Mappable ECU's - The definitive thread

[Redracn]

It is possible to place the various main load sensors into two categories.

Those that must be tuned to the engine and can not detect and correct for certain changes that may be made to the engine such as a set of headers. They can not even compensate for engine wear. In fact long distance enduracing racing has compensations applied over the length of the race. (Closed loop control or self learning can of course work wonders at least until the Lambda sensor looses calibration). 

MAP and TP are in this category. MAP can also be used along with air temperature to calculate air density but this is not a flow.

Those that can detect and compensate for improved or reduced engine breathing tend to measure flow either volume or mass. In the case of volume additional compensations can be applied to calculate mass flow.

MAF, AFM (Velocity) (of which there are many variations) are in this category

The flow measuring category by their very nature tend to be more restrictive than the other methods. There have been many attempts with varying degrees of sucess to reduce the restrictiveness of these sensors.  

 

 

[bumble]

 

The comments about Alpha-N are at best misleading and at worst, simply wrong. I would love to know what site it is on. I am going to guess it is in some way connected to a particular engine management system, because it seems to talk about limitations that not all have.

http://support.moates.net/category/all-articles/

There are several articles headed Theory.  IMO they were referring to 'pure' Alpha-N as a primary technique with no other compensating strategy.  In that case they may have a point.

[Redracn]

 

They have no point in relation to Alpha N. In fact about 50% of the tech stuff on that site is wrong or misleading.  To drive up a hill (not a mountain) requires no other compensation. If you move your foot to maintain speed more air will flow into the engine and get more fuel to go with it (because you moved your foot to a new position in the look up table). Remember every possible combination of throttle and rpm has been tuned (or should have been). If you keep your foot in the same spot you will slow down. This results in a new rpm position in the tuning table being looked up which will provide more fuel per injection as the slower spining engine is now under increased load due to each cylinder filling better. For a constant throttle if you connected a MAP sensor the value would go up (less vacuum) as the RPM goes down. (Ignoring CAM shaft)

I also do not know of anyone that has fitted Alpha N and not also fitted a MAP sensor and connected it to the plenum or left it open to the atmosphere to act as a Barometric sensor. With this done even mountains are not an issue. 

I would stay well away from anything these guys do as they simply do not know their subject.

[SimonN]

http://support.moates.net/category/all-articles/

There are several articles headed Theory.  IMO they were referring to 'pure' Alpha-N as a primary technique with no other compensating strategy.  In that case they may have a point.

Having looked at the site, which is actually quite interesting because they are one of the main suppliers of emulators, I now understand their comments. They are right up to a point, but the way they word it is misleading.

 All their articles are really written from the point of view of hacking existing ECU's with an emulator. as such, they are correct that people doing this are unlikely to come across Alpha-N because it is rarely used as the primary strategy on mass production, road going cars. I don't agree with some of their other comments about Alpha-N, mainly because of poor wording and not relating them back to to their core market. If they had simply said that it was going to be a short article because Alpha-N rarely used in mass produced cars and their customers probably won't come across it, then all would have been fine.

But some of what they say is simply wrong. The "hills" comment is the funniest. Consider how many rally cars are mapped with Alpha-N (most or all?). Do they go up hills or are the ones that are mapped that way at a huge disadvantage? The comment that implies it is only good for WOT is also laughable, although they would have a point that it isn't so good for handling idle, which is where other additional strategies come in.

[Redracn]

They are also wrong in relation to speed density and in particular what VE (Volumetric efficiency) is.

Many cars have done Pikes Peak with alpha N plus a BAP sensor. The air pressure at the top is nearly half that of sea level. Alpha N handles idle fine when used with an electronic throttle as all air that enters the engine goes through the throttle. 

[SimonN]

So I have been thinking about what next on this thread and the obvious thing would be to compare the ECU's that are on the market. After all, although my experience is with Motec and it would take a bit to move me away from what i know and the support network I have, I know that there are alternatives that will suit others and maybe I am being a little to blind. So i thought it would be simple. Just type into Google "compare ECU's" and away we go with the research. A dozen search phrases later, and I don't have a lot to report. There are a number of articles comparing plug and play stuff for certain US muscle cars, but I haven't found a resource that helps us.

I am a bit time poor at the moment, but will be doing some more research as time goes on. In the mean time, has anybody else found a good resource that compares what is on the market which might be suitable for our cars?

[clutch-monkey]

Whatever ECU the tuner is comfortable with using. Haltech is good. I have an autronic on my 3.6. Goes well. Our engines aren't demanding a lot from them.

