Aaron Cake

iTrader: (1)

The Microtech after start is hard coded for 10 seconds. Unlike every other ECU where you can adjust the length and amount of fuel applied. So you need to compensate by adding a lot of fuel with the afterstart then taking that up with the water and air tables. Crude, and it results in a car running WAY rich on startup.

One you have set matrix mode, you are tuning on that table. No switching back and forth between the LOAD maps and matrix. The matrix table becomes your load map referenced by manifold pressure and RPM. Basically like every other ECU out there.

The correction tables like air and water still apply to the matrix table.

I've said it before, don't waste time using the carburetor mode. Switch to matrix. It will be way easier to deal with and is how every other ECU is tuned.

Leeroy_25

Thanks again Aarron.. Okay will switch to Matrix! I assumed it was just advanced version of the current mapping but from reading more I find that is not the case.. Are any of the load or idle numbers I have made and water temp correction valued going to populate over to matrix mode to give me a base interpolated map it is it complete start from scratch? If scratch as I feel it might be can I note my temp correction value's and load map settings and input them somewhere through the middle of the matrix to give me a base to start from?
The manual says there is an idle map and load map in Matrix mode. But reading other threads people have reported the idle map part doing nothing? Any advice to offer on that side please. Or any tips o the best way to get started from what I currently have would be great.
and there is also still cruise and WOT maps in matrix mode? Or is this just all part of the one big matrix table. Manual does not explain. It great. I guess it is one big table and in theory you only hit one cell for any give load/rpm combo no other compensating factors should be needed for any cell?
Thank you

Aaron Cake

iTrader: (1)

The manual has instructions how you can copy your current values to the matrix table but essentially you are starting mostly from scratch. That's one of the reasons I had mentioned you might want to switch to matrix as soon as possible. Water and air correction values apply to the matrix table just like they did to the other maps.

The matrix table idle map is only active if the engine is less than something like 900 RPM. So most rotary setups wont be using it and will just idle off of the main matrix.

I hate to keep saying it, but it's another limitation of the Microtech.

The matrix table replaces ALL the silly correction maps. No more cruise and WOT map.

Another confusing aspect is that the LOAD line on the matrix table is just the 4000 RPM column.

Leeroy_25

I read the bit in the manual that says to start by tuning the 4000rpm column and work out from there? But I did not see anything about copying existing values from the load map over to the matrix mode. I will have another look later.
Very strange the load line only shows the 4k column in the matrix mode? Is that where it will copy existing load maps to then? Just populate the 4K column? That doesn't sound like it would give you anything close to usable to start with? I would imagine values running diagonally up through the load and rpm? I could screen shot my idle map and load map and reckon I could probably manually enough to get it running at idle and maybe off idle but the I would need a dyno from there anyway to get some load on it.
The only reason I had been persisting is I didn't understand that matrix mode would be a from scratch deal and also I wanted to know that the running problems were just map related and nothing else silly. To be fair matrix mode actually sounds like it would be easier to tune just maybe a little more time consuming. If you only have to monitor one cell to get a correct fuel value for a given load it should be far simpler. Rather than have to go to a different table to remove fuel for 18"Hg at cruise. Then another to add some back in for WOT conditions?
So the only other things to tweak around after that would be pump settings for throttle response if I remember what I read the other day.

P.S! One thing that would be super handy with Microtech would be offline tuning! Even if just to look at stuff whilst talking about it and checking out info and requirements!

Aaron Cake

iTrader: (1)

I think, and I say "think" because the manual is wonderfully unclear, that when you switch from normal to matrix the normal LOAD map becomes the 4000 RPM row and the ECU reads those values if the rest of the matrix isn't populated. It has honestly been a long time since I made the switch because every Microtech I see gets immediately switched to matrix. I don't even try to fire the car the first time in normal mode.

There are instructions in the manual on page 25 under "All maps the same" to copy the load values to the matrix table. Never actually done it though.

And yes, the matrix table is a one stop table for the whole load curve as opposed to all sorts of correction tables. Every other ECU uses a 3D table to represent fueling, ignition, etc. It's far easier to deal with because you can see the cursor move around the table as the engine load changes. You can make your changes there, then go into the graphs to make sure the curve is still smooth and make minor changes to even things out.

Microtech users have been requesting the ability to save and view maps offline for 17 years. If you really want to weep as to what Microtech users are missing, download the free Megasquirt tuning software called "TunerStudio" and load up one of the example MS3 tunes.

BluRR

iTrader: (1)

Yes Aaron, if you activate the "All maps the same" function, it copies the values from your "load map" (4000 rpm) to all the other rpms in the matrix table. Just so u have a fully populated matrix map where you can add or subtract fuel for cells above and below the 4000 rpm line.

