Rotate86

Out of interest what dictates the limitation of mapping points? Be it Ve table or ignition table. Is it Ram or limitation of CPU or just software?
If it is software is there a reason more columns and rows eventually maybe able to be added to table?
Asking as I like the way adaptronic allow the user to add more where they see fit. With current 16 load scale I have to balance it between vacuum and boost. Would be nice to have a little higher resolution in both areas.
Thanks in advance

dznutzuk

iTrader: (1)

I don't know about other versions but the ms3 pro can do 16x32 or 32x16

muythaibxr

As far as adding larger tables, we don't really see the need. As dznutzuk says, you can use 2 tables if you are having trouble balancing things with 1.

Rotate86

ok so its not a hardware limitation then correct?
i like to have a few extra vacuum areas as it is a street car but means there are quite large jumps in the boost areas when running up into 20+ psi.
with the 2 tables it gets messy. especially as i have flex fuel sensor and already going to be running to tables as it is.i would end up with 4 tables correct?

Aaron Cake

iTrader: (1)

The 'Squirt has always interpolated very well. And as it turns out, at high load in boost, fuel is almost (ALMOST!) linear so you don't really need much resolution above 100KPA.

With the current hardware, tables are not really limited by hardware. I will be corrected if I'm wrong but there is still plenty of memory left.

Dual table is the way to get more resolution if you need it. Set up one table below atmospheric, one in boost. Only slightly more annoying to tune as it's two tables. But even with 20 PSI, I can't see the need. Personally, tuning up to about 20 PSI as well in a few installations, I've never run out of resolution. But every vehicle is different. If you have some crazy large port and ITBs...then maybe using MAP isn't the best idea outside of boost.

rx72c

iTrader: (8)

What do we think about this with cars that are running 40-50psi?

muythaibxr

Again, if you're running that much boost and you feel the need for bigger tables, you can run a second table and put an entire 16x16 table in boost, and entire one off boost.

Its not really a hardware limit so much as we have not seen a need for bigger tables, even on engines running 40+ PSI. Plenty of folks have tuned engines like that with 16x16 tables.

There is a current software architecture limitation in that we have flash tables in the current code that are 1K, that allows room for 2 tables, (or 1 bigger table) but would require us to completely rearrange all our config data to make larger tables fit (since we couldn't just make 1 table bigger, we'd have to make at least all our fuel and ignition tables bigger, and there are several of those). That is a huge task, so we have challenged people to provide data that shows that larger than 16x16 tables are even needed. That thread was open for 4 years on msextra.com with nothing but anecodotal evidence presented. We want hard numbers or logs showing that there was some problem that couldn't be tuned out with a 16x16 table.

As such, we feel that using 2 tables with switching is sufficient for those who feel they need more resolution.

Ken

Rotate86

thanks ken thats exactly what i was after info wise as to its a code thing not hardware limitation.
doing some further research on the subject i found a post from Andre Simon a highly regarderd tuner who runs High Performance Academy.this was his question and answer.thought it would benifit this thread.runs pretty closely to what ken has just said.

question:
Ok so on megasquirt ECU’s they are advertised and the manual only really talks about 16×16 fuel and ignition map, And thats all I thought it was capable of…

Until recently I found its actually east to get it to use a 32×32 Map but its very obscure in the manual and you actually have to use 2 individual maps with one map going from say 500-5000rpm and the second map is 5000-9000rpm.

So I queried why its such an obscure function, why its never advertised it has the ability and since the software/hardware can obviously handle it why isnt it a obvious option and all on one map rather then 2…



The response I got was that no engine ever really needs more then 16×16 maps… I said what about very high rpm engines or very high boost engines to which they said at higher rpm’s and using a VE method the fuel maps are very smooth and dont need that many zones. same with spark advance.

They put out a challange awhile ago for if anyone can show them conclusively a motor that indeed needs larger maps to be tuned correctly and cannot be tuned well with a 16×16 map but seems no one has?

But either way I still dont get that if you have the ability to more zones then why not give the option easier and actually advertise the fact, because needed or not it makes the specs of the ecu look better next to another ecu. if one has 16×16 and the other has 32×32



Answer:
I don’t necessarily disagree with Megasquirt’s assertion. My own technique as described in both our practical courses is to use maps with approximately 500 rpm and 20 kPa zoning. This will cover most of the engines we are likely to see adequately with a 16×16 map. Sure if you are running to 10,000 rpm or 4 bar of boost this may ‘seem’ like it would need more zones but this isn’t necessarily the case.

I ran a webinar a while back where I ran a fully tuned engine with a typical map (say 16×16 or thereabouts). I then reduced this to a 2×2 map and the results in both power and AFR control were almost identical. This really begs the question – Are all those extra zones actually making your tune any better, or just taking up more time? Of course nobody is going to tune an engine with a 2×2 map as we don’t get the fine control we need to control AFR and ignition, particularly and most importantly around idle and cruise. As rpm and boost rise though we can usually get by with wider spacing of the zones.

What we tend to see is that the VE of an engine will usually change smoothly, particularly at higher rpm. At low rpm, particularly if the engine is heavily cammed there will often be areas of large and erratic VE change, and this will require tighter zoning to get good control over the AFR. When the VE isn’t changing rapidly though, we don’t actually need to have a lot of zones as you will find that there is minimal change required in the fuelling for example between rpm sites. Remember also that the ECU interpolates between zones.

What this means is that if you are running an engine to 10,000 rpm, you will probably find that above 5000-6000 rpm the rpm zones can be spread to 1000 rpm. Likewise with high boost drag engines, we usually find we will get to a certain boost level after which there is minimal change in VE and hence the fuel map becomes flat. Remember that there is still a background fuel equation going on that alters the fuel injector pulsewidth based on manifold pressure so we are only accounting for changes in VE in the actual fuel map, not changes in manifold pressure.

So in short, yes a 16 x 16 map is probably adequate for most engines. I still find that there is a tendency for most tuners to add more sites than necessary and the result is more work for no actual improvement in performance or accuracy of the tune.

muythaibxr

Good to have what we have been saying for years and years confirmed!

We've never really been interested in the table size pissing contest, and this is why.

Ken

rx72c

iTrader: (8)

Ill be testing a ms3 out soon. see how it all pans out.
I don't really care if it is 16X16 or 32X32, if it works that is all that matters.
Haltech have a bucket load of features that just don't work. No point having them.

elturbonitroso

any update.

peejay

I have a bridge port and ITBs, and I could probably get by with a 4x4 table. N/A of course, but still, the engine is silly-linear. My bins are 2k apart at the top of the rev range, although it's pretty much a straight progression from 2k to 6k and from 6k to 10k so even that is really redundant.