DES 8/1 Propane Conversion project

[dieselgman]

Glad to hear that you are making progress... fine tuning will probably take care of your remaining symptoms.

Here is the inlet restrictor as used with the HRG... 2 1/2" inlet reduced to 1 3/8"



dieselgman

[BruceM]

I worked the rest of the day on adjustments- solved the no load issue with the idle bleed screw.   Running the 2 stage compressor for a load, I then gradually increased ignition timing to increase power.  I have some "head room" via throttle now with the compressor pumping into a 150psi tank.  Before advancing the timing, it was maxed out at 150psi.

I'm not hearing any signs of pinging (precombustion) under load so far, so it seems I can probably increase the compression ratio back up to 14:1.

The governor spring was replaced with the famous Home Depot spring and it worked good enough for where I'm at right now, about 40 rpm from no load to "full" load (air compressor).

I was then going to try some loads on the ST-3 but was stymied by it not working.  First there was gunk under the brushes so no contact, then no joy despite flashing.  It has an AVR that was hooked up by CGG to bypass the harmonic winding, and I'll have to check that out tomorrow.  The bearings are moaning already, on a new unit.  Grrr.

It seems Gary's CA110 has saved the day!

[BruceM]

Got my timing checked via strobe, and I'm still way to retarded.  Seems there is more lag to the hall effect switch than I anticipated.

The dilema is that the IP cam lobe on this engine is so seriously retarded (17 degrees ATDC) that I can't advance the timing any more without fixing that problem. 

I'm going to pull the camshaft,  move the IP lobe, and have it welded.   Anyone know what the spec'd angle is? 

[Thob]

Got my timing checked via strobe, and I'm still way to retarded.  
...

Don't feel too bad.  Sometimes I'm way too retarded too.

Is it possible to move the magnet to the flywheel instead of where the IP was?  I think a lot of engines run this way - it gives you an extra spark on the exhaust stroke, which doesn't harm anything (unless the spark is way retarded and happens on the intake stroke).  You can vary the timing by moving the sensor around relative to the flywheel.

I was thinking a little about your injection pump cam and wondering "how could it be that far off?".  Then I realized that the IP doesn't inject at the peak of the lobe - it injects somewhere on the upstroke, before the peak.  It pulls in more fuel on the downstroke.  So maybe it's in the right spot?

Sounds like you're getting close!

[Tom]

I'm still way to retarded.

I've got that problem too. Glad to hear about the progress. Can't wait to hear a video of it running sweet under load.

[dieselgman]

Lister published a specification that says "fuel starts" at 18 to 20 degrees btdc. Actual lobe position may be considerably off this mark because fuel starts sometime during the ramp lifting the pump tappet.

dieselgman

[BruceM]

Gary, thanks, I did look up the specs.   I guessed at about 14 degrees BTDC for the cam, which puts IP firing 5 degrees before the lobe tip.  I'll find out tomorrow how it turns out.

Tom, yes, I run a bit retarded most days, myself. Noticeable lack of pep...  

Thob, yes, there's always the wasted spark method via magnet on the flywheel (at twice the power consumption), but I've grown attached to using the IP cam follower.

 I pulled the camshaft and moved the IP cam lobe to where I think should be for diesel operation.  I needed to pull the flywheel to fix an oil pump plumbing leak anyway, and I wanted to raise the compression ratio back closer to 14:1.  So it all came apart, again.  

I was going to use 38ac's set screw method for testing purposes, but I could not get even a cobalt bit to drill through the cam lobe collar; it was that hard!  By that time I was pretty tired so out came the JB weld.  I think that should be adequate without welding since I have no IP.

Tomorrow I'll get it all back together and see how I did on the cam lobe timing.  Then I'll find out if I must add circuitry for retarded starting timing when I get the running timing properly advanced.  

[BruceM]

I've been delayed by other parts of Jeff's new homestead project but wanted to update this thread.

In checking out my modified IP cam lobe placement, which came out well, I found a big oops.  The exhaust valve cam lobe has a big problem.  It was not really closed when I set my lash by 38ac's method, it was actually .003" open, on a flat part of the cam.  So it was cracked open during the intake stroke, and for about 20 degrees.   If I open the lash to avoid the long flat ramp , then the exhaust closed event is way off.  Either way the exhaust open event is way off.

This may account for some of the odd temperature related behavior on the CA55 carburetor, where it would start and run fine, the start missing as it started to warm up.  

I'm annoyed with myself for not fully mapping out ALL the cam lobes, and that's next.  