Leeroy_25

I had to run the car up today to do some other bits. So thought I would update my findings. Water map is miles out again having sorted the main idle map better! Need another go at that when I have time. Also I am not sure what and how much air temp correction I should be adding in as currently ambient air is 0 degs so I am sure that must have an impact. I don't want my water map ending up over rich for normal airtemps? Any input here would be great. Once all up to temp the previous values I ended up with for 18hg. 1.10 Ms and either side of this 1.14ms on the idle map seemed to work fine. Marginally rch for my liking but any meaner amd any load coming on will drop the voltage enough to lean it out a little too much and then it runs away with itself and dies. I check as I waited for the fans to kick in and this time watched voltage and Afr readings. It pulled down about 0.5v or so and this pulled about 0.5AFR or so. So my 11.3-11.5 or so was 11.8-12.0. At 12.5-12.8 it will start to stutter and die. Hence if I run my 18hg down at 1.05 which gives a lovely AFR around 12 from memory. Then when the fans come on that would pull too much out. So I will look onto rigging the file control up in due course and seeing if I can wire the fan on into the ECU aux input for idle up. I also noticed I have to apply throttle more than with the old injectors to get it to fire up? Is that cranking adjustment. I messed with AFt_St a bit and took it up to 12% didn't seem to help. I need to hold throttle for a period of time before it levels and I can let it idle on it's own. More AfT_St? Didn't want to get carried away!
I had my heater fans running all the time today so that may have been having a bit of an impact with an initial draw on voltage? Did seem to take the volt reading a few seconds to come up to 13.8v after starting? So certainly some playing to be done with corrections and idle control. Time to get the map into matrix mode and over to my tuner think though!

Happy New Year!

Leeroy_25

side thought..
Is there any way I could rig heater fans and engine fans to the aux in so if either was on the fuel mix is adjusted and idle up comes on? I cannot just connnect the wires other wise they would turn each other on?
Other options is could you rig something out of the alternator. So whatever wire the ECU reads voltage from so that if the voltage drops below say 13.5 volts there is a fuel mix added. I guess there is no way of scaling it for varying voltage drop giving varying fuel? I cannot remember what all the Aux inputs are but I recall it is just a switch on/off.
I am not electric guru but that doesn't seem to much of a task?
Looking how the volt drop effects AFR I would say there is no voltage compensation at all on the ECU despite what Microtech said!

Leeroy_25

Hi Guy's could really do with some input on the last couple of posts please to see if there is anything you can suggest? Mainly regarding, Cranking, Aft_st, Air Temp adjustment, and if there is anything I can do to rig some kind of switch up to trigger with voltage drop from the Alternator?

Thank you
Lee

rx3sum

Are your water and crank maps similar to the ones in the first post of this thread?
https://www.rx7club.com/microtech-11...e-maps-343297/
Again air temp corrections are not important at this stage.
Have you switched to matrix or still in normal mode?

Leeroy_25

Leeroy_25

So I have changed the alternator pulley for a stock one now and swapped a couple of hoses So ran it up to check for leaks and see what voltage looked like. Voltage is just over 14v at idle now. Check maps against the link you posted. Water map is in a similar region. Generally a little richer. Crank map the again similar but richer. That has not been touch since running 1600 cc injectors. Aft-St is 10 currently and I would say it ideally wants more but not sure how much to go?
air temp values were zero after 11degs and that was only 1, then 3 for 5degs and up.. As water warmed up it is clear the cold air temp is having and impact so I added in a couple of percent at 5degrees and then -1 degs up to 7%. This seemed to help. Not sure if I should add in some for 18degs as stock map suggests or not? When warm it idles at 900-950rpm pretty well until you get any voltage load. Interior fans or rad fans pull down the voltage and it leans and dies. This is the main thing to solve now. Will slightly higher idle help or adding in some timing somewhere? If I could rig something to detect voltage drop that would be great. And wire it to the ecu aux input. A right bummer it is so sensitive to the drop in volts. I noticed you get a bit of delay in the voltage dropping on the ECU compared to when the injectors start to react to it?
need to get over to matrix mode though as the TPS keeps messing around and catching out. When its cold the tps needs adjusting out so it does it drop the car into the load map instead which is richer. As soon as it warms up again the tis is fine but consequently then slow to react to application if throttle so you remain in idle map when applying real light throttle! All good fun this!

rx3sum

A lot of the maps are % based. So water,crank,afterstart % are added to your normal idle/ load maps. So if it runs fine at normal temp these % will remain similar for all rotary engines and are not effected by injector sizes.
So any changes you make to idle/load map may require these other settings to be fine tuned.
These old tps do tend to drift around a lot with temp.
I personally dont use the air control valve and just use warm idle of about 1100 rpm so that it runs well when cold with no throttle required.
Im no expert just a self taught enthusiast and have found the more you ask of the microtech the more issues you come across.
Hence im happy with higher idle and richer AFRs than ideal for my weekend toy that has no emission standards to pass.
Does your car have smog tests etc to pass?