[SimonN]

Some on here, including me, intend to do our own mapping. While I prefer to use an ECU I know, Motec is expensive so if it was worthwhile and i can get the results and support, i would consider something else. So what is good and how do you come to the conclusion that a particular ECU is good vs bad, or maybe that should be less good.

I think the differences come down to a number of things

Hardware - This includes processor speed, connections, number of sensor inputs and number of outputs, availability of sensors, harnesses etc., types of injectors managed, ignition etc. ease of installation

Software - Ease of use, variety of mapping methodologies, structure and hierarchy of look up tables, available starting maps.

Support - Manufacturer support, documentation, training (online and courses), forums and other user communities

Porsche specifics - are there any Porsche specific resources or parts

So this is a quick list post Melbourne Cup drinks! What have I missed? Maybe it's possible to draw up a definitive list we can then research against and by filling in a spreadsheet, people can draw their own conclusions.

[Redracn]

Hi Simon

Seems you want to choose a horse withou taking into account what sort of cart you need it to pull. I would start by specifying your actual job. This should include what it needs to do now and what you have planned for the foreseeable future. In my experience once installed and tuned most people just drive and enjoy. If you want to go down the development route and add sensors that are not used in the ECU then a dedicated logging box is better. Having a couple spare inputs on the ECU is good but why pay for a lot that will never be used. Also rating how important you consider each item. Someone starting out would consider tech support high on the list while a pro would have it lower and value reliability and bug free software more highly. Processor speed means nothing as MHz is not an indicator of actual capability to get the job done no if's no but's and getting the job done is what matters. Items like power dissipation and max operating temperature are important as they affect the mounting position and reliability. These days why pay more for an ECU with expensive low impedance injector drivers. You can get hugh flows in fast high impedance injectors so why use old tech low impedance ones.

[symsy]

So guys what is most pervasive .. here in Aus and what is the depth of local experience? I posted something in another thread and do want to pull you too far sideways ..but for example with my RSR 993 1996 3.8 super cup spec motor (Gonna use in a Hotrod Roadcar) .. to be mated with Jenvey Or PMO unless someone tells me other wise… whats the most used down under config... after all its accessibility to good techs alongside good gear is it not that gets best results?

From other experiences, I find a lot of Non Pcar guys love Haltech or is there a particular mating of ECU to Porsche motors that seems to come up trumps here or worldwide.

PS thanks for the Brain melting thread

[bumble]

Just catching up on this thread.

Megasquirt used to be good when it was a kit as you could build it for about $300. Now it's a full deal and costs about the same as an entry level Haltech.

You might want to look at this then...  https://www.diyautotune.com/shop/megasquirt-kits-components/

 

[SimonN]

symsy

There is zero doubt in my mind that you need to speak to Colin Belton at Ninemeister in the UK. He has fitted Motec to a large number of cars with similar spec engines, including the high powered specials he used to do for Singer. He will sell you a kit of everything you need including an electrical harness and a base map, and because Motec is well known, you shouldn't have trouble finding somebody to finesse the map. IIRC, he says that to fit the Motec is something like 3 hours work (don't hold me to that!). 

If you want to map yourself, there are other options, but I suspect you don't want to. 

Other than that, the one thing I am sure of is that with carbs, you are leaving a fair amount of power on the table, so even if you don't go down the 9M route, you should see if there are any local tuners who can fit a mappable ECU. To some extent, it doesn't matter which one, so long as they are experienced and good with it. I have my preferences, so I would go to PR Technology, but then again, they use Motec!. they do have their own rolling road and it is in use almost all day, every work day as they map engies ranging from full race Time Attack to normal race and rally and also road/track cars.

Considering that you could consider Australia home to the mappable/aftermarket ECU with so many manufacturers here, you should be able to find something somewhere.

[bumble]

So guys what is most pervasive .. here in Aus and what is the depth of local experience? I posted something in another thread and do want to pull you too far sideways ..but for example with my RSR 993 1996 3.8 super cup spec motor (Gonna use in a Hotrod Roadcar) .. to be mated with Jenvey Or PMO unless someone tells me other wise… whats the most used down under config... after all its accessibility to good techs alongside good gear is it not that gets best results?

From other experiences, I find a lot of Non Pcar guys love Haltech or is there a particular mating of ECU to Porsche motors that seems to come up trumps here or worldwide.

PS thanks for the Brain melting thread

Motec...

 

[MFX]

There is zero doubt in my mind that you need to speak to Colin Belton at Ninemeister in the UK. He has fitted Motec to a large number of cars with similar spec engines, including the high powered specials he used to do for Singer. He will sell you a kit of everything you need including an electrical harness and a base map, and because Motec is well known, you shouldn't have trouble finding somebody to finesse the map. IIRC, he says that to fit the Motec is something like 3 hours work (don't hold me to that!). 

