[rusEfi] Hunchback 1991 Miata NA/NB2 rusEfi official racecar #20

Your chance to introduce yourself and your vehicle
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Nicely done!
It looks like the VE tune is quite solid already. I also really like how stable the VVT angle is.
One thing: There are quite a few trigger errors and some of them severe enough to show up as drop in rpm and sudden change in VVT angle. It shouldn't really hurt performance all that much, but since my '03 doesn't have these, I wonder what causes them. Bad sensor maybe? If so, better have spares handy for the race.

And for the next session: To the limiter and beyond!
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

How much dyno time do you have? Enough to tune VVT and spark?
I'd probably try to run a few different VVT tables. Probably with fixed advance and see what angle makes the best torque for a certain rpm band. Then assemble a final table from these values.
Same with spark.
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Any more progress yesterday? What time's the dyno?
Abricos
contributor
contributor
Posts: 849
Joined: Mon Aug 18, 2014 12:32 am
Location: Carteret, NJ 07008

Re: 1991 Miata #20 Hunchback

Post by Abricos »

stefanst wrote:Any more progress yesterday? What time's the dyno?
1:00pm
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Spark wire pickup did not work right so dyno was not showing RPM, there was also a bit of a mis-communication so only did a few pulls.

120whp

https://svn.code.sf.net/p/rusefi/code/misc/logs/1991_miata_hunchback/20170501_quick_dyno_session.7z
Attachments
20170501_123451.jpg
20170501_123451.jpg (7.41 MiB) Viewed 35524 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: 1991 Miata #20 Hunchback

Post by puff »

lower than projected?
yet another cool project - building your own dyno!
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Pretty much dead-on with expectation. I'd guess we could find a few more hp with more aggressive spark. Maybe re-map the VVT curve.

A few observations:
- We still get some errors 6036, 6037 and 6045. Nothing that's going to kill anything, but definitely odd
- From low rpm @102kpa to high rpm at 94kpa, we're losing quite some pressure in the intake. It looks like we're seeing quite some intake restriction. How old is that airfilter? Is the MAF still installed? If we had 8kpa more at top rpms we'd have 10hp more!
- The VVT angle seems to change suddenly, not smoothly. Is the VVT target top row still @85kpa? If so, I would increase that to 120kpa or so.
- In the 12:22:54 log we see AFRs shooting up at around 6600rpm. That's before the throttle lift. Any idea what happened there?

@: I don't think the increased pressure from detonation actually increases torque. From what I learned, about 25 years ago (so take that with a grain of salt) detonation causes the pressure peak to occur too early in the power stroke, actually reducing torque/hp. So tuning for MBT (maximum brake torque) should avoid detonation in theory. Still, verifying detonation independently would definitely be worth it. Blowing up engines is no fun.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

stefanst wrote:How old is that airfilter? Is the MAF still installed?
AEM 21-205BF DryFlow Air Filter was ordered on 12/23/15
Yes MAF is still there, it holds the filter on the intake pipe. Now that we have no chance to stock ECU it's time to remove it.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

stefanst wrote:We still get some errors 6036, 6037 and 6045
Also pretty scary cranking if it's not me who is trying to crank it. I know a better way, the guy at the dyno did not:

[video]https://www.youtube.com/watch?v=RDG9R2GkHGo[/video]

[video]https://www.youtube.com/watch?v=wpTcVkiYLCQ[/video]
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Actual dyno pull

[video]https://www.youtube.com/watch?v=dLd5Z8nhBWs[/video]
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Sounds healthy!

BTW, just for comparison purposes, stock 1.6l (which is where you're coming from) usually makes 90hp on the dyno. So you already gained about 30hp!
Somebody
Posts: 6
Joined: Wed Sep 14, 2016 10:47 pm

Re: 1991 Miata #20 Hunchback

Post by Somebody »

No torque reading off the dyno huh? I'm impress by 120whp, that is at least 140 at the crank. I'll do some digging to find some sort of representation of a torque curve to figure out the best shift point. Those starts by the Dyno guy worry me, I mean I didn't screw the starts up that bad...
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Somebody wrote:No torque reading off the dyno huh?
No RPM reading is probably the reason why no torque reading. No RPM reading kind of suprises me a bit, they were puzzled by our wasted spark with half COPs half wires and coils in the middle of all this.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

We believe dyno runs were done in 3rd gear, does it matter which gear is used?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: 1991 Miata #20 Hunchback

Post by puff »

lower gears increase torque...
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

On a non-turbo engine it really doesn't matter all that much. Power is measured by the rate of acceleration of the inertial load. Engine torque is then calculated as power [W]/rotational speed of the engine [1/s]. Since we don't have engine rpm, we don't get torque.
On a turbo engine I would recommend higher gear, so we'd be in more of a static state- more time spent in each VE cell, but on an NA engine it really doesn't matter.

