Potential alternative to Emerald ECU...

[ro_butler]

...the Link G4.
I have a spare ECU should the worst happen and I managed to 'unscramble' the one that came with my car (thanks Jools!) so I am not actively in the market but I came across this and it sounded interesting.

I mailed the company and they say that it will cope with a flat plane V8, which is usually the main problem. Also it looks to have plenty of interesting features such as the potential for 'drive by wire' (get rid of throttle pots), rev limits until engine is up to temp, valet mode, full sequential injection etc...

Someone has fitted one to his supercharged Tuscan (albeit a SP6), replacing his Emerald so it should be possible to fit one to a cerb, although by no means easy.
http://pistonheads.co.uk/gassing/topic.asp?h=0&amp...

Website and specs here: http://linkecu.com

Would probably help with emissions for the all important MOT as well. If my ECU's break I would certainly consider this.

[GreenV8S]

ro_butler said:

'drive by wire' (get rid of throttle pots)

How does that get rid of throttle pots? Surely you still need at least one throttle pot (at the pedal), and you've replaced a simple reliable bit of wire with a motor and associated control system (maybe even with its own throttle pot if it uses positive feedback). TBW seems like a good idea if you're programming the computer to stop the driver making mistakes, but it seems like a very expensive and complicated replacement for a simple cable otherwise.

[Redmist336]

I'm running a G4 Plus in my 350z twin turbo powered offroader. It's a fantastic low cost aftermarket ECU with high market features. However, I think you're ignoring two of the biggest costs with aftermarket ECU's, the wiring and tuning.
Drive by wire actually adds considerable complexity. I've removed it from my 350z engine simply because I believe that I can out "think" the computer in racing situations. For a road car I could see use in "valet mode" "warmup" or for gradual opening of the throttles (which actually gives you more torque through the curve).

[Mars]

ro_butler said:

I mailed the company and they say that it will cope with a flat plane V8, which is usually the main problem.

I thought the issue was the angle of th AJP V rather than it having a flat plane crank.

[ro_butler]

GreenV8S said:

ro_butler said:

'drive by wire' (get rid of throttle pots)

How does that get rid of throttle pots? Surely you still need at least one throttle pot (at the pedal), and you've replaced a simple reliable bit of wire with a motor and associated control system (maybe even with its own throttle pot if it uses positive feedback). TBW seems like a good idea if you're programming the computer to stop the driver making mistakes, but it seems like a very expensive and complicated replacement for a simple cable otherwise.

All good points. However my thinking was: Replace the 2 pots with a potentially 'better' one attached to the pedal. It might survive longer in the less harsh environment. Also the cable couldn't jam open or snap (although you could then be hostage to electrical gremlins). My point was not that drive by wire is better than a cable per-se but it might be better in the long run than a 'TVR' throttle cable

[ro_butler]

Redmist336 said:

However, I think you're ignoring two of the biggest costs with aftermarket ECU's, the wiring and tuning.

My current MBE is remapped anyway so if I needed to replace it I would have it done again, it really does improve the car. I don't have any copies
of the map since the so called 'professionals' who mapped it can't find it and anyway they put a 'scrambler' chip on the ECU and can't remember how
that works either

You are right though, wiring would be a pain in the @rse!

[ro_butler]

Mars said:

ro_butler said:

I mailed the company and they say that it will cope with a flat plane V8, which is usually the main problem.

I thought the issue was the angle of th AJP V rather than it having a flat plane crank.

Given that it allows full sequential injector and spark operation I thought the angle would not be an issue, but the firing order might pose a problem. Could be wrong though.....

I am just hoping that if my ECU breaks there are more than 2 options open to me (try to find a 2nd hand MBE [currently expensive] or Emerald).

[jesfirth]

MBE's are availble from the breakers at reasonable cost - a few people also have spares - i did until i sold it to somone on here who was desperate for one in a hurry and TT Tim has one - its in my garage!

memo to self - must give it back -- Tim are you at Castle Coombe?

[450Nick]

Emerald are good and I have one myself, but there are plenty more options than Emerald and MBE! Have you looked at Omex? DTA? GEMS? Megasquirt? Autronics (sp)? Motec? etc? They all come in with varying features for varying costs. Not to mention useability, try out the mapping software as you're more than likely going to want to tinker with it at some point, and a nice user interface like the Emerald (I've not seen better), makes it a pleasure to do so.

For a road car, full sequential isn't really necessary and you won't really notice any difference over wasted spark. EGT is useful but only if you're planning on running things close to the limit and things like dual lambda are useful for closed loop adaptive fuelling, though a single lambda is fine for the road.