Leeroy_25

Okay.. So if I cannot get any joy with getting the Idle valve working then I will up the idle a bit. as I recall when I got the engine in the donor car many moons ago the idle was set around 1200rpm.. So maybe that is the reason. but then the timing also turns out to have been way advanced static. so now that is sorted the idle naturally dropped right off and I had to turn it up a bit in fact. My hunch is it will be much happier with a bit of timing added at idle for the richer set-up and maybe bring the idle up around 1100rpm as you say. if everyone else is running without stalling issues and without Idle control valves then no reason I can't!
I just like to get things the best I can! I am going to research and find out if I can find some kind of voltage drop switch I can wire in to make the idle up function or at least fuel add function work off the aux input.
My tuner will be much better equipped to know if my Cranking and AFT-St need adjusting up more. I think Aft-St does but I would sooner just stop messing with it now and leave it to someone to sort properly. I am going round in circles a bit and giving myself carbon monoxide poisoning slowly!
I will live with idle as rich as it is (low 11's) but would love to get it to run high 11's-low 12's if I can.
Do you find your idle is on the verge of hunting rich or not that bad? Mine is during warm-up and a bit less so when warmed up.

I have to get an MOT which needs some kind of emissions test but I think due to it being an engine conversion and year of engine I can get by on an old school test pre-Cat test. Not sure it will get through on that or not still?! If not I will fudge something either map or exhaust wise to get it through!

Leeroy_25

Was speaking to my tuner today.. Sounds like he wasn't planning to tune it in matrix mode but is going to now that I asked! Also agrees after a bit of a chat that it should help with idle issues. He believed he could get a decent tune on normal mode but then as we talked more he came around! So I am just trying to get a date to plan that in and maybe get him over to do some basic set-up before hand to make sure it all runs okay in matrix mode to start with. He had some good thoughts on helping idle control. He suggest a couple of options. locking off the vacuum to the fuel pressure reg and normal idle so that it kept fuel delivery more stable. Do this with solenoids and one way valves. So I will see what the pressure fluctuates like and just try it with a clamp on the vac hose maybe.
Another option was have a stock ECU along side the microtech to control the idle valve? not sure if anyone has done this?
I am also still trying to looking into a switch that will trip off the voltage coming from the alternator and switch a source to the ECU Aux input if it drops below a certain level so you can then use the idle up functions as if you had connected say the cooling fans to the Aux in.
So some thoughts there and will look at it in time. Mainly I want to get it on a dyno now and get the main map done in matrix mode with warm idle sorted so I can pull the engine and rebuild it. I will mess with the cold idle once it's all back together and try some different things.

Cheers for the help. Any thoughts or input on the above I would love to hear.

Thanks
Lee

Aaron Cake

iTrader: (1)

My personal opinion is that if your tuner wasn't going to tune in matrix mode, he better have a damn good reason why, because otherwise he would be satisfied with turning you loose with a massively inferior tune.

I don't know much about your emission laws on the other side of the pond but the chances of passing most sniffer tests with a Microtech are about zero. Its fuel accuracy is far inferior to the stock ECU. The best you can hope for is making one map to pass the tests, tuned excessively lean but just before the NOX increases in the areas the test measures. Then swap to your regular map afterwards because that emissions map will be very unpleasant to drive.

You ABSOLUTELY DO NOT WANT to lock fuel pressure. Your injector times will have to increase widely under boost. Plus it will be much harder to tune your idle and light load.

I know, I'm starting to sound like a broken record, but step back and look...You are thinking about hacking fuel pressure and hacking other devices onto the Microtech to control idle...even to the point of wiring in a 2nd ECU! Replace the Microtech with something that doesn't suck.

