I suspect Tracy's PIC has its hands full at high RPM
 reading this :) The single pulse by it self is probably
 rotor one. The wide pulse is probably rotor 2.

 I have no idea what Tracy is doing. He probably
 considers the single pulse is the same as the
 old 2nd gen crank angle sensor.

 The easy way is to use 2 or more sensors, one for
 each rotor, with only one tit on the wheel.

 The pictures say it all. The router motor and
 Harbor Fright (:)) motor control is the hot tip.
 This thing runs smooth to ungodly RPM's (up to
 22,000) thanks to Jim Poole's accurate machining.

 If you want a RX8 trigger wheel hub and you need to ask
 the price you can't afford it :) I have a spare one.

 It is putting out a very strong signal with .2 inch
 between the sensor and the wheel. On the order of
 8 volts peak to peak. That is a stock, high buck, RX8 sensor.
 Ill try it soon with the cheap $16 ABS sensor.

 Be careful. Don't let it come apart. It can kill.

 Paul Lamar

 Paul,

 That's great progress.  I would be interested to see what effect a
 non-concentric timing wheel would have.  I found that mine had a
 .0012-.0013" variation in the gap.  I decided to go ahead and switch over
 to
 the Renesis water pump which will use the Renesis trigger wheel and hub.
 This should cure the problem with changing gap, and possibly cure the
 miss
 too.  It sure won't hurt anything.  ;-)

 Mark S.

 That out of round is not going to help. It will vary the amplitude
 of the signal once per rev. Jim Mederer tells me that is a common
 problem with the e-shaft on the 3 rotor.


 Good idea to go back to the 2nd gen CAS. Does Tracy have a different
 software package for that or will it work with either?

 I have no idea what Tracy is doing but reading the RX8 trigger wheel
 and firing a three rotor is not an easy  thing to do in software.

 How are you coming with the Hytech ?

 Paul Lamar

 Paul,

 We had our two grandsons for the weekend, so I haven't had time to mess
 with
 it.  It was all I could do to keep them from burning down the house.  I
 did
 manage to slip away and go to the airport on Sunday for a few hours (as
 described above).

 Anyway, I'm a bit confused on the Hytech's purpose.  I know it is a data
 logger, but you have connected some sensors.  Is this to do compression
 checks, or verify manifold vacuum, or both?  I didn't want to exceed the
 limits of the sensors by hitting them with 100 psi if they are meant for
 vacuum.

 Also, with the water pump off, I won't be testing anything for a while.

 Mark

 Those are for manifold vacuum and a crank angle sensor fed into
 one of the A to D.

 Your right. The sensors would blow up. These are the same sensors
 Tracy's use in the EC2.

 What you will get is a lot of raw data that looks like this.


 BTW I don't think the 15 thou run out on the end of the
 e-shaft is the cause of the miss. At what RPM is it missing?

 I think you should borrow my RX8 trigger wheel hub and test
 the EC2 with out the engine running. I think it has a software
 problem over about 6500 RPM. You won't need the water pump
 for that as they engine will not be running. If the injector
 pulse width is not getting wider that is the problem.

 In fact come to think about it I suspect Tracy's system
 quits working after the torque peak as it is based solely
 on manifold press and max HP comes well after peak torque
 AND peak man press. If it has trouble sensing RPM
 over 6000 it will not feed the additional fuel needed into
 the engine at the same man press at the higher RPMs. Particularly
 since you moved the torque peak down.

 A car EFI systems uses a mass air flow sensor so
 air flow is proportional to RPM as WELL as throttle
 opening.

 That explains a lot.

 We need to hook up the Hytek and see what is going
 on in Tracy's system over 6000 RPM with out the engine
 running.

 It is like trying to determine what some one is thinking
 solely by observing their actions :) The hood is welded
 shut :)

 I have an early EC2 from the days when nobody went over
 6000 RPM. Tracy has changed the software from that time.
 We tried it on the p-port eng on the dyno but could not get
 it tuned. We gave up at $650 a day and put the Weber on.
 Perhaps we were having the same problem.

 It might be interesting for me to hook it
 up with the CAS simulator and see what it does
 with my scope.

 The bottom line is what does yours do?

 Paul Lamar


 Paul,

 I agree, it would be very helpful to do a bench test on my EC-2.  Bobby
 Hughes has a scope we can use and I'm pretty sure I can rig it up using my
 trigger wheel and my router/speed controller.  Bob Darrah could fab an
 adapter to fit my router.

