Odd Dyno Graph

#1

We went to the dyno today to diagnose the high RPM “miss” I had at Road Atlanta.

Sorry for the poor quality, but below is a photo. The strange things I see are:

  • good peak HP (160) but a very steep curve getting there (i.e. very little area under the curve - I am down 11 hp from July’s run from 5000-5500 rpm even when adding 4 hp to adjust from Dynojet to Dynapack)
  • Air/Fuel goes very rich (14.8 to 13.9) at 5250 rpm - and stays at that level

We put in a new (yes new) AFM which was richer across the whole A/F curve, but still had the 5250 drop. Swapped the ECU with another used unit. Other swaps for new parts included crank sensor, coil, wires, rotor & cap.

Does the ECU go to an open loop at higher RPMs?

The theory we are using now is that I have a bad valve spring that is allowing float at high rpms. We changed everything electronic we could think of, so it now seems to be mechanical. Any ideas?

Steve D.

#2

Steve D wrote:

Last time I checked 14.8 to 13.9 is lean and not rich. Ideally under full throttle I would want right around 13 for an all motor car. Rich would be 10s.

#3

These motors have a harmonic where the motor goes lean at a certain rpm. There is always a horsepower drop associated with the motor going lean. Looking at your dyno report, I strongly suspect operator error…it just does not look like a M20 motor dyno plot. You need to scan the printed version and post…make sure hp/tq/and fuel air ratio are shown.

BTW, the motor is dangerously lean at 14.5 to one.

If you have a broken valve spring I doubt the motor would make 160hp. It is easy enough to check just by pulling the valve cover and moving each rocker arm to check for play. Chuck

#4

I don’t know enough to say yes or no to operator error, but I doubt it. PFM has done a few pulls in their day and Fowler’s no slouch either.

Below is a HP/TQ plot. I don’t have air-fuel plotted on this one. The scale is rather large, which makes the changes in the curve seem very dramatic, but even so it seems like a wierd torque curve.

Anybody got any suggestions other than “it’s not an M20?” :laugh:

Steve D.

#5

160 is probably right at Carter’s HP cap so you should be fine. I’d just run it as is.

#6

IndyJim wrote:

The post of a very threatened man. :stuck_out_tongue:

Again, I am more than happy with that peak number, but I gotta work on the curve…

Steve D.

PS - IndyJim - I am just doing this for Nationals. Afterwards, I will crank it back to “mediocre” so you and Geegar have a shot.

#7

Yeah for nationals you’ll want at least 180hp and a flat 175-185 or trq.

You can probably work a deal to borrow Craig’s motor.

#8

Steve,

It looks like your VANOs is installed backwards. :silly:

#9

Are you sure you were getting a good tach signal for that whole run? Something is very wrong about that dyno plot. You don’t just lose 70 ft/lbs in 1000 RPM.

#10

The dyno plot is screwed up. It is a mathematical formula between hp and tq…the plot lines MUST cross at 5250 because this is the constant. Find someone else to dyno your motor…that plot is useless. Chuck

#11

Oh, yea, Scott. If you feel like removing the small bumpers and hardware to attach I will buy them from you. Chuck

#12

i think they DO cross at 5250, the scales are just different. look at hp at 5250 and torque at 5250, seems like the same value to me.

#13

Steve - you did check fuel pressure? My lean problem, similar to yours, turned out to be a clogged fuel filter.
Ed

#14

edavidson wrote:

[quote]Steve - you did check fuel pressure? My lean problem, similar to yours, turned out to be a clogged fuel filter.
Ed[/quote]

Ed -

Thanks. We will check that out. The strange thing is that it goes from one relatively steady state to another. It seems like a clogged filter would cause an increasingly lean condition as revs rise. But as you can tell I am learning here.

Steve D.

#15

cwbaader wrote:

Chuck -

As OriginalSterm pointed out, the plot is correct. The scales for HP & TQ are different.

I appreciate everyone’s input, but telling me that the dyno operator is incompetent isn’t accurate - or helpful. PFM’s resume is pretty strong. Plenty of Runoffs podiums, Spec Miata national wins, etc.

Is there something peculiar about dyno’ing an M20 engine? I could have posted a dyno plot on a 20-200 scale and it would look like I had a little dip in TQ, but I think the plots I posted illustrate the big dip in torque. Also, the lack of a nice horsepower “hump” is puzzling.

Steve D.

#16

Ok, y’all…this is what a good Spec E30 dyno graph should look like. Note it still has the harmonic rich point at 5K. Chuck Am trying to upload the dyno sheet. [file name=scan0003.jpg size=0]http://spece30.com/media/kunena/attachments/legacy/files/scan0003.jpg[/file]

#17

Here is one of my old graphs for reference:

Click Here

#18

Your dyno plot looks ok until around 4300RPM. Then, it goes crazy.

Here is Mr. Junky’s plot.

#19

cwbaader wrote:

Here’s Chuck’s dyno plot. I am going to try to compile some of the publicly-posted graphs on a matched-scale basis for comparison.

Thanks everyone for the data and input. And thanks to IndyJim for the protest fodder.:woohoo:

Steve D.

#20

Check the Throttle Position Sensor. It needs to know you’re at WOT or it will try to adjust to stoich (run lean). Tests are in the Bentley.