Half the battle though is finding a good mapper to do it all. I've had mine done by several and they all seem to give the car a slightly different feel due to their style of mapping. Also it seems that any mapper will map the power curve, then a bit of cruise, and then move on to the next car. You pretty much have to tie them down and force them to map things like cold/hot start, cold fueling, low speed around town running, air temp, coolant temp, pressure, lambda, voltage, egt corrections, stepper motor etc. In a production car, a mapper will literally spend weeks and weeks perfecting a map to make it work well in all areas. Most mappers on aftermarket ECUs will take between 3 and 6 hours.

This means that much of the bad press of any ECU is not the ECU but its setup, which can completely ruin your perception of the car and thus, the ECU. A well setup Emerald, Megasquirt or similar should not really feel any different to an Omex, DTA or GEMs unless you are using the car at an exeptional level (racing, hard use extreme environments etc). If the mapping isn't as good as it could be, throwing money at a more expensive ECU won't solve the problem.

So in short, there are plenty of alternatives to the Emerald all with varying costs and functions. But be careful with it, a lot of the stuff you're paying for on the more expensive ones are things you will never need, and the major problem with most aftermarket ECUs is mapping, so spend your money here, and insist that its done properly.

[julian64]

Can I add that if anyone has a duff ECU and are about to throw it away, could they throw it in my direction. Easy fixes and I'll return it.

[rev-erend]

Nick - that sounds like good advice ...

Guess you must have been there, seen it, done that.

[Mars]

ro_butler said:

Mars said:

ro_butler said:

I mailed the company and they say that it will cope with a flat plane V8, which is usually the main problem.

I thought the issue was the angle of th AJP V rather than it having a flat plane crank.

Given that it allows full sequential injector and spark operation I thought the angle would not be an issue, but the firing order might pose a problem. Could be wrong though.....

I am just hoping that if my ECU breaks there are more than 2 options open to me (try to find a 2nd hand MBE [currently expensive] or Emerald).

As I understand it, the AJP's angle leads to a unique firing sequence. It is more akin to a heartbeat, dada, dada, dada as opposed to a regularly-spaced sequence of sparks.

Best check with someone who knows for certain - Joo or Karl @ Emerald I suppose.

[eLSerbera]

450Nick said:

For a road car, full sequential isn't really necessary and you won't really notice any difference over wasted spark. EGT is useful but only if you're planning on running things close to the limit and things like dual lambda are useful for closed loop adaptive fuelling, though a single lambda is fine for the road.

Hey just wanted to ask, but I always thought that it was for race cars that full sequential injection and ignition was unnecessary, as most time is spent at WOT and that the fully sequential injection reduced excessive fuel burn and helped with emissions most when at partial throttle openings?

I am sure I've read that somewhere, and when you are running batch injection aren't you really making the injectors work in a similar way to a carburettor?

[450Nick]

rev-erend said:

Nick - that sounds like good advice ...

Guess you must have been there, seen it, done that.

Jeez, you wouldn't believe it... I've had maps made too lean, too rich, no cold start, no corrections, too much correction, no adaptive, too much adaptive the list goes on. It actually seems to be better to get someone to map it, then take it back and have them tweek it once you've found the problem areas, and then learn to do it yourself and iron out the problems in a way which suits your driving style. You most certainly need to set the car up on the road as well as the rollers though, and in more than one ambient. You really can't get drivablility from the car unless its had a good bit of time being mapped on the road as the rollers can't simulate the real world for some reason.

[450Nick]

eLSerbera said:

450Nick said:

For a road car, full sequential isn't really necessary and you won't really notice any difference over wasted spark. EGT is useful but only if you're planning on running things close to the limit and things like dual lambda are useful for closed loop adaptive fuelling, though a single lambda is fine for the road.

Hey just wanted to ask, but I always thought that it was for race cars that full sequential injection and ignition was unnecessary, as most time is spent at WOT and that the fully sequential injection reduced excessive fuel burn and helped with emissions most when at partial throttle openings?

I am sure I've read that somewhere, and when you are running batch injection aren't you really making the injectors work in a similar way to a carburettor?

To be honest, I'm not sure when full sequential is really needed. I'm sure you can squeeze a little bit more economy out of the engine so perhaps its good for emissions? When I looked into getting it, I was told that you wouldn't really notice any gain in fuel economy unless you really carefully measured it, and the TVR emissions regs weren't so tight that you needed it. So I've always assumed it would be good to try and squeeze the last bit of power out of a race engine by ensuring all fuelling is very tightly controlled. Then again, I'm no expert on it!

[spitfire4v8]

450Nick said:

rev-erend said:

Nick - that sounds like good advice ...

Guess you must have been there, seen it, done that.