Do you have an idea of the ballpark cost of that kit? Then people can start getting some costings together. 

[clutch-monkey]

So guys what is most pervasive .. here in Aus and what is the depth of local experience? I posted something in another thread and do want to pull you too far sideways ..but for example with my RSR 993 1996 3.8 super cup spec motor (Gonna use in a Hotrod Roadcar) .. to be mated with Jenvey Or PMO unless someone tells me other wise… whats the most used down under config... after all its accessibility to good techs alongside good gear is it not that gets best results?

From other experiences, I find a lot of Non Pcar guys love Haltech or is there a particular mating of ECU to Porsche motors that seems to come up trumps here or worldwide.

PS thanks for the Brain melting thread

I was going to go a haltec but tuner more was more comfortable with autronic. It will work just fine. Uncle on here used haltech in both his 911's I believe.

my ITB setup was almost 8k all in and done and tuned on my 3.6. PMO's would have been 6k for the same peak power (I guess I could go an even more expensive shower injection with the efi to gain a an edge again)

[SimonN]

Just catching up on this thread.

You might want to look at this then...  https://www.diyautotune.com/shop/megasquirt-kits-components/

 

People think  am a Megasquirt hater and this might confirm their feelings are right, but all I can do is repeat what I read and am told.

First the good news. A well set up Megasquirt system works well, it is cheap and it is fairly easy to learn and there is good support. So why does it have a bad name among professionals?

The single biggest reason is that more than any other system, people run into trouble, do shit installations and the pros hate being asked to sort it out. That is not really megasquirts fault, but the nature of a system aimed at teh DIy market that gets all the boy racers with zero skill trying their hand. However, part of it is because everything is very generic with a bit of a "one size fits all" policy, so to do a good install still needs the knowledge and the skill.

There are 2 big gripes that come up time and again. It seems that many think the interface software is very average, being more complex and less intuitive than it should be. This is made worse by the haphazard way that the product has been developed. There are menus all over the place with no logic. This is also reflected in the firmware updates (2nd gripe) and I have read a lot about problems in this area. It seems that they aren't very good at version control and backwards compatibility so it is possible to end up with an inability to access an already programmed ECU in future years. 

I believe that some of these issues have been improved with the latest MS3.

I suspect that if you are really going it alone and won't be going to a pro to help you out, Megasquirt is well worth a look at. You need to make sure you are 100% confident and that you do a really solid job on installation, because one of the biggest issues reported by pros is that the DIY nature of the product leads to distinctly DIY installs. If you have a tuner who is happy to help you out of you get stuck with MS, great. For me, I don't know anybody in Sydney in that respect, but for other ECU's, I certainly do. I hope that is seen as being as balanced as it can be. 

[symsy]

Will share my info and budgetary quotes for my enquiry to Ninemeister..if its helps others when they reply 

[SimonN]

Do you have an idea of the ballpark cost of that kit? Then people can start getting some costings together. 

I will go back and check, but I think that on a 964, all the sensors, specially made harness, COP, intake and MAP, TPS etc and a pretty good start map, it will cost something like GBP3,000 of just under AU$5000. I could be a little out of date on this. I believe that in the UK, for supply, fit and map he charges GBP4400. That also comes with a power guarantee - if he cannot produce the shape of curve and peak power he claims, he will put the car back to standard and refund the money - assuming the engine is sound in the first place. Colin is that confident in what he produces. I remember back in 2005/6 lots of people on forums  saying his claims on standard 964RS's was simply BS - 320-25bhp with curve to match- and that is when he first started making the money back offer. I haven't heard of anybody taking him up on it, and I have been in cars mapped by him with peak power of between 320 and 400 bhp. The headline peak power doesn't do justice to how the car drives. 

I personally have no affiliation with Colin and haven't even been a customer, although I know people who have. There is a reason why people like me, with zero connection, think so highly of him. He is a true Porsche enthusiast who has spent a fortune on this sort of thing. He recently claimed that his investment in Motec development cost more than his 993RS

[Redracn]

I was going to go a haltec but tuner more was more comfortable with autronic. It will work just fine. Uncle on here used haltech in both his 911's I believe.

Some of the ECUs are getting a bit long in the tooth and while they may get the job done locating a second hand laptop that can run the tuning software is a very real problem.

[Redracn]

 This is also reflected in the firmware updates (2nd gripe) and I have read a lot about problems in this area. It seems that they aren't very good at version control and backwards compatibility so it is possible to end up with an inability to access an already programmed ECU in future years. 

