Two engines controlled by one ECU POSSIBLE???

[Eryk Sokolowski]

I want to know is it possible to control separate two engines (singles or maybe twins not bigger than 300cc each) with fuel injection, running with a common intake (20mm restricted) with a common ECU.

I would be grateful if you could provide me with some information such as:

<UL TYPE=SQUARE>Is it technically possible?

Could you provide us with such ECU (rough estimate of cost for custom ecu's)?

Is this a good idea?[/list]

I would need something like: a custom ECU which could control both engines independently, it would possibly have to have two crank sensors inputs etc.


HELP

[Eryk Sokolowski]

I want to know is it possible to control separate two engines (singles or maybe twins not bigger than 300cc each) with fuel injection, running with a common intake (20mm restricted) with a common ECU.

I would be grateful if you could provide me with some information such as:

<UL TYPE=SQUARE>Is it technically possible?

Could you provide us with such ECU (rough estimate of cost for custom ecu's)?

Is this a good idea?[/list]

I would need something like: a custom ECU which could control both engines independently, it would possibly have to have two crank sensors inputs etc.


HELP

[Jimmy01]

I could be wrong but I doubt any commercially available ECU is set up to look at two different engine speeds. It's certainly possible however it seems much easier/simpeler to just run two ECU's. The only way I can think of to run two engines with one ECU is to connect them via the crank so they always run at the same speed and timing, but then you may as well just use a bigger engine!

One restrictor wouldn't be a problem, unless your engines are quite far apart, then you might have some piping issues.

[Eryk Sokolowski]

Do you know if there is a way maybe through a microcontroller to make them talk to each other... Maybe by can, I want to have a way of making sure that they run at the same speed, and if not then the spark is retarded to slow one down...

The reason for this is that we can have only one throttle prior to the restrictor.

Do ecu's support spark retardation on can request??

[ibanezplayer]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Eryk Sokolowski:
The reason for this is that we can have only one throttle prior to the restrictor.
</div></BLOCKQUOTE>

Much easier to get the TPS signal to two ECUs, or to put two TPSs on your throttle....

[Zac]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Is this a good idea? </div></BLOCKQUOTE>

The magic eight ball says:

"All signs point to no"

[Adambomb]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Eryk Sokolowski:
Do you know if there is a way maybe through a microcontroller to make them talk to each other... Maybe by can, I want to have a way of making sure that they run at the same speed, and if not then the spark is retarded to slow one down...

The reason for this is that we can have only one throttle prior to the restrictor.

Do ecu's support spark retardation on can request?? </div></BLOCKQUOTE>

The best way I know to make sure 2 engines are running the same speed would be to join them at the crank, like Jimmy said. Trying to do it by making 2 ECUs talk to each other and alter the tuning appropriately sounds like a HUGE pain.

But again, if you're doing that, what have you gained compared to just running a bigger engine?

[Hector]

Who says he's not trying torque vectoring, aka "Skid Steer"? Definitely don't want your cranks operating at the same speed then http://fsae.com/groupee_common/emoticons/icon_wink.gif

[AndrewUofL]

I could actually see it being possible and simpler than what everybody thinks. You could connect the engines via the crank so that the pistons in both of the engines are paired so that you now basically have a 2 cylinder engine that would operate as a single. From here you could run design a single intake that would be similar to a normal intake on a 2 cylinder. This would give you a single throttle, plenum, intake and TPS for both engines. Best of all you may be able to use one ECU set up for a single cylinder engine and splice wires since the engines are operating as a single cylinder.

The other way that this may be possible is use the same style intake but offset the engines so that the one crank is 180* offset from the other so you simply have an engine that operates as a 2 cylinder engine. You should be able to run this configuration with a simple ECU like a PE. Also, as far as tuning, you may run into an issue that the cylinders would be running different AFR's (not uncommon) and you can simply use a cylinder compensation when running with this method to correct the issue. If I was to try joining 2 engines, I would try this method first.

Now if one of these methods would work, I would guess that it would be heavier and a lot harder to maintain and tune. Another problem that you may run into is your method of connecting the engines. If you connect them together in series, you will have 2x the force in one crank because it would be transferring the load into the driveline from both engines which may cause that crank to fail because it was only designed for the load of one cylinder. So you would have to have you "connector" between the 2 engines also be you location of energy transfer into the driveline. also, with both set-ups, since you are tricking the ECU into thinking that it is controlling one engine, you only need one Crank Pos Sensor.