 There are two reasons I think this may be related.  First, I can get
 different results by changing the gap between the timing wheel and CAS
 pickup.  Second, Tracy recently commented that he had a misfire caused by
 vibration of his CAS trigger bracket.  He built a stiffer bracket and the
 miss was gone.  As I see it, a wobbling trigger wheel would cause the same
 symptoms.

 So, I think I will finish installing the Renesis water pump with new crank
 hub and see if there is any improvement in how it runs.  I can then do the
 bench test if it still runs poorly.  Keep in mind that before changing
 engines this ignition setup would give me 7200-7300 rpm during the first
 few seconds of my takeoff run.  So, something has obviously changed, I
 just
 don't know what it was.

 It may be that my exhaust system is too restrictive and is causing exhaust
 gases into the intake charge at higher rpm.  Lynn Hanover has stated that
 the P-port engines don't tolerate exhaust restriction well.  That makes
 sense as a restrictive exhaust could force exhaust gases into the adjacent
 intake during the overlap period.  Come to think of it the rpm at which
 the
 engine began rough-running dropped after I installed the new muffler.  I
 may  be on to something here...  So, this may not even be ignition
 related after all.

 One question, what size air inlet should I have feeding the three 1-7/8"
 ID
 runners?  Currently, I have a 3" SCAT hose with K&N filter.

 Mark

 I'll loan you my hub. No need for Bob to make one.

 Well if you just go on area alone three runners are  8.28 square inches.
 3 inch scat is 7 square inches. There is also skin friction drag pressure
 loss in the tube due to velocity squared so I would go up a size or two.
 There is 3.25 OD and 3.5 OD SCAT and SCEET.

 SCAT is rough inside. SCEET is smoother but the inner liner
 has been known to come adrift and plug it up. If you switch to
 SCEET sew the inner liner to the outer liner with small diameter
 safety wire.

 Thin wall aluminum tube is best but I think about 3.5 inches OD would be
 needed. You might want to roll your own out of .025" thick aluminum.

 Send me some pictures of the P-port version of the 3 rotor.

 Paul Lamar


 Paul,

 I've enclosed two pictures that show how it all came together.

 You've confirmed that I need to enlarge the size of the intake tube.
 Maybe
 I can get some mandrel bent aluminum tubing and weld the pieces together
 so
 that I only have two small pieces of flex hose.

 Mark

 Nice tangential muffler.
 Can you get a Harbor Freight inspection camera up the exhaust
 pipe far enough to see the small holes in the center pipe?
 Some could be failing.

 Also nice job on the K&W duct on the oil cooler.

 Can I see a picture of the other side of the engine? :)

 Burns Stainless has some .065 wall 3.5 inch OD aluminum
 90 degree bends for $105. Too expensive and too
 heavy.

 Carve it out of Styrofoam, wrap with vinyl electricians
 tape and make a layup. The foam can be dissolved with acetone
 and the tape fished out.

 Paul Lamar

 Paul,

 Yes, I have a Harbor Freight LCD inspection camera and use it regularly.
 I
 can work it up the exhaust to where I can see a few inches into the
 muffler.  So far, everything appears to be in tact internally.

 Mark

 The muffler looks pretty good from the outside. How many holes
 in that center tube and what is the diameter of the holes?
 In other words what is the total area of the holes?

 Somebody, I forget who, said that tangential mufflers can build
 a complex internal pressure field due to the internal vortex.

 The center of the vortex is low pressure while the outside
 of the vortex is higher pressure. If this is true the center
 of the field will be cooler as well. You might be able
 to verify this with one of those non contacting infrared hand held
 temperature gauges. The outer part of the vortex would then
 apply more back pressure on the engine. That is the theory anyway.

 Therefore you might try a separate 2 inch down pipe from the outside
 of the can leading out, down and back. Make it temporary so it can
 be removed and plugged up if it does not help. Leave this for
 later after you have analyzed the electrical situation with the Hytech.

 Robin is out of town on biz but I may be able to come down for a day
 or two next week. Remind me to bring my infrared temp gage.

 Paul Lamar


 Paul,

 That is a generous offer, but I work days (M-F) and our daughter is coming
 to town Wed through Monday (6/16 - 6/21).  I'm not sure what all they
 have planned, but I really want to spend some time with her.  She lives in
 Denver, so I don't get to see her very often.

