4- 1850# good spray
5- 1800# generally good spray, but streams

I replaced #5 with a 1850# injector with a good spray from a junk yard set I had.

It still ran the same.

So I decided to do a compression check and this is where it got interesting. I picked up a Harbor Freight unit and then proceeded to pull out all the injectors. I had been reading up on doing compression tests and the jury was out on whether to use the heat shields or not, so I decided to use them. I screwed the test adapter into the #1 injector slot. It kind of had some resistance through a few pulls then snugged up nicely. I cranked the engine and it registered 410#. Not a bad sign.

But when I pulled out the test fitting, this is what I found

http://www.peachparts.com/shopforum/...1&d=1346632005

The injector test fitting had literally punched the center out of the heat shield. It didn't register what happened at the time, so I went on to #2, #3, #4 and #5. Then it hit, where is the center piece of the heat shield. Panic time. Amid visions of replacing the pre chamber, pulling the head and possibly damaging a valve or worse, I found the center piece had been blown out on the ground.

So here's what we've got. Notice how the bottom of the injector is different from the compression test fitting in the HF kit?

http://www.peachparts.com/shopforum/...1&d=1346631955

There are actually three problems. The first is with the configuration itself and getting a good seal. But there is another problem and that is the mismatch in the size of the bottom piece. The arrows show how much more volume there is under compression compared to with the injector in the cylinder. Some threads have speculated that the hole in the adapter will affect the compression ratio, but this is dramatically more than that.

http://www.peachparts.com/shopforum/...1&d=1346632005

And the last is the quality of the kit itself. The first one would not release the pressure (I took it back) and the second one would not hold pressure (I'm going to take that one back too)

Anyway, here's my results on a cold engine

1- 350#
2- 360#
3- 350#
4- 310#
5- 310#

Notice that the above results reflect redoing #1. The first test was 410# and the next one was #350. I can only conclude that the mismatch in the injector-tester configuration produced the difference. With the heat shield "jammed" in and eliminating the mismatch area, it read considerably higher.

I was going to cross reference to readings at the glow plugs but gave up when I realized that 2, 4 and 5 were inaccessible.

I believe it is possible to fabricate a good test adapter from an injector and some JB weld. The connection to the compression tester is a simple 1/8"NPT fitting. I'll post any success in creating a good test fitting.


After everything is back together, I've still got a rough idle when cold, but the missing and hot idle roughness have gone away.

Originally Posted by Diesel911 (Post 3004024)

My Harbor Freight Compression Test Kit has an adapter, a Tube/Ring and a couple of Copper Washers. I had always assumed some combination of them would work.

Originally Posted by mach4 (Post 3004028)

Good call. I just took a look and you're right. It appears that the adapter, then the copper washer, then the spacer and finally a heat shield would work perfectly. I'm not going to rip it apart to retest right now, but that's the answer.

I'm still dissatisfied with the overall quality of the kit. Maybe #3 is the charm.

Thanks

Originally Posted by mach4 (Post 3003978)

I've been working through a problem with a rough idle that has just recently become what feels like a gasser running with one or two spark plugs unplugged.

The first step was to pop test all the injectors and see if I could find a suspect injector. The results on my new DIY tester were as follows

1- 1850# good spray
2- 1800# good spray
3- 1800# good spray
4- 1850# good spray
5- 1800# generally good spray, but streams

I replaced #5 with a 1850# injector with a good spray from a junk yard set I had.

It still ran the same.

So I decided to do a compression check and this is where it got interesting. I picked up a Harbor Freight unit and then proceeded to pull out all the injectors. I had been reading up on doing compression tests and the jury was out on whether to use the heat shields or not, so I decided to use them. I screwed the test adapter into the #1 injector slot. It kind of had some resistance through a few pulls then snugged up nicely. I cranked the engine and it registered 410#. Not a bad sign.

But when I pulled out the test fitting, this is what I found

http://www.peachparts.com/shopforum/...1&d=1346632005

The injector test fitting had literally punched the center out of the heat shield. It didn't register what happened at the time, so I went on to #2, #3, #4 and #5. Then it hit, where is the center piece of the heat shield. Panic time. Amid visions of replacing the pre chamber, pulling the head and possibly damaging a valve or worse, I found the center piece had been blown out on the ground.

So here's what we've got. Notice how the bottom of the injector is different from the compression test fitting in the HF kit?

http://www.peachparts.com/shopforum/...1&d=1346631955

There are actually three problems. The first is with the configuration itself and getting a good seal. But there is another problem and that is the mismatch in the size of the bottom piece. The arrows show how much more volume there is under compression compared to with the injector in the cylinder. Some threads have speculated that the hole in the adapter will affect the compression ratio, but this is dramatically more than that.

http://www.peachparts.com/shopforum/...1&d=1346632005

And the last is the quality of the kit itself. The first one would not release the pressure (I took it back) and the second one would not hold pressure (I'm going to take that one back too)

Anyway, here's my results on a cold engine

1- 350#
2- 360#
3- 350#
4- 310#
5- 310#

Notice that the above results reflect redoing #1. The first test was 410# and the next one was #350. I can only conclude that the mismatch in the injector-tester configuration produced the difference. With the heat shield "jammed" in and eliminating the mismatch area, it read considerably higher.

I was going to cross reference to readings at the glow plugs but gave up when I realized that 2, 4 and 5 were inaccessible.

I believe it is possible to fabricate a good test adapter from an injector and some JB weld. The connection to the compression tester is a simple 1/8"NPT fitting. I'll post any success in creating a good test fitting.


After everything is back together, I've still got a rough idle when cold, but the missing and hot idle roughness have gone away.

Originally Posted by barry123400 (Post 3004272)

Quick test. Shut off the return line from the injection pump and see if the idle is better. Indicated compression readings are not so far off that the idle should be noticeably bad from that anyways. If you have not checked for valve lash do so. I am assuming this is an older 617 engine that has mechanical lifters.

Originally Posted by Diesel911 (Post 3004350)

He would be better off temporarily replacing the Cigar Hose with a section of 5/16" Fuel Hose; that can easily be pinched off with a Pliers or a Vice Grips.

If the idle gets better by pinching off the return Hose you have an issue with your Pressure Relief/Overflow Valve.

Originally Posted by Army (Post 3004729)

It seems as though most people ignore the glow plug hole as a connection for a compression tester / leak down tester - is that right?

Originally Posted by Army (Post 3004729)

It seems as though most people ignore the glow plug hole as a connection for a compression tester / leak down tester - is that right?