Get a Megasquirt MS3 Pro or a Haltech P1000. Make a sub harness to connect it to the existing Microtech harness. That way you can salvage most of the wiring, just adding the few extras when necessary.

j9fd3s

iTrader: (3)

i'm not sure what your limits are, but here is roughly what the engine will do with various equipment.

stock EFI + air pump + cat = 50ppm HC, 0 Co, and Nox is listed as .21, not sure of what.

take the cat off of the stock setup, and you'll be, 2,000ppm HC, 6% Co, we didn't measure nox.

tune the ecu, and with the air pump and no cat, and you can be down in the 200 range for HC's, and probably around 1% CO.

take the airpump off and you'll be back in the 6% co and 2000HC range no matter what.

the engine will either lean misfire and have high HC's, or it'll be rich enough not to misfire and have high HC's, its a fundamental design thing. the air pump is added so that the engine can be run and not misfire, and the cat can still see a "normal" AFR

Leeroy_25

Points taken.. Just to clarify the fuel pressure lock idea was only going to be for idle with the thought being it would help keep it stable with fluctuations. But it was just a thinking aloud thing. Not going to do anything with that. I agree if I have to start doing things like that then it is a case of bin the ECU off for something else. I will persist with what I have until I cannot live with it any longer. Loads of people out there run microtech. I just will have to live with some compromise or if I can't get a new ECU! After it has had a proper tune it will be a more fair assessment anyway. I will up the idle rpm a touch too.

Aaron Cake

iTrader: (1)

I promise I'll only say one more thing about swapping ECUs:

Yes, lots of people use Microtech. I have on in my car (until it is swapped). I'm not saying it doesn't work. In fact it works very well within its limitations. For a track car that doesn't need data logging and just basic EFI control it is highly reliable, robust, and somewhat easy to tune.

But it is also so basic, backward, and lacking features most deem necessary for a street vehicle that it is essentially like running an early computer controlled carburetor. OK, it's a little better than that, but only marginally. The ignition method forces you to accept the limitations of a distributer because it so brilliantly emulates one. The fuel table is fixed. And there is no closed loop, idle control (that doesn't suck), aux inputs/outputs, 3D ignition table, engine protection besides boost cut, A/C control, data logging that doesn't suck, save to disk, autotune, configurable tuning software, configurable fuel table bins, and I could go on and on and on. It would have been a revolutionary ECU in 1994.

All I'm saying is, why waste all that time tuning the Microtech when inevitably you're going to have to make a load of compromises to end up with just an acceptable tune. When that tuning time (dyno and tuner time is EXPENSIVE) could be put into a much better ECU. Sell the Microtech. Someone will pay decent money for it.

Just do yourself a favour and check out the MegaSquirt's TunerStudio software. And compare it to the recompiled Windows 3.1 application which Microtech uses:

https://www.rx7club.com/attachment.p...1&d=1485619749

OK, I'm done.

Leeroy_25

Okay.. So that does look a lot nicer.. Just on the face of a screenshot I can see some useful functions there! Like it has injector settings!!And was that an AFR table in the back.. So I assume it can self tune if you plug into a AFR gauge of some sort?
I have heard of megasquirt before.. I just thought they were some cheap nasty ecu not to be touched? Maybe I need to look into it more.. even if not for this car the next as it will need a custom ECU of sorts! Haltech is the name you hear of most and I know they don't come cheap.. is megasquirt pony by comparison?
I thought it would have gone like. Haltech (King), Microtech (queen), Megasquirt (Joker) but maybe I have my cards all backwards? or is it just the megasquirt has done a better job on some basic function where microtech has failed and is not so good elsewhere?

Aaron Cake

iTrader: (1)

Haha, got your attention with that one, eh?

Yes, there injector settings. Off the top of my head, including but not limited to:
-injector dead time compensation
-individual injector trim
-individual injector battery voltage compensation
-non linear compensation for small pulsewidths (really useful for big primaries)

One of the purposes of the AFR table is indeed to allow the ECU to autotune the fuel curve. Set the ratios you want on the AFR table, turn on the autotune feature and drive around while the ECU takes care of the tedious work. I use autotune a LOT to do all the initial work and get the table to the AFRs I want. And then I can spend my valuable time fine tuning that table and adjusting it for best response, drivability, mileage, etc. Also have used autotune in boost. Generally by choosing a very safe ratio (ie. 10.5:1) and then once autotune has produced a nice smooth "load" table, just scale it to the appropriate AFR. TunerStudio allows you to highlight any section of the table and apply transformations (increase, decrease, multiple, formulas).

And under closed loop operation, the AFR table can be used to set the ratios the ECU will achieve via feedback from the wideband. Yes, of course this is all possible because the ECU can actually read and understand the wideband output! You can even run the ECU fully in "observer" mode so that the AFR table applies to the entire non-load section of the fuel table, always compensating and setting the AFRs you have specified. I like using this as a tuning aid. By opening the AFR correction gauge I can make changes to the table until the corrections are 0%. Also a useful trick when tuning Haltechs, by the way, because they don't have as robust an autotune feature.