Jeez, you wouldn't believe it... I've had maps made too lean, too rich, no cold start, no corrections, too much correction, no adaptive, too much adaptive the list goes on. It actually seems to be better to get someone to map it, then take it back and have them tweek it once you've found the problem areas, and then learn to do it yourself and iron out the problems in a way which suits your driving style. You most certainly need to set the car up on the road as well as the rollers though, and in more than one ambient. You really can't get drivablility from the car unless its had a good bit of time being mapped on the road as the rollers can't simulate the real world for some reason.

You're right Nick .. every rolling road session should be finished off by running the car on the road .. I can spend up to an hour on the road refining the part throttle map for best driveability, you just cannot replicate the feel of the car on the dyno, plus the majority of dynos cannot replicate downhill sections of road on trailing throttles etc .. rolling roads are excellent for geting you 90% of the way there but they're not perfect.
I can't think of one occasion when ive brought a car away from a paid-for mapping session where the map hasn't had to be tweaked in some way afterwards to make it drive nice at low loads, except for emerald mapping our race T350 where i've not touched it.
I know one mapping place that actually refuses to drive the cars on the road, doesn't make sense to me other than they would then have to spend another hour on the car ..

[ridds]

It's not just the aftermarket either.

Even with some of the £10m engine dyno cells I've used you still have to put the Calibration in a car to check it's ok and usually you'll end up with some lean hot spots in it.

[350Matt]

450Nick said:

eLSerbera said:

450Nick said:

For a road car, full sequential isn't really necessary and you won't really notice any difference over wasted spark. EGT is useful but only if you're planning on running things close to the limit and things like dual lambda are useful for closed loop adaptive fuelling, though a single lambda is fine for the road.

Hey just wanted to ask, but I always thought that it was for race cars that full sequential injection and ignition was unnecessary, as most time is spent at WOT and that the fully sequential injection reduced excessive fuel burn and helped with emissions most when at partial throttle openings?

I am sure I've read that somewhere, and when you are running batch injection aren't you really making the injectors work in a similar way to a carburettor?

To be honest, I'm not sure when full sequential is really needed. I'm sure you can squeeze a little bit more economy out of the engine so perhaps its good for emissions? When I looked into getting it, I was told that you wouldn't really notice any gain in fuel economy unless you really carefully measured it, and the TVR emissions regs weren't so tight that you needed it. So I've always assumed it would be good to try and squeeze the last bit of power out of a race engine by ensuring all fuelling is very tightly controlled. Then again, I'm no expert on it!

Full sequential injection is essential in getting the best out of a very high revving engine (15,000+) getting injection timing out by 30° on the F1 engine cost about 30Bhp.

However in the less rarefied world of 6Krpm then you won't get any extra power with the std/ back of the valve injector position. It will however allow you to run a leaner fuel mix without 'shunting' and will typically save 10 -15% fuel on a road car.

If you want to make power with sequential and at road car rpms then you'll need a dual injector set-up with one at the very end of the inlet and the other in the std back of the valve position. Your ECU should then switch over between the two injectors at a certain load level, this keeps the engien driveable at low speed and provides the extra power at the top-end.

[ro_butler]

spitfire4v8 said:

I know one mapping place that actually refuses to drive the cars on the road, doesn't make sense to me other than they would then have to spend another hour on the car ..

Could well be down to insurance, or accident worries. I know one mapper who also only maps on dynos.

Nick, I know only too well the trials of mapping. My Skyline was constantly on the rollers being tweaked, and the map was never quite right. I realise that the mapping is important but a lot of the cost is down to wiring the ECU and sensors so if you go for an unusual chice of ECU you could spend lots of time/money getting it to work at all. If somebody has got this working with a speed6 then at least part of the work has been done.

I also fully agree with the ease of use of the mapping software as like you say, it would be great to tweak the mapping yourself.

I favour sequential operation simply because it makes sense to me, in a way that batch firing doesn't. I am aware that I wouldn't get any more power and not much extra economy. I also like like the sound of the multiple switchable base maps and numerous compensation maps. Loads to play with

Emerald are certainly a good option, tried and tested, but the Link looks far more advanced. I too have a spare MBE ECU but this ECU looked interesting, offering endless hours of fun/frustration

[GreenV8S]

It puzzles me that there isn't a general common way of mapping cars on the road. I mean, the algorithm for doing the steady state setup is going to be basically the same (drive it towards a target lambda over a 2-D load/rev map) and I suspect there would also be common algorithms to set the quantity and duration of the AE enrichment to stabilise the AFR during throttle transients. It seems like the sort of job that would be suit a feedback loop and a Kalman filter with a suitable driver to access the ECU configuration and logs.

I'd write one meself, but ... I'm lazy.