Anyway, im making this too long and I dont want to do your design for you if it was something that you are considering. Also, in case anybody was wondering, Rule B8.1.3 states that multiple engines can be used as long as the total disp. does not exceed 610cc and all the air comes from one restrictor.

Andrew
Louisville Motorsports

[Eryk Sokolowski]

Joining engines with a common crank is not a solution.

We already looked in to that solution...

I am not looking for mechanical solution but i would like to solve this problem via electronics, 2 engines 1 ecu or 2 engines 2 ecu's which can communicate with each other...

Any ideas please?

[AndrewUofL]

As far as 2 engines 1 ECU, the above is the only solution that I can see. I am curious though, Why would you want to run 2 engines on 1 ECU that are not joined by some mechanical means? If you are looking at running 2 engines on a formula car to propel the car, there has to be some mechanical connection between the two even if it is the road surface. I could probably help you out better if I have an idea of what type of system you would need to control 2 engines with 1 ECU. I may just be misunderstanding what you are trying to do/ gain with multiple engines. Please explain further what you are trying to do.

[js10coastr]

Anyone know if the JCB Dieselmax had one ECU? I can't find any all encompassing literature at the moment.

“We have talked this through, and so long as the engines are close enough to each other in specific output and run at the same speed, they will regulate themselves mechanically,” Leverton explained.

from ingeria.org

[wweissin]

Just a stab in the dark here but...
If you were to run both engines as batch fire or a TBI setup you might be able to use the crank and cam trigger from one motor to run both motors.

Basically piggy back the second motor. You wouldn't really have control over much because on the second motor you would have no reference or crank speed and the tune would be for the first motor that is sending the crank and cam info to the ECU but if the engines are close enough to each other it might work. And if you do not use sequential fuel and spark you would have more leeway on your ignition timing and would really have to worry about injection timing. Just my quick .02

[L B0MB]

Two engines; why?

Too many drugs for breakfast?

[Drew Price]

Do they have to be able to 'talk' to each other, I.E. do you need to be able to specify the torque request for each independently?

Using 1 ECU to control both seems nearly impossible to me - you will never assure that both engines will start in the same part of the cycle (or even in the same number of cranks) and as soon as they're out of phase, the 'master' engine will start, the 'slave' engine will just have a coil and plug firing happily away not matching what the hardware is doing as soon as the other spins up.

Why not 2 engines and two ECU's - and one throttle cable?

I am with Justin (js10coastr) that two similarly configured engines receiving the same inputs will self regulate. They are actually coupled - through the tyres of the car. (the ground couples them!)

There have been multi-engined tractor pull tractors forever, and if the cranks aren't physically joined, they are relied on to just self regulate.

And they are often carbureted!!!

There was a twin engined Hyundai Tiburon drag car right when the car came out, one engine in front, one in back, two sets of everything and each had it's own ECU, they just received the same throttle command signal to open the throttle the same. I remember seeing an interview with one of the tech directors or someone saying that there was some initial calibration to sync them, but that it just worked pretty simply.

The engines have to maintain very close to the same spec, and any deviation in sensors, injectors, fuel pressure, things like that will throw it all off.

How will you take into account accel and warmup enrichments for each running at slightly different temps (where will the radiator for each engine be mounted? or do they share? are they in airstream?)

Best,
Drew

[RenM]

if they are not connected together mechanically and you can not assure a constant ignition spacing your engines will hardly work at all if you are using one intake. (which you have to do because you have to use one restrictor)

There will be a huge difference in the amount of air every engine gets dependent on your ignition spacing! If they both suck air at the same time they will get far less air then with a 360° difference. It will be impossible to tune the engine mapping because you will either run extremly lean or extremly rich, to the point where your mixture wont be able to combust.

[ZAMR]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RenM:
if they are not connected together mechanically and you can not assure a constant ignition spacing your engines will hardly work at all if you are using one intake. (which you have to do because you have to use one restrictor)

There will be a huge difference in the amount of air every engine gets dependent on your ignition spacing! If they both suck air at the same time they will get far less air then with a 360° difference. It will be impossible to tune the engine mapping because you will either run extremly lean or extremly rich, to the point where your mixture wont be able to combust. </div></BLOCKQUOTE>

2 stage intake

[murpia]

Eryk,

If you won't explain why you need to do this, any advice is likely to be way off, and many will just accuse you of trolling.