[BruceM]

The big oops is entirely my own goof-  when I reinstalled the camshaft I got it off one tooth. 
With lash adjusted for center of spec exhaust closed, and intake closed, my errors are reasonable;

exhaust open: 56 degrees BBDC (1 degree out of spec)
intake closed:   5 degrees ABDC (10 degrees out of spec)

The relocated IP lobe  (formerly 17 degrees ATDC) is now centered on 4 degrees ATDC.  This solves my timing advance problem.

Now for some start/load testing with spark BEFORE TDC!

[BruceM]

Back to load/ignition timing tests yesterday afternoon.  At CR=14:1  I have LESS power (opening carb has little effect when at full compressor load) than I did at CR= 13.25.  Knocking is pronounced unless spark retarded to an inch (flywheel)  ATDC.

 When last running at CR=13.25, the engine spark was 4 inches ATDC (retarded) as that was the limit of advance until I moved the IP cam lobe. With the full compressor load (pumping at 150 psi near 780 rpm) I still had advancing throttle response.

I plan it to fiddle some more at 14:1, perhaps get the ST-3 going for a known wattage load test, then reduce CR to about 13.5.

[BruceM]

I'm getting the best power at a bump clearance of 0.098", and spark timing near TDC at my 5600 feet elevation at about 70F.  Advancing timing further does not add power under full load, only raises the no load speed.  Increasing compression to 14:1 did not increase power.  I'm a bit short on power, though the engine is not broken in.  My operating rpm is  800 at 60 Hz on the generator head.  It will sustain 2000 watts, not much more.  850 rpm is the rated speed for the 8/1.  I'm not sure how much I could cheat that upwards and the next B pulley size larger for the generator head will exceed that.  Any thoughts on a safe max speed for the DES 8/1 on propane?

I did move my magnet to the flywheel with a wasted spark.  It gives me more stable timing, the IP cam method varied 3/4  inch or so on the flywheel as the hall effect sensor I'm using is not the stabilized linear variety.

I should have some help today to do more load and fuel mixture adjustments.  And hopefully a video.

I may hook up a CA125 that Gary loaned us today.  Starting and running on the CA110 has been problem free since I opened the idle bleed screw for improved no load running.  The main fuel adjustment seems to have little effect. No EGT testing, yet.

[dieselgman]

The 8/1 is designed for 850rpm... some of the makers pushed them up as high as 1,000. I do NOT recommend running that higher speed. 850 is fine, maybe 900 would not be pushing it too far.

I cannot really say what additional stresses might be encountered with the propane fuel but it should all be fine unless you are experiencing any pre-ignition knocking or pinging.

dieselgman

[BruceM]

Thanks, Gary.  I'll keep the speed below 900.   I think we are very close to meeting our power goal at 810 rpm/60Hz , so perhaps no pulley change will be needed.

Had some help yesterday and found better power at further reduced compression.  Bump clearance now at 0.122", spark timing at 1/2" ATDC.  I was able to run the air compressor at 160 psi and add another 500 watts on the ST-3 (no throttle left).  With the compressor unloaded (but belted as a parasitic load) at 2000 watts of load on the ST-3 there was still throttle response.  There is still a bit of knock when opening the throttle at full load, so we may try another reduction in compression, and I still need to try spark timing adjustment at the newly reduced compression.

I also need to try running my 1/2 hp submersible well pump, then starting the washing machine, with the compressor belted but unloaded.  That's our max load for Jeff's system, also.

The ST-3 on a single B belt driven flat on the flywheel seems to work fine.  The single B belt for the air compressor is OK so far but did have to be adjusted for some stretching so I'll keep an eye on it.  I went with single B's to keep a possible rpm change affordable.

CA125 carb testing soon, EGT measurement also.

[buickanddeere]

Sounds like you are the right track. Just too much mechanical compression ratio with a low rpm and a short duration cam . The cylinder pressure and heat from compression is taking the LP above it's auto ignition temp.
    Iirc about 11 to 1 is about max for LP withow rpm industrial engines.
  Is there any quench in the combustion chamber or is it open ?

[BruceM]

B&D- the Lister CS is an indirect injection engine with otherwise little in the way of concavity in the head.

I hooked up the Impco CA-125 carb today.  It starts and runs nicely without any intake restriction.  I do have markedly less power, and more clearly noticeable knock at load, but the adjustments all work and have an effect, including the main rich-lean adjustment.
I may fiddle more with it once the compression is lowered, again.