 

What is most important is forward upgradeability I.e you should just be able to install new firmware/software as an upgrade and make no adjustments and everything continues to work as it should with any bugs fixed and new features available if needed. Some manufactures are better at this than others. Motec M800 and earlier ecu did this very well. But with the M1 series this was thrown out the door and major user/installer intervention can be required to do a version upgrade. Worst case is if an upgrade is done and the tuner is required to make changes that are not done or done correctly things can turn very bad very quickly. But with most ecu not getting any upgrades for the second half of their production cycle they become religated to only being suitable for engines of the era they were developed in. Ie. A 20 year old ecu will do a very good job at fuel and ignition on an older V8 with a distributor. But is no good on an engine with continuously variable cams or electronic throttle. These are pretty much pre coil on plug, pre simple variable cam, pre electronic throttle and continuously variable cams.

Given the long production life of many ECU like Motec M4 around 18 years and the M800 around 15 years and counting you need to be confidant that tuning software will run on new computers and that current tuning software will run on older computers otherwise you could be forking out for a new laptop every few years. Not an issue for a shop as the laptops never last very long unless you actually need an old laptop to get the job done. No winners here.

 

[clutch-monkey]

Some of the ECUs are getting a bit long in the tooth and while they may get the job done locating a second hand laptop that can run the tuning software is a very real problem.

agreed, i pulled a wolf ecu out of an old triumph.. got me buggered who would be able to play with that now.

Haltech are a pretty good bet though these days, they are always doing new things it seems.

[smit2100]

I tend to agree, that MoTec is a fantastic aftermarket computer system, but I think ultimately we should talk about bang for buck. I have no idea what  MoTec costs to set up at the moment (but I remember when I had my Skyline that is was very expensive), and you also need to factor in how long you are going to spend on the dyno getting everything perfect. 

I think it is worth working to a $ per hp. As it always comes down to a budget.

Bringing up  $per hp  is a great point.  I have no clue what it costs to mod these air-cooled examples.  Can anyone throw some indicative numbers on what a tune and exhaust mods and will get you in terms of extra hp and the associated cost with all work done in a workshop.  Eg I'm limited to having bosch ecu's messed with on watercooled 2.7l and 3.6l twin boosted engines.   I would like to compare $ per hp against my baselines on a 3.6 water-cooled example (boosted) eg First 80chp is running at circa $75/chp and then a 3 to 4 multiplier on that rate to get the next 120chp.  Also any issue with stock clutches once you messed with getting some extra power / torque.  PS That 3 to 4 multiplier allows for a bitsa clutch setup, eg bitsa aircooled 964 parts like a lwfw  and ring gear mixed in with bits of 997gt3rs clutch disc and shit bits like 944 throwout bearings. 

[Redracn]

  Also any issue with stock clutches once you messed with getting some extra power / torque.  

There are always issues and this is where a specialist workshop is worth every cent as they save in the long run by getting it right the first time with both advice and implementation.

[SimonN]

Clutches and flywheels are an often overlooked "tuning" opportunity, besides the need to handle the extra power. In the 964/993's there is a pretty well proven path, using the RS light set up which is also 20% uprated. That is usually good to about 330bhp and because of the weight, it noticeably improves throttle response. The downside is that idle can be problematic, even in a stock car with stock ecu mapping. One of the biggest reasons for getting the chip remapped on the RS was to sort this problem, which can be done with Motronics, but not to the level of perfect idling you see with a standard 964. The best idling with lightweight flywheel I have come across is with aftermarket ECU's, and I am thinking of 2 particular cars, one with Autronics and one with Motec.

[smit2100]

Clutches and flywheels are an often overlooked "tuning" opportunity, besides the need to handle the extra power. In the 964/993's there is a pretty well proven path, using the RS light set up which is also 20% uprated. That is usually good to about 330bhp and because of the weight, it noticeably improves throttle response. The downside is that idle can be problematic, even in a stock car with stock ecu mapping. One of the biggest reasons for getting the chip remapped on the RS was to sort this problem, which can be done with Motronics, but not to the level of perfect idling you see with a standard 964. The best idling with lightweight flywheel I have come across is with aftermarket ECU's, and I am thinking of 2 particular cars, one with Autronics and one with Motec.

There are always issues and this is where a specialist workshop is worth every cent as they save in the long run by getting it right the first time with both advice and implementation.

Yes that's why I am going with a bitsa and always like to go with the extra margin clutch setup as dropping drive trains aint cheap.  I am a fan of the  lighter rotational mass of the LWFW, gives the butt dyno feel you have another 20 plus hp and the good tuners will allow for LWFW.   Also be interesting to see how much difference in pedal effort with the LWFW over stock as not a fan of the feather light clutch pedal I currently have.