Flat Shift Suggestions

Hi guys,

I’m tuning the closed loop flat shift on a 500hp turbo 4 cylinder race car with an AGB10 6 speed sequential box. The car shifts perfectly, however it’s having huge backfires due to 100% ignition cut + 20 deg retard.

I also tried retard only without cut, and set the total retard amount to 45 degrees. This made the shifts quiet, but the driver noticed that more force was required on the gear lever for shifts. I am just unsure if this method is safe for dogs or will cause premature wear.

I have the option of using 80 or 90% ignition cut, however also not sure if this is safe for the dogs if a cylinder fires mid shift before they are engaged

my other option is a fuel cut + retard. I’m not a fan of cutting fuel at 25psi boost, as there will be some lean cycles while the fuel puddle is being emptied / topped up from the cut.

How do you guys normally set up a box like this?

cheers, Dean

I assume you're using a load cell on the lever to trigger the cut - do you have the option of having the ignition cut for, say, 100mS and the injectors for, say, 80mS? That should give a clean cut and reduce the exhaust noise by reducing the fuel in the exhaust and leaning it out so it doesn't ignite easily, whilest ensuring it doesn't lean out in the cylinder by re-wetting the ports if required?

You are going to have to cut the fuel to eliminate the combustion in the exhaust. You should cut fuel and ignition, and with your 25 deg retard I would think there would be no ignition-induced knock, regardless of how lean your mixture. I hear your concern, but I think if you consider the shift time, and engine speed (post shift) you will find there are very few cycles without fuel. The boost will keep the fuel puddle from evaporating, and when the gear shift is complete you will be putting the correct amount of fuel back in.

I will admit, I have no experience with 500hp turbo engines, nearly all of my work is with normally aspirated. And to eliminate exhaust pops, we cut the fuel.

Thanks guys,

There is no option for a percentage of one or the other. You can do both together with an overall cut percentage.

the car was initially setup this way with 100% fuel + ignition cut. It would pop around 50% of the time (probably from where an injection cycle has completed but the plug hasn’t fired).

There were large lean spikes in the log from this, and so I ran a 200ms recovery time to blend the timing back in slowly as I don’t know how much of the lean spike is artificial from the ignition cut, or actual from a lack of fuel buildup after the cut.

What sort of recovery times do you use after a fuel cut? Do you use 100% cut or much less? How would you know if your strategy is causing excess wear on the dogs or not?

Cheers

You mention large "lean" spikes, I would remind you that the lambda can only read the oxygen content in the exhaust gases. If there is an ignition only cut and there's any oxygen that wasn't used up by fuel burning in the exhaust from being ignited by the heat there, it will show up as a 'lean' mixture. Same thing with fuel cuts, the oxygen in the air is passing through the engine to the lambda.

Doesn't necessarily mean you aren't getting a momentarily lean condition, but don't get too hung up on it.

If there's still a load on the gear lever, it means there's still an excess of drive torque and perhaps the best solution will be to add a LOT of ignition retard to maintain turbo' speed and boost, but bringing the timing back ove a few hundred milliseconds to reduce shock loadings on the transmission and reduce the chance of tyre or clutch slippage on the change?

In the end, though, you're going to have to figure out what works best for you, the driver, and the vehicle in your specific application.

Yes, I’m aware that the lean spikes are from the cut, but was more concerned about actual lean cycles after a fuel cut with boost. The lean spikes from a fuel + ignition cut were much larger and longer than from an ignition only cut with similar cut times. I’m not sure if a single knock event at 125whp per cylinder is generally tolerated or devastating to engines.

When I used retard only (no cut), I was at 20deg ATDC for the shifts. The ecu will allow a further 15 degrees retard, but at this point I’m in unknown territory and don’t know if there are negative effects to the engine going that far.

I think the original plan with fuel + ignition cut will be the best. Looking at the log, it only takes around 50ms from the peak of the lean condition to return to a suitable AFR, so the recovery time can be much shorter than I originally thought.

thanks for the advice!

I find fuel cut not much good for gear shift with a port-injected engine, it is slow to reduce torque and slow to recover so the shifts feel long and are much less seamless. A nicely tuned shift should feel and sound almost seamless and shouldn't need a lot of driver effort.

Try reduced ignition cut with a positive fuel trim added, this usually eliminates or at least softens the bangs a lot. ~90% cut and 15% extra fuel often works out for me.

I have seen some maps from older LMP2 and BTCC level cars (NA engines) that are using retard only for gear shift (~60deg), but I haven't personally experimented much with that yet. Although I suspect at this level the transmissions would likely have a more optimised dog design, less undercut etc than we use as peasants in club motorsport.

Thanks for the advice Adam!

Ill give this a try next. Are you using any retard in combination with this? I think I will have to keep some retard in the recovery so the tyres don’t break loose. This probably doesn’t help the bangs.

Shifts did work (although a little tighter) with no cut and 45 degrees of retard, so 60 should be even easier on the box. Good to hear of a real world example with that much retard.

The flat shifting I have setup using sequential box usually feels much smoother from the drivers perspective using ignition retard only.

Make sure you have a decent amount of preload on the dogs before igniton retard is applied. Check this tolerance with barrel position and set threshold points accordingly.

Only 'downside' I've seen from using lots of ignition retard for gear cuts is some slight boost creep during the retard. Can see 1-2psi increase briefly but the boost wont drop at least, so is usually faster.

Cheers Mike,

The voltage for each of the 6 gears have been calibrated, but the threshold is set at the default +/- 0.1V. Is this the tolerance you are referring to? If so, do you have a good process on how to accurately set this?

If you are referring to strain gauge force, I set this by pulling on the gear lever with the engine off and looked at the Voltage required to naturally fall out of gear. For memory, when the strain gauge was at rest it output 2.5V and it took around 1.1V to naturally come out of gear. So I set the cut threshold to 0.9V for slight preload.

Can you give me an idea of what sort of retard values you are using for the shift. I have only tried 45 degree retard which put the ignition angle at around -20deg during shifts, but haven't tried any further than that yet.