I suspect you want to either torque vector a rear axle, or run a 4WD setup with one front engine and one rear engine. (Otherwise you would just link the crankshafts and your problem goes away).

Either way, you can just use 2 ECUs, there is nothing in the rules preventing this. The rules mean you are limited to one throttle, but that can have 2 sensors. Anyway with such a setup you are probably better off with a MAP sensor in each engine's plenum to get the fuelling right.

A better query would be 'how do I modulate an engine's torque given a fixed airflow', since I think that is what you want to do. That in itself is an interesting debate as it relates to traction control, etc. and applies to many engines and cars. But unless you explain more, no-one can realistically help you.

Regards, Ian

[dazz]

If the two cranks are not rigidly coupled, then I think that it would be extremely difficult if not impossible to achieve this with a common ECU. (unless there is an ECU that is capable of running two different engines completely independantly)

With both cranks are rigidly coupled, then you can treat them as a single engine and use a common ECU. On a slight tangent, a bloke here in Aust is/was using two Motecs to run fuel injection on a Meteor V12 Rolls Royce converted to twin turbo and running custom DOHC heads. I believe he was treating each bank as it's own straight six turbo!

I can't see why you couldn't run an ECU for each engine? Split the intake imediately after the TB and run 2 MAF sensors, one before each manifold. This would help avoid tuning issues from one engine starving air from the other. Seems like a lot of hassle though? Care to share any more details of your concept?

[RenM]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ZAMR:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RenM:
if they are not connected together mechanically and you can not assure a constant ignition spacing your engines will hardly work at all if you are using one intake. (which you have to do because you have to use one restrictor)

There will be a huge difference in the amount of air every engine gets dependent on your ignition spacing! If they both suck air at the same time they will get far less air then with a 360° difference. It will be impossible to tune the engine mapping because you will either run extremly lean or extremly rich, to the point where your mixture wont be able to combust. </div></BLOCKQUOTE>

2 stage intake </div></BLOCKQUOTE>
you will need 1 restrictor and thats the problem

[Diablo_niterider]

there was this 1 car with 2 kawasaki zx12r engines (1 for front and 1 for rear), it was featured on a show with tiff needell (5th gear host), i dnt rmmbr the name but i guess it had a single ecu and it looked like the caterham. and it looked fun to drive and tiff pushed it so hard, he ended up blowing up 1 of the engines, do check it

[Eryk Sokolowski]

tiger z100 and it was rear wheel drive, they made only one with 4 wheel drive for acceleration record...

they are no longer in production...

already spoke to the company making them...

the engines are completely independent, it has two gear boxes and they are coupled by a Y diff...

[ZAMR]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RenM:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ZAMR:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RenM:
if they are not connected together mechanically and you can not assure a constant ignition spacing your engines will hardly work at all if you are using one intake. (which you have to do because you have to use one restrictor)

There will be a huge difference in the amount of air every engine gets dependent on your ignition spacing! If they both suck air at the same time they will get far less air then with a 360° difference. It will be impossible to tune the engine mapping because you will either run extremly lean or extremly rich, to the point where your mixture wont be able to combust. </div></BLOCKQUOTE>

2 stage intake </div></BLOCKQUOTE>
you will need 1 restrictor and thats the problem </div></BLOCKQUOTE>

Have the restrictor go to one plenum. Then, have 2 runners go to 2 separate plenums. Have the two separate engines draw from the two separate plenums. That dampens the pressure effects and intake event problems.

[Wesley]

Though your flow losses will mount significantly.

[Sisyphus]

I like the way you think but you are thinking too small.
Back in the 70's there was a CanAm car with 4 two cylinder, 2 stroke McCullock engines in it--one for each wheel. No ECU's in them days--all mechanical linkages, etc
I think it might have made one lap...

[RenM]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ZAMR:

Have the restrictor go to one plenum. Then, have 2 runners go to 2 separate plenums. Have the two separate engines draw from the two separate plenums. That dampens the pressure effects and intake event problems. </div></BLOCKQUOTE>
If you are working with multiple plenums and a lot of volume you will have a catastrophic throttle response and still not be able to completly compensate the effects.

[Adambomb]

We've sort of "bench raced" the idea of using two engines (never seriously), more so that you could get rid of the diff and even get "torque vectoring" as Hector mentioned. Obviously, as Drew mentioned, both engines would necessarily need to be able to run at different speeds.