 I need to order parts in order to get the opto-isolator board wired up and
 ready to go.  You mentioned a Panasonic part, do you have the p/n?

 The tube running through the tangential muffler has 109 slits made with a
 carborundum cutting disc (too &^%*$# difficult to drill holes in
 inconel).  The holes were then spread apart with a punch to form a bunch
 of
 diamond-shaped holes.  So, I would guess that the holes are 1" long and
 3/8"
 at the widest.  I realized too late that it would have been better to
 angle
 the slits to "encourage" the gasses into the inner pipe.  I figured I
 would compensate for that by using lots of holes (slits).  I had thought
 about adding an extra exit, or two.  Or, I could open it up and replace
 the
 2.5" inner pipe with a larger pipe.  Or, I could remove the inner pipe
 altogether.

 Mark

 I sent the jpg with the pin number on it.
 Did you not get it? Here it is again.
 It is a Digi Key CNS 7s something. 77 cents each.

 Your muffler could be acting like an air compressor
 with the hot air swirling around what amounts to a bunch
 of cowl flaps. Air could be flowing into the exhaust pipe :)
 Just add another exit and see if it helps.

 Paul Lamar

 Paul,

 OK, the opto's are on their way.

 Mark

 At 250 RPM and .05 inch the signal is 5 volts peak to peak.
 If the engine starts it must be close enough.
 The scope picture is one half of a revolution.

 Paul Lamar

 Paul,

 What is the output voltage at around 7k rpm with a gap of .0050?

 Mark

 About 25 volts peak to peak but don't try to  cut that down. That
 is Tracy's job.

 The picture says it all. The upper character line line
 on the scope say it is 1 volt per division but
 I am using a 10X probe so it is really 10 volts
 per division. Hence 25 volts.

 The sweep is set to 1 ms per division
 so as you can see there are about 8 division
 between rotor one and back to rotor one. 8 ms
 is about 7500 RPM.

 Anything you might do to preserve the
 wave form AND cut it down will affect the engine starting.

 Set the EFI computer up on the bench. You can simulate the injectors
 with a high wattage 5 ohm resistor connected to 12 Volts.
 The bottom of the resistor will be switched on and off by
 Tracy's computer. Use opto circuits  on the injector signals and
 feed them into the digital inputs on the Hytek.

 Connect the CAS simulator. No need to connect
 the manifold pressure sensor hoses at this time. Leave them open.
 That simulates WOT.

 I think you have other problems and nothing to do with the CAS.
 If you are using an MSD ignition or something like that
 you can use the Hytek to check the integrity of
 the signal coming out of Tracy's computer.

 Set the Hytek (poor man's storage scope) sample rate
 to 50,000 samples per second.

 You only need to run it for a few seconds as it will
 acquire a mass of data on your hard disk very quickly.

 This way merely checks to see if the ign pulse
 is going out in a timely fashion. In that case use
 opto circuit and feed the 5 volt output pulses to digital inputs
 on the Hytek. The data will show any missing pulses
 if any and the width if it is greater than. 1/50,000
 of a second.

 Paul Lamar


If that scope trace is with the probe on the 'hot' connection to the CAS,
the sensor is connected backwards.  The pulse must start with the positive
going edge.

tracy

If that was IMG_0547 one wire had a diode in series. Something
Mark S tried. Sorry I mis-led you about that.

These two scope images are more typical. No diode.
The first is with the wires one way and the second with the
wires the other way. Which IMG number and corresponding
connection do you prefer?

People do not always use the stock connector. In this case
I used small alligator clips. I see the stock connector is
keyed. Can you provide a sketch looking into the stock
unit connector and identify which is which?
Or have you done that and I did no hear about it?

I take it you are mainly  concerned with the single pulse
in the center of the wide gap?


Paul Lamar

Paul,

The zener diode was connected through a resistor to ground.  There was
another resistor (approx. 20k) in the "+" lead, best I recall.  I'll look to
see if I can find a sketch of the circuit.

Mark

OK I'll reproduce it and see what it does to the wave
shape on the storage scope. If you have any spare diodes
you don't need drop me one in the mail. Or let me know what
value it was and I might find one at Fry's.

I still think that is not a good thing to do.
Tracy's appears to be dependent on a fast rise time.

Paul Lamar