[Hillmanjames]

I would like to request that we can blend the MAP and TPS based acceleration enrichment in a similar way to how you can on a Megasquirt. I have attached screenshots below of what we currently have and how the mega squirt does it. It would help me greatly getting my ITB's smoothed out (either has a stumble coming off slight throttle using TPS based AE, or can't add enough in when required above 100kpa when in MAP based AE. A blended mode sorted this out on my last build)

Current Speeduino way of doing things:

Speeduino AE.png (52.4 KiB) Viewed 8324 times

How it works on megasquirt:

Megasquirt blended AE.png (90.63 KiB) Viewed 8324 times

Thanks in advance, James

[george-angelosvoudouris]

James can you please share a pic of your VE map?
As it happens I daily drive with my ITBs and it's on TPS only map so I can share plenty of info in that kind of matter.
Normally the stumble you describe is due to the mixtures going to lean as you open the accelerator.
When you tune ITBs you need a closer grouping for the first couple of degrees of the butterfly because airflow is non-linear at that point... after that, you can expand your grouping gradually until half your throttle and beyond that, you can go for a 20% change until you reach WOT.

Also setting up the accel enrichment I would make it give a bit more fuel on smaller throttle plate angles to reduce the lean spike.... in general, there is always a delay before accel enrichment kicks in the goal is to make it as small as possible

[PSIG]

Also setting up the accel enrichment I would make it give a bit more fuel on smaller throttle plate angles to reduce the lean spike.... in general, there is always a delay before accel enrichment kicks in the goal is to make it as small as possible

+1. This is the issue. TPS AE has never worked quite right due to coarse reading frequency and % required, and while you can blend to "improve" it, the correct fix would be to develop both TPS AE and MAP AE fully. It works fine for a drag car or other heavy throttle action, but not for maintaining Lambda during smaller or finer throttle changes.

[Jakobsen]

I also have the same issues as above

Is there going to be an update on the firmware to help this issue ??

[jonbill]

Better would be good, but I have no problem with AE. I have ITBs with Alpha-N so the MAP isn't used at all. I have done what George said above - the bottom 5 rows of my tables are 0,2,4,8,12 %TPS. I've also got TPSDot threshold low at 5% and no filtering on the TPS.

[Hillmanjames]

I have tried two routes so far:

First off i used tps load only, bottom rows i set to 0,2,4,8,8,12,16. after 100 miles of road tuning i was still getting lean spikes so tried 0,2,3,4,5,7,9,12. It was better, but still not great (90% of the lean spikes were between 2 and 10%)

Next up i used switchable maps with it switching at 94kpa from a MAP based fuel table to a TPS based fuel table. Couple days tuning that seen an improvement over the best TPS only fuel table but still not 100%

Using MAP based acceleration enrichment and a switching fuel table has given the best results so far.

I think that using TPS only isn't working as the difference in air flow between 0 and 1% throttle is far too large on the throttle bodies i have. (cbr1000rr throttles - roughly 44mm diameter)


I might try altering the TPS calibration so that 70% actual position shows up as 100% in tuner studio to get a better resolution on the tps?

James

[george-angelosvoudouris]

James, I've been running my ITB's quite some time now only on Alpha-N (aka TPS only) map...

Even with a single TB, you'll never gonna get an Accell that doesn't have a spike whatever ECU you gonna run... As ECUs are man-made machines can't predict natural phenomena or drive quirks or call it whatever you like... The only thing that they can do is to follow up on the phenomena that are happening at that time...

For that reason, you'll always get a lean spike even if it's a little tiny one that your physical body can't notice or the equipment you have can't track it...

Your only option is how fast you can cope with that spike... If for example, you're running an engine with ITBs on the lean side (lambda >1) the difference will be greater then running it a bit on the rich side... The main reason that ITBs are getting big lean spikes when you lightly touch the accelerator is because of the large total crossectional area (you with 44mm ITBs have an area of 1520.53mm^2 times 4 (6082.12mm^2 total) and someone with a 75mm single body has 4417.86mm^2).

In terms of tuning now the ITBs, I tend to go with a TPS scaling of 0,2,5,10,15,20,30,40,60,80,100 and never had an issue..
on the Accel side of things what I usually do (options depending) is to ramp the fuel in terms of how fast the rate is changing (small throttle change small quantity added big change big quantity added). If you don't have the option to adjust your fuel accel rate of change and just the quantity you have to compensate for that in your fueling map...

If you can provide some info on the setup might also give some hints on why you get the lean spikes. (engine displacement, an idle valve on the setup, etc. also some logs and the map will be good for troubleshooting)

[Hillmanjames]

I have been fiddling on over the last few days. To simplify things i went back to a TPS only map and started from scratch. This got me back to where i was initially, so i don't think that the VE table is having a massive effect on the lean/rich spikes.

I then tried lowering the Throttle position sensor filtering from 50% down to 20% (lowest i could go without any noise becoming an issue) This helped , but did not solve the problem.

A few days ago i printed and fitted some 50mm trumpets as before it went straight from the throttle inlet to the air filter backing plate. This improved the flow, requiring the whole map to be richened up between 1500 and 4000 rpm and also amplified the rich/lean spikes.

I data-logged my whole drive home today and studied it before writing this. I noticed two scenarios that are causing me the issues tuning the accel enrichment which i doubt ill be able to remedy using solely TPS based enrichment.

Scenario 1: cruising along at 2500rpm TPS increases from 48% to 60% giving TPSdot of 40 and enrichment of 180% resulting in a rich spike (9.5 afr) and the car falling on its face until it clears. Solution, Reduce accel enrichment.

Scenario 2: Coasting along at 2500rpm. TPS increases from 0% to 7% giving TPSdot of 40 and enrichment of 180% resulting in a lean spike (18.2 afr) and the car jerking before picking up again. Solution, Increase enrichment.

So with these two scenarios, the ECU will not see any physical difference so i dont know how to get around this without using MAP based accel enrichment as well (as i have in the past) to over come this. It makes it very awkward to drive as setting off causes the engine to buck (scenario 2) which, when tuned out by richening it up, makes Scenario 1 worse.

Below are the current VE, Target AFR and Acceleration enrichment tables.
The engine is a Mx5 1.8 on CBR1000rr throttles. Stock mazda triggers, injectors, Yaris COP's and i am using a DIY EFI Core4 standalone ecu

accel enrichment.png (263.05 KiB) Viewed 8025 times

AFR Table.png (384.35 KiB) Viewed 8025 times

VE Table.png (288.32 KiB) Viewed 8025 times

[Hillmanjames]

I have attached my .msq to this post, The log is massive so i will try to get a smaller one showing what i'm up against.

There is also no idle valve used on the car. It uses advance to control idle speed (800rpm when warm) When warmed up and no idle control is on it will idle at 1100rpm.

hillmanjames TPS tune.msq

(57.89 KiB) Downloaded 266 times

Thanks, James

[jonbill]

I know you're question was a request for blended TPS and MAP with AE, but....
IMO, the VE table looks wonky. in scenario 1 (48% to 60%) you're going VE 52 to VE 56 at 60% which is the same value you have at 100%, and you're at peak torque it seems around 2500. whereas in scenario 1 you're going from VE 16 to 19 for a change which represents adding quite a lot of air - on my car, the same TPS transition goes from say VE 16 to say 30.
Are you sure the VE table is good? I think you can't tune AE until you have a good VE table.
(apologies if I'm teaching you to suck eggs).