I've got a turbo 20B which I have rebuilt twice now.  I've
sorted out    wiring issues, had Tracy's EC repaired, checked the
timing, checked    for fuel and spark; repeatedly.    The engine was,
 possibly among other things, flooded.    Today I even tried engine
starter fluid.    I cannot get the engine to    even cough or
sputter as if i's even trying to start.

    Is there anyone within a reasonable driving distance of Canton,
 GA, an    hour north of ATL.  With Rotary and preferably Tracy EC
experience as    well that might consider    giving me a visit for
some assistance getting this thing started?  I'd    be happy to pay
your travel expenses.  Ive exchanged numerous emails    with Tracy,
 this newsletter and Mr. Anderson to no avail.    I am at my wits end.

    Jay

    Ed Anderson might fly in from NC.

    Flooding is unlikely for a 20B.
    Is the compression now good?
    I assume the spark timing has been checked.
    Do the injectors squirt?
    Do you have a manifold pressure gage on there?
    Can you borrow an oscilloscope from some body.

    There is also a free program that uses the audio microphone jack
    on all PC's as an input to a software O-scope.
    Use that to check the CAS. It works great on my Linux machine.

    http://www.zeitnitz.de/Christian/scope_en
    http://www.picotech.com/picoscope-oscilloscope-software.html

http://www.electro-tech-online.com/general-electronics-chat/38689-
good-cheap -free-pc-oscilloscope-program.html

    I do recall reading your emails on this subject.

    Send a picture of the intake manifold and injector locations.
    Apply 18 volts to the starter. Three 6 volt batteries in series.
    Don't let it crank for more than 15 seconds at a time. You
    could over heat the starter.

    If all else fails I fly UAL (retired A&P wife) for a modest cost
if    you could    meet me at ATL and provide room and board for a
couple of days. Or I    could rent    a car. The wife would want to
come along :)

    Paul Lamar

Jay - I am in Roswell - less than a half hour away, and would be
willing to help. 770-815-5140

Victor Roberts

Jay,
Start with the basics.....the very basics.
1. Turn off your fuel pumps and leave them off

2. Make sure you have compression. Do two test. One as it sits, then
squirt 20cc of auto trans fluid in each rotor and do another
compression test.
3. Check ignition. Simply pull the plugs out, clamp them for a good ground
and look for spark.

4. Make sure the timing is correct. I prefer an old style timing
light. Also can just make sure rotor 1 is on tdc and the timing mark
lines up on the distributor.
5. If all the above is working and correct, install the plugs, leave
the fuel system off and use a can of carburetor cleaner (berrymans
b12) and spray it in the manifold while cranking. It will fire up
and run if it has compression and correct timed ignition. Do this a
couple of times now just because you've worked so hard to hear
something. Its quite awesome to hear it fire up.

6. Now work on the fuel system.

7. fuel pumps still off, put the system in primer mode, push the primer
button and listen for the injectors to click. It should be easy to
hear. If you do not hear anything theres a problem. Either wiring or
computer. I would bet on wiring.

8. If you do hear the injectors click, pull them out, turn the fuel
pump on and do the same test. They should spray fuel only when
primer is operated and should not leak or drip otherwise. Be careful,
 obviously your spraying fuel around. The best way is to put the
injectors in small glass jars.

If all the above is working, your only problem should be mixture.

As a reference. I do not start my 20B in cold start mode. I squirt 6
primes in, turn my mixture knob to full rich, and it fires up every
time. As it warms up I start leaning it out

Hope this helps

Bob Mears

Thank you very much for the replies, suggestions and offers so far.

Although probably not the only/primary problem the engine did appear
to be flooded before my last start attempt.  I took the plugs out
and motored the engine and all sorts of mist spewed out the plug holes.

Compression on all three rotors is about 80 psi.

I get spark both during the EC3 diagnostic test and when spinning
the CAS by hand.