And even then, taking into consideration what has been mentioned here, it sounds like exponentially more pain than what it is worth (mostly due to the fact that you are forced to run a single restrictor). I have a hard time believing you could get it to run right.

If torque vectoring is what you want, IMO you'd be much better off with one larger engine and an active diff!

On the other hand, one other thing we came up with when discussing multiple engines: Baja rules don't specify that you can only run one engine...

[Adambomb]

Also, forgot to mention: ISU has a mini-tractor pull team (Power Pullers). Their rules state that they can use as many engines as they want, as long as they are un-modified spec. Briggs and Stratton v-twins (I'm thinking they are like 16 or 20hp) and total vehicle weight is under 900 lbs. This year I think they're running 5. One of their old cars has 4 of them, and they are timed for even-fire, it sounds vaguely reminiscent of an old flat-head Ford V8.

[ZAMR]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RenM:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ZAMR:

Have the restrictor go to one plenum. Then, have 2 runners go to 2 separate plenums. Have the two separate engines draw from the two separate plenums. That dampens the pressure effects and intake event problems. </div></BLOCKQUOTE>
If you are working with multiple plenums and a lot of volume you will have a catastrophic throttle response and still not be able to completly compensate the effects. </div></BLOCKQUOTE>

Heisler page 263. 2-Stage intakes exist. They are deisnged for multiple cylinder engines that have overlapping intake events.

You need 1 restrictor. yes. You need 1 intake. Yes. You CANNOT tune an intake if you CANNOT predict the intake event timing. You need 2 plenums, with a damping plenum beforehand. If not, you may have intake events that send pressure waves that interfere with drawing air through the restrictor. If you want to use Helmholtz business, you need to separate the intake events.

Perhaps you will have flow losses, yes. But I see this as the only way to maintain consistent airflow and wave properties, ergo a consistent fuel map and consistent air/fuel mix and VE for a given RPM/throttle. Otherwise you can't drive your car.

If you can make sure the ECU lines the cranks up (impossible or very hard) then you can make a normal intake.

I don't agree that the idle will be harder to control. Your total plenum volume for the 3 plenums for the two engines should only be slightly larger than if you ran 1 engine with equivalent volume. Airflow will be smooth because of the damping chambers, but you should be able to control the plenum pressure just as effectively as a single engine.

[L B0MB]

Airflow meters would be the only way to run 2 engines off one restrictor as MAP would be constant across the x many plenums whereas engine speed would vary depending on available grip at the driven wheel

Edit: 2 engines in fsae is a silly idea

[ZAMR]

I agree wholeheartedly.

[Wesley]

Zach, a sufficiently large plenum would also be an acceptable design, it would seem.

And I wasn't arguing that flow losses would make that intake design not work, I was arguing that running two engines is a poor idea because of the compounding of flow losses, as well as friction, rotating inertia, and overall power to weight reduction.

[RANeff]

I just want to know why this question came up. Noone would ever use this idea (in FSAE that is, some other comps may http://fsae.com/groupee_common/emoticons/icon_wink.gif )

[AxelRipper]

For some reason, the fact that it is SPECIFICALLY pointed out in the rulebook makes me think that sometime back in the days of 2 strokes and rotaries that someone tried it as a way to get around the 20mm restrictor rule and run 2 separate restrictors.

does anyone happen to know if this ever did happen?

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by L B0MB:
Airflow meters would be the only way to run 2 engines off one restrictor as MAP would be constant across the x many plenums whereas engine speed would vary depending on available grip at the driven wheel </div></BLOCKQUOTE>
I disagree.

Yes, with linked plenums the MAP would be the same or similar for each engine, but the fuelling calculations for a MAP based system take engine rpm into account. It's often called 'speed density' and the clue is the first word... Usually a volumetric efficiency table is used to estimate cylinder charge from plenum air density (calculated from MAP sensor plus air temperature sensor). MAP based systems can bet setup to run without a TPS sensor although it's useful for other compensations such as transient fuelling, flood clear, detecting idle state etc.

Speed-density is one of the default options for the Megasquirt system, and using one Megasquirt per engine would be a good place to start for the original poster.

I still see no reason why two separate ECUs do not work in this application, unless this is not for FSAE and instead for another series where only one ECU is permitted.

Regards, Ian