Monitoring AFR balance in a boosted engine
Turbo Mini Forums

With sequential code getting close to testing, there is one thing that has been bothering me. When calibrating up the staging to a particular engine/Cam/inlet combo you need to measure the imbalance between outer and inner cylinders at a number of load and RPM points. Now with a natasp this is easy, as you just weld a pair of lambda sensor bosses in and either do 2 sets of logs, or for the well heeled fit 2 widebands. For a turbo you cannot do this as nernst cells neither like the high temps or high pressures of a pre-turbo exhaust.

So the options left are to calibrate the injection without a turbo, or to fit a pair of EGT probes. The latter is possibly the better solution, but then you need a way of logging these. Luckily we have the technology (actually there are 3 options that spring immediately to mind) provided you can find a good welder for cast manifolds (those with custom tubulars are laughing at this point.

Thoughts?

Edited by bill shurvinton on 12th Mar, 2007.

id go for the ex temp probes bill , the innovate aux box can take a couple i think.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

It can and logworks is now MS compatible ( and a great analysis tool), but that is an additional cost to those on a budget. I'll need to check, but there might just be enough analogue inputs on MS2 to log them natively with a little additional PCB, which would at least be neat.

Hi,
Is it just a case that the reading is out by a percentage due to the turbo back pressure, or does it kill the O2 sensors?
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary

Both. You need to do a correction for EBP (which none of the currently available hobbyist units do) and they get very grumpy above 750C.

I know from Carl Austin's mini that with the centre-branch seeing twice as many pulses as the outer branches, it ran a lot hotter. I don't see that this will therefore be accurate enough?

My original plan was to do the testing on my system with the wastegate jammed open, then again at a ridiculously low boost like 2/3 psi to see if there is any change (which I doubt there will be).

I'll have an EGT in the centre one to monitor the temps - but only to prevent cooking the wideband.

I'm now not sure I'd bother to pursue the 'low boost' test; I really don't want to cook or blow my wideband.

The correction for EBP might not be necessary if you only want to see relative variation of AFR between cylinders. So for timing purposes, you could just see if you get the same reported AFR for outside and middle branches but to get the correct AFR, you'd need to have a wideband after the turbo for both pressure and temperature reasons mentioned.

Dave, if you have an EGT sensor then you could monitor closely the temperature and go as high as possible with the boost whilst keeping the temperature in the correct range. I doubt you would fry a sensor at 750C or even 800C very fast even though the reading would be completely wrong.

Bill, there is already a daughterboard for MS (mainly MS1/extra) that has 2 EGT conditioner circuits on it: the error* board sold by glensgarage. This uses the AD595 EGT amplifier which is a bit pricy but seems to work well. There are 2 unused ADC inputs on MS2. However, they are not configured for EGT inputs but that could be changed. The standard options for those ADCs is 2nd O2 sensor and knock sensor.

Jean

http://www.jbperf.com/

I use the extra board, and have a small stash of the AD 595 for one offs. The problem is that you cannot easily fit MS2 and the error* board in a standard case, and the wiring is a bit of a monster. When the higher case becomes available this will be eased. OTOH it does have a basic knock detect on as well, which could be of use.

Wouldn't take much to knock up a custom board for this of course.

It would probably be a good idea to have a board more dedicated to the MS2 V3 setup. The error* board was designed for the MS1/extra V2.2 setup and has some no-longer necessary stuff.

I'll have a look at that and see how an extra board and my p&h board would fit in the original case with MS2. The p&h board will be useful with an upcoming version of my code with full sequential option.

Jean

Edited by jbelanger on 12th Mar, 2007.

http://www.jbperf.com/

how about temp sensor rings placed under each plug ?

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

i liked stevies idea of small diameter pilot tubes comeign out the centre and side branch, in which wideband lambda's are located.

with a suitable restriction the pressure in the tube where the lambda sits would be neglegable, and as it is remote from the manifold temp sould not be an issue either.

teh only possible problem here would be lag from whats in the manifold being read by the sensor, but this would only be 1 engine cycle at most.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Sadly you have violated some laws of physics there. You could build a venturi off the main exhaust, but that is a lot of work compared to a pair of EGT probes. Remember, after you have balanced things then you can do all the tuning post turbo.

I must admit that I was thinking along the same lines as TD.

To be fair I would probably not start with a forced induction motor for testing / familiarisation (sp?).

Every day is a school day ...........

How fast and how expensive ...... the same question...

I agree that until a tuning method for injection timing has been established and initial ballpark figures have been found, I wouldn't try to tune a boosted engine. I was just saying that once this is done then the EGT probes could be used to check that the lambda sensors are safe. You could also see how closely you can rely on EGT for balance tuning purposes.

Jean

http://www.jbperf.com/

I agree Jean. If I was to do a boosted motor at this time I'd do the same as TD suggests, lock the wastegate open and run it fairly gently.

Every day is a school day ...........

How fast and how expensive ...... the same question...

The temperature of the exhaust gas will be the same it is just that the outer runners will have more time to conduct heat away before the next pulse. Given the slow reaction time of EGT this would be averaged out.

I don't have the answers, but I think it is worth discussing.

The problem with running the engine without boost and/or gently is that you then have a smaller injection pulse width. This means that you're using a smaller percentage of the injection window which means you have to play more with the injection timing going back and forth to find the correct angle.

One solution might be to use smaller injectors when doing the initial tuning to use a bigger percentage of the injection window. The disadvantage of this is that you may not have the same injector characteristics with the different injector sizes.

Jean

http://www.jbperf.com/

That is a good point Jean, less of an issue for staged injectors, but for a single pair of injectors it would be more so.

I still think it is the most proactical option at present. So long as you made a reasonable attempt at calculating the window size and position across the rev range it wouldn't take too long to establish the operational boundries.

True it would mean a little more time on the rollers for the first few people as we rather data but in the long run I am fairly sure a set of generic rules could be formed that would work for most people. Those people running something out of the ordinary would be back at square one (maybe two) with each engine but those people would probably be the sort who'd want to see it through from first principals anyway!

Lets face it, most people will buy a manifold that comes off the shelf, a turbo off the shelf and cam all at the same time. These "standard" build specifications should be easy enough to advise for as you'd select injectors such that you weren't hunting for the very last % of window.

Every day is a school day ...........

How fast and how expensive ...... the same question...

bill i was talkign about having two relaitvely small tubes comeing off the manifold and venting to atmosphere with a small chamber and the wideband in the flow. as manifold pressure will always be greater than atmospheric, it will allow for constant sampling of the exaust gasses. only trouble i see with this is contamination of the sample in one runner when another runners port is exhausting. more complicated than 2 (or 3) EGT probes but gets around the issue of the centre branch running hotter than the outer two, as the centre branch is bound not to be exactly twice as hot as the outer branches and that then leaves the question of how do you know what temperature difference indicates a balanced mixture.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Hmmm. I call that a wastegate. But one that is always open.

The centre branch wouldn't be twice as hot with a balanced engine. You have twice the mass of gass per second, but at the same temperature, so losses (Qdot in the textbooks) would account for less heat loss. If you were to wrap the first few inches of header where the EGT probes should be anyway, then this should be minimised.

Of course another option is a tank of leaded and plug cuts...

.

Edited by bill shurvinton on 14th Mar, 2007.

I get you. . thermodynamics was never my strong point :) still i dotn see a problem running dual wideband in the setup i described, if sensible sized capillary tubes are used the pressure/flow loss from the perspective of the turbo would be very minimal still EGT is certainly cheaper and cleaner to implement.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.