Shift points, btw, are determined by the power curve, not by torque curve.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Cranking is still not great, but once car is running it's running strong.

Checked timing with a timing gun and it's spot on. Too often it helps to floor throttle pedal in order to get into "cylinder clean-up" mode, wonder if I am dumping too much fuel on cranking and it's the extra fuel which is back-firing so violently?

https://www.youtube.com/watch?v=6mHJWxmevoU
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Things look great so far!

We have failed tech inspection because chief inspector was angries than usual - he has failed us for exposed fuel tank filler rubber neck behind driver (miata thing). But we got lucky - a random guy at the track just happened to just buy a $150 miata shell just for suspension which has that parcel shelf metal we needed to make car legal again.

Starting is pretty good now. What's not so good is that random miss we have every once in a while - but we can totally live with it I think.

Logs: https://svn.code.sf.net/p/rusefi/code/misc/logs/1991_miata_hunchback/20170512_njmp_test_day.7z
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Looks like the miss is pretty much always associated with error 6037 - Custom Synch Count Mismatch. Not sure what that tells us, but maybe @ knows....
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Timestamps are weird but I think these are Saturday wet racing day

https://svn.code.sf.net/p/rusefi/code/misc/logs/1991_miata_hunchback/20170513_njmp_saturday_wrong_timestamps.7z

2 hours Andrey, 1.5h Ethan and 1h Rich
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

russian wrote:Timestamps are weird but I think these are Saturday wet racing day

https://svn.code.sf.net/p/rusefi/code/misc/logs/1991_miata_hunchback/20170513_njmp_saturday_wrong_timestamps.7z

2 hours Andrey, 1.5h Ethan and 1h Rich
I was spectating this day and it was a miserable, awful rainy affair. I was in the garage, and even reasonably protected from the rain it was miserable. I was soaked just from being outside every once in a while.
The cars were SLOW! The Miata is capable of running 1:45 laps at this track, possibly a little better, and I believe the best time we had was 2:30-ish.

A testament to how awesome racing is, is the fact that the drivers still loved it. They were completely soaked after minutes (windows on a racecar are underrated), but still were enthusiastic. So enthusiastic in fact, that some of them, in their exuberance, extended the track onto the grass!
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Today was not auto racing, it was relay swimming. Videos of puddles #1 and #2 would be epic! Dirt creak was also pretty cool.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Preview of the Saturday swimming session
https://www.youtube.com/watch?v=ac8ioQcGeNs
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

I have a feeling that I was mostly passing in the rain on Saturday while I was mostly being passed on Sunday.

I wonder if I did so well simply because our windshield was perfectly transparent while too many people were struggling with poor wipers and fog on the inside?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

russian wrote:I[...]I wonder if I did so well simply because our windshield was perfectly transparent while too many people were struggling with poor wipers and fog on the inside?
That may be a factor, but the most important factor in a rain race is big brass "equipment". Not caring if you live or die can make you go faster.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Well, Sunday shows that I am a cautious slow driver :( Need skills.

[video][/video]
[video][/video]
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

Could you please post a few more logs- especially from Sunday? i would like to see if I can isolate a few full-throttle runs and analyzer with virtual dyno.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1991 Miata #20 Hunchback

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: 1991 Miata #20 Hunchback

Post by stefanst »

As some of you may already have heard, I'm having some trigger-troubles with my MS3X on the '99. While researching this, I cam across something that may be of interest to the Miata VVT crowd:
Interesting side not...

We are working on the ME221 based PnP for the MX5 NB - We use the Max9926 as our input conditioner - currently we are seeing the crank signal 'dissapear' at around 2000rpm then come back, then go again (at around 500mS interval on the logic analsyer). I am putting this down to the fact we are use 470ohm pull-ups to 5v as opposed to 3k3 (as per OEM) so I suspect the sensor is overheating or something to that effect - going to try 3k3 today and will report back Maybe helpful for those who have had this issue (if any!)

Matt
Matt later confirmed that this did fix his problem.
https://www.miataturbo.net/megasquirt-18/abe-jasoncs-nb-cam-crank-input-circuits-25789/
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: 1991 Miata #20 Hunchback

Post by kb1gtt »

Does the Miata use a 12V pull up?

As noted on the Frankenso schematic, I expect for a 12V pull up that you'll want 3k, so 3.3k seems reasonable. However if you have a 5V pull up voltage I would expect you want something more like a 1k pull up. The default for Frankenso is a 5V pull up with a 1k resistor. I believe this matches your Miata sensor close enough.

13.4V/3.3k = 4mA
5V/1k = 5mA
Welcome to the friendlier side of internet crazy :)
Post Reply