Dissapointment at the dyno

[Scott R]

So, four pulls on the dyno to 6200 RPM and my best HP was 141. The dyno guy tried a ton of tweaking on the EFI system, but nothing made any difference. The best we can tell is that the injector pulse width gets maxed around 4K.

I can post a pic of the dyno sheet, but it's pretty depressing. And based on this my EFI system is not the way to go on this engine. It's to finecky, and does not appear to be able to support my engine configuration.

What would you do? Find a large runner 79 CIS system? Switch to carbs? I would like to hear what others with modified engines are doing.

[ianc]

Have you tried contacting Tony directly? I'd be interested to hear his response, and he really should be notified to provide advice...

ianc

[Scott R]

Originally Posted by ianc

Have you tried contacting Tony directly? I'd be interested to hear his response, and he really should be notified to provide advice...

ianc

I have, he's been looking at my logs for weeks, not too much in the way of replies though.

[sfoster13]

something is not right.

what was your AFR and how did it look across the rpm band?

what kind of torque numbers did you see?

Out of curiousity, what kind of dyno?

[stlrj]

When I got totally frustrated with my OBD I Chevy 2.8 TPI EFI project on my 78 3.0, I always went back to the stock CIS. This gave me time to cool off, think about solutions, and then go back and try again.

Best part about going with this domestic system was the incredible free support I got from the Chevy service department.


My final solution was updating to an 86 3.2 and tweeking that...HEI ignition, GM Multec injectors, HEI ignition cables, cap and rotor etc.


Needless to say, I feel your pain.



Cheers,

Joe

[Jim Sims]

Were your expectations calibrated to account for NA engine power loss due to altitude? There in Denver the loss will be about 15%.

[sus911]

Don't worry too much.
Different dynos will you give you different readings.
Check this for related discussion:
stock 3.2 Carerra engine out
good luck,
Phil

[Steve W]

Originally Posted by Noah

You need to post your dyno scans, especially the AFR's.

+1, and if you are maxing out your injectors, you'll just need larger ones.

[Scott R]

Here is the highest dyno sheet, have several more that were slightly less as we tried to lean the top end out.



Here is the AFM readout from my LC1, and the corresponding logs.

[ischmitz]

Scott,

The second chart is an eye-teaser Am I correct in that the first half is the pull and the second half is the engine idling?. What is the step in the AF ratio half way into the pull. That does not look right.

It looks like you run very rich before the step but it is hard to see the units on the graphs. Can you blow this up a little bit? From the graph it looks like you run at less than 8. That is strange.

Are you sure the wide-band sensor data is correct. I have played with a MS-II (not the TBITZ) and Megatune on a turbocharged Type 4 engine. There are many ways to mess up the calibration of the LS-1 wideband sensor data.

Are you using Megatune or something else to tune the Megasquirt?
Is it Megasquirt or Megasquirt-II?

Cheers,
Ingo

[2.7RACER]

The second chart is not decipherable. Where is the RPM's?

[Scott R]

Originally Posted by 2.7RACER

The second chart is not decipherable. Where is the RPM's?

6000 peak, it's the yellow line. I can read it but I'll try and blow it up.

[Scott R]

It's a MS 1, 42lb injectors, and we calibrated the LC1 before the test. I'm attaching three datalogs for different pulls. We went from both rich, and lean, in the tables, and when we leaned it out, the HP went down.

http://home.comcast.net/~sramnit/datalog200710060909.xls
http://home.comcast.net/~sramnit/datalog200710061014.xls
http://home.comcast.net/~sramnit/datalog200710061025.xls

The car was giving everything it had, but it's a slow poke for sure. I'll also add that this system has never run well, always requires tinkering, the idle has always been an issue, never stable. I have people look at the logs and they say, "looks good."

Little background on the engine, it's 1000 miles old, 9:8:2 JE's in new jugs, full rebuild top to bottom, with 964 cams and SSI's. The injection system has been smoke tested for leaks and nothing was found. All ignition components are new, or rebuilt, distributor, MSD, coil, wires plugs. It just won't make any power.

[coollx]

If you haven't already, you should probably do a leakdown test on all cylinders to make sure the valves and rings are sealing well. Just a thought.

By the way, what HP level were you hoping for?

[Scott R]

Originally Posted by coollx

If you haven't already, you should probably do a leakdown test on all cylinders to make sure the valves and rings are sealing well. Just a thought.

By the way, what HP level were you hoping for?

Leakdown was done at the major service last weekend, less than 2% all around. I was at least hoping for stick numbers. (btw engine was built by a very well know tech here in Denver.)

[DUK]

Scott, How are you controlling your spark? Can you include your msq? I'm no expert but I can try to help. Looking at your datalog, I'd say it's something in the accel/decel area. Just my .02

[Scott R]

Originally Posted by Duke Zink

Scott, How are you controlling your spark? Can you include your msq? I'm no expert but I can try to help. Looking at your datalog, I'd say it's something in the accel/decel area. Just my .02

Spark is controlled from the original SC distributor, this feeds the MSD unit, it's not controlled by the MSI.

Here is my MSQ file:

http://home.comcast.net/~sramnit/megasquirt200710061203.msq

[ischmitz]

Scott,

I guess you are aware that the LC-1 controller can be configured to translate the measured A/F ratio into different voltage ranges. For example you can have it to put out a linear 0 - 5 volt range corresponding to 10 - 20 AFR. This setting is independant from calibrating the O2 sensor in free air.

In MegLogViever you then need to select the previously choosen mapping method of the LC-1 to make sure the voltages reported by the LC-1 controller to Megasquirt are translated into the correct AF ratios. If you have a missmatch your real AF ratios are slightly off. If you for example have the LC-1 map 10 - 20 AF into 0-5V but have chosen "default LC-1" in the Calculated - Wideband O-2 AFR option you are off by about 2 counts on the AF.

I assume that you are not running MS-1 in closed loop mode, correct. Still, what you see in Megatune for A/F can be missleading if the calibration methods of the LC-1 and Megatune are not matched.

I have plotted your data with two different settings in MegaLogView. The first if with "default LC-1" choosen and the second is with 0 - 5V for 10 - 20 AFR choosen. The seconds looks more "correct" I assume this is what the LS-1 is set to. However, this is just to illustrate how you can be off by a little bit and leave power on the table.

BTW, I don't see the injectors maxing out in this plots. They look fine. I wonder why the LS-1 has these blackouts. This seems strange. Is there a wiring issue with your setup?




Here the same data with "10 - 20 AFR 0-5 volts" choosen

[DUK]

Scott, what's the timing graph look like? What's your max advance, idle setting, etc?

[Scott R]

The LCI blackouts can been seen on the Logworks software from time to time as well, it shows "heating" or "calibrating" when it does that. I have it driven off it's own relay as they suggest, same for the MS I, and the MSD.

When I was saying injectors maxed out, I was referring to the pulse width on the graph, it seems to be over 90% then right up to the limit of the chart at or before 5000RPM. Is that normal?

Given that the AFR's were grossly off, which would be strange, since we tried all of the 4500 - 600 bins at both lean and rich conditions and nothing happened, well other than one of the pulls going down to 138 hp. I just don't seem to be getting performance, or stability out of this system. Two years, and an engine later and it still won't idle without hunting!?!