I'm using after market fuel rails where the injectors simply
slide/snap into place. I 'm not sure how I would test them off the
engine with fuel pressure. The injectors would simply squirt out.
There is fuel vapor coming out of the exhaust and using an engine
stethoscope I can hear them firing when I turn the CAS by and during
the EC3 injector diagnostic test.

I'm using the stock intake manifold.

I have a manifold pressure gauge,  I'll have to look again but I
don't think it's fluctuating during the starting.

I have two 12 v batteries wired in parallel.  Although trying 18 v is
certainly worth a try at this point.

Paul I would not only pick you up but let you stay at my house as well.

I don't think I have access to an oscilloscope but I will check
around.  I assume that's to check the CAS output?

Victor, I'll  be back in town tomorrow afternoon and will call you
Friday or Saturday to see when it would be convenient for you.  I'd
gladly bring the engine to you but it's mounted on a test stand and
it would be impractical for me to load it up for travel.  Even if I
had something big enough to carry it.

Bob, I believe I've done everything on your list at one time or another.
Most of them several times. I even tried twisting the CAS sensor at
90 deg intervals just in case I've managed to get the timing really
screwed up.  Not even a cough.  As mentioned I tried engine starter
fluid this last time and again, not even a cough.

Someone asked about my electronic prowess.  Other than a ham radio
license I got in the early 70's  I'm about one notch above beginner.
I have found wiring issues with the EC3 in previous troubleshoots
but I THINK I have it squared away now.

You know how sometimes you can struggle endlessly with a problem pulling
your hair out all the while with no apparent solution in sight?
Until someone walks up looks over your shoulder and spots the
problem right away? I think that's where I am.

Jay

I think Bob's checklist is an excellent idea. Go through that once
more and let us know what you get.

It should spit or pop once in a while. Maybe the timing is 180
degrees out. Do you have  an old fashioned timing strobe light? Do
you know how to find TDC on rotor one using a screw driver? You
stick it in the spark plug hole and feel for it :) A more accurate
way is hollow out an old spark plug and connect a clear plastic hose
held straight up. Fill the combustion chamber with oil and rock  the
e-shaft back and fourth so the oil level is a max in the hose. That
is TDC.

A very accurate way of marking top dead center is use the ring gear.

Paul Lamar

Jay, I assume you are using a 20B CAS. In that case there are two
CAS signals. One sends out a more or less square wave of 12 pulses
per engine revolution and the other sends out one pulse per engine
revolution around or before TDC.

This is why a scope is handy for checking the existence of these signals.

These are tiny magnet based  generators so they do not need any
voltage from the electrical system to work.

The 12 pulses per rev allows the computer to know when to fire the
plugs and activate the injectors at a particular e-shaft angles plus
or minus 3 degrees.

The three rotor would fire the plugs roughly around TDC for rotor
one. 120 degrees for rotor two and 240 degrees for rotor three.

If these wires are swapped going into Tracy's computer it will
confuse the hell out of it to say the least. Frankly I have no idea
what would really happen :)

The 12 pulses per eng rev is is white and red wires. The white is
ground I think. The one pulse per rev is green and white with a
black stripe.

There are 24 tits on the lower 12 pulse per rev wheel as the CAS
rotates at half e-shaft speed. There are two tits on the top wheel
for the same reason.

It is possible to make an engine simulator out of a solid state CAS
simulator and a hand vacuum pump from an auto parts store. The CAS
simulator is basically a 555 osc followed by a divide by 12 counter.
Some of your ham friends could knock that out in a heart beat. Here
is an oscillator part. I have not added the divide by 12 counter
yet. Probably two CMOS CD4018's would work. One divide by 6 and the
second divide by 2. We are talking $10 in parts.

This allows testing Tracy's system without actually turning the
engine over or running it.

Paul Lamar



A DMM can be used in frequency mode for a quick indication of CAS operation.
The sensor for the 12 pulse/rev with have a higher frequency than the one
pulse. Timing is a different deal but it will tell you the triggers are
functioning and wired correctly or not.

Dale Davies