A datalog can also be taken and used to autotune the VE table offline via log analysis.

The Megasquirt certainly does have a very mixed reputation. And somewhat deserved. The early MS1 units with the 2.2 board required extensive hardware mods and often firmware changes to be usable in all but the most basic applications. The V3 board improved on that somewhat while the MSII processor added a lot of needed functionality. Problem is that the built it yourself nature of the ECU meant that a LOT of people who should NOT have building and modifying ECUs were doing so. With typical disastrous results. I have seen Megasquirt installs that made my skin crawl, all the while the user blaming it on "This POS Megasquirt". Meanwhile it looked like "my first soldering project", poorly built DIY mod circuits hanging everywhere, wiring made of lamp cord, etc. The MS3 is currently the latest incarnation still available as DIY, or fully assembled. But the version I'd recommend for the non-electronics user is the MS3-Pro. It's a fully assembled MS3 with better hardware, sealed case, and designed to be the equivalent o f buying something like a Haltech.

Actually, I realised I am gassing on when I've already done so. Take a look at my Megasquirt tutorial. While it is geared towards the FC with a 13B, the principle is the same:
How To Megasquirt Your 2nd Gen RX-7

Leeroy_25

Thanks Aaron, I will check that out..
So bit of an update while I am sticking with the Microtech for now.. So quite uncanny you should reply today!
My tuning guy came over to drop the ECU into matrix mode and check it was all okay and do an initial set-up in the garage before we go to dyno in a couple of weeks. First thing I will say is why did you not tell me about matrix mode sooner!! I joke! Chalk and cheese.
To be fair. After my last setting up with the water map I let it run up in carb mode and it was okay.. struggling a bit.. then behaving a little strange as it warmed up. I think the idle valve was actually trying to operate in someway where it is connected but not in a helpful way. So I decided to up the idle to just over 1000rpm about 1050rpm hot where it was previously about 900-950rpm. Well that in itself made a huge difference. right away I noticed I could put on an electrical load or all off them and the voltage did not move.

So, then we flicked over to matrix mode and did some free revs up to around 2500rpm and set-up a small area of the table. Seems to be able to pull lower values in the table although it seems to read slightly richer at the gauge? I dare say we could get a bit more out. I think the idle cells are now down to 1.00 and it is happy and smooth the best it has ever been! I think raising the idle just that small bit has made a world of difference. but matrix mode is soooooo much better and for me a lot easier to get your head around.. No other compensation tables to worry about.
So after a while of running and happy with it I let it warm right up so the engine fan would kick in. The engine fans made a slightly dent on the volt reading for a second but it corrected itself and carried on running fine. I put cabin fan on at the same time as well and it was happy as Larry! So if it stay that way then great.
Idle AFR is around 10.8-11.0 with the 1.00 in the map. Might try and take it down a little more but I suspect it will start running into problems. I had AFR more around 11-11.3 before okay.. so maybe it will go a bit lower but that is the least of my worries now! I think it is the restarting that becomes a problem if you pull the value lower.
On that point too.. it turns out AFT-ST was too high and that is why it did not like it. in matrix mode you could clearly key on and it would run rough as hell for 10secs then die! tried adding more AFT-st and it made it worse. then pulled it down to +2% and bang started right up and pulled itself into shape straight away.
So a bit more playing about to do and see if I can lean it off but much, much happier and now just regretting wasting a few weekends dicking with carb mode!

One other small point to mention.. Voltage reading when we started was 14.2 as it warmed up this seemed to gradually drop to 13.8-13.9v? Didn't seem to change AFR or running at all strangely? Is that normal? Same RPM and even a quick rev didn't seem to put it back up to 14.2?

I am definitely onto Megasquirt for the next build I think though unless a Haltech comes up cheap! Too many limitations that shouldn't be there.

Aaron Cake

iTrader: (1)

Like I always say, don't even bother trying to tune without first enabling matrix mode.

I suspect your voltage is higher after start because the battery voltage is lower, causing the alternator to output a higher voltage to charge it back up quickly. Then the output of the alternator and battery average. 13.9V is about right for a running vehicle. Should be a bit higher, but SO MANY factors can effect the reading I wouldn't worry about it at all.

Leeroy_25

Maybe.. battery had just come off charge and was 12.75v before I initially started the car.. I thought they usually sit around 14.1 or so and I am sure it read that the entire time I was running the car the last time I ran it up. Could having it in matrix mode have made a difference? A bit strange for sure.

Aaron Cake

iTrader: (1)

Some alternators don't put out full voltage below about 1500RPM. Most Mazda alternators seem to be like this.