Conversion to five speed stalling issue

I recently had my 1990 XJS converted to a 5-speed manual by a local shop. They used a kit from the Driven Man. The car is back together and is a lot of fun to drive with one exception. When you disengage the clutch and take your foot off the gas at speed (like maybe you are going to turn a corner) the rpm’s fall to zero and the car stalls.

Does anyone have any suggestions as to how to remedy this? Could this be an issue that needs to be addressed through the onboard computer?

Yes. It’s why the Suppllier should mention it and preferably include a fix with their kit. Contact Roger Bywater at AJ6 Engineering (Google it)

The V12 fuel injection cuts off injectors at closed throttle above 1200 RPMs. When you put a manual transmission, the drop in RPMs is to fast for the engine to recover on decel so by the time the injectors turn back on, the engine has pasted the point of recovery. The fix is to have that function on the ECU disabled. Roger Bywater is the one that knows how to do this.

My mind is on 5 speeds too. The results for Bywater ECU are guaranteed but not cheap! For those of us with less money and more time, is there something we can do to retard the change in TPS signal? Basically keep it from hitting “closed” for a second or two so the RPMs drop below 1200 before the closed signal is sent. Maybe a teeny tiny capacitor… in-line from the TPS signal.

I am sure it can be done because Roger Bywater did it for my car. If you don’t want to send it to Roger, then you can search and see if someone else can do it. As a temp solution, you can adjust the throttle position sensor so the ECU does not sense that the throttle is closed. This will have an adverse effect on idle quality though.

I understand what you are getting at, but are you aware that free wheeling is not good?
Coasting in neutral robs the drive train of the friction needed to safely make a turn.
I believe it’s actually illegal in most states because it is not safe.
The only time you should have the transmission in neutral is during a shift or at a stop.

You can probably defeat the sudden injector cutoff.
Using a diode in series with the TPS signal to the ECU, plus a parallel resistor across the diode, then followed by a capacitor.
This will delay the falling voltage signal from the TPS when you close the throttle.
Fiddling with resistor and capacitor values might give anything from 1 to 5 seconds delay before cutoff.

The TPS will need adjusting to give the correct throttle closed signal to the ECU.

Smart idea, but I think that with only 5 seconds delay it will not completely eliminate the problem.
There will be cases, like in a prolonged downhill, injectors cutoff after 5sec delay, you press the clutch and engine will stall…

Maybe a switch it the clutch pedal that when depressed will give a momentary signal to the TPS (something like 1V for 1second, easy with a couple of transistors) making the ECU think it’s not on idle anymore and fire back the injectors with the hope that the engine will have more time to recover ? Just a thought… there might be other implications when the engine is already on idle?


The ecu may also know the engine is at idle via the vacuum sensor it has connected, by hose, to the crossover pipe.

A proper bypass would deal with both signals, TPS and Vac.

Better yet call Roger.

Did I miss something ?
I thought the original complaint was about pressing the clutch pedal when driving with the engine over 1200RPM ( the injector cutoff limit ) and throttle closed.
The action of pressing the pedal decouples the engine from the drivetrain and the engine stalls before the injectors fire up again as the RPM drops below 1200RPM.
I assumed the driver just pressed the pedal and closed the throttle for a second to change gear, not to cruise along with the pedal pressed or gearbox in neutral.
Are we on the right track ?

If we are on the right track delaying the ECU from seeing the throttle closed signal from the TPS for second or more should fix the problem.

Roger has a file describing the somewhat complicated issue of the over run valve which might also be an issue.

Further to Roger’s file on over run valves, I do have a noticeable jerk on opening the throttle on the 5 speed coupe under some conditions.
The Motec EFI system used in the coupe does have a variable setting for extra fuel added when the injectors resume after the cut off and/or over run condition. This is supposed to help reliably fire the mixture on resumption. Maybe this EFI function needs a bit of fiddling to get just right.

I think that a tricky situation would be when you would depress the clutch after cruising downhill, gear engaged, throttle off, above 1.200 RPM, and for more that a second or two so the delay would be off.

Best regards,

“My mind is on 5 speeds too. The results for Bywater ECU are guaranteed but not cheap! For those of us with less money and more time”

Trying to solve this problem is a bit above my pay grade. Fortunately (I don’t know what it costs) some amount of money will solve this problem with relative ease but, what happens somewhere down the road (hopefully a really long way down the road) one of us or maybe some random dude is doing a 5 speed swap and Roger is not around to solve this problem. If Bywater’s know how is the only knowledge base that keeps these type of upgrades moving forward then the future looks pretty bleak for the rest of us as well as those to follow.

I do hope that someone is able to find a solution to your problem.

I suppose at that point we’d have to move on to a Megasquirt fix.

Hopefully Roger has documented his mods and has some plan for their future.

If not, then it’s reverse engineering on a modded ECU.

You have a point there Aristides, that scenario would negate the TPS delay fix.
I do not see how Roger reworking the ECU would overcome that unless he disabled the injector cut off.

That’s exactly what he does – but I don’t believe it’s to address a stalling issue. He disables the overrun cutoff to prevent “shunting”. Because with a manual transmission you can actually be engaged and driving at 1500 rpm where the fuelling resumes, you can get into a situation where the fuel is kicking on and off while you’re driving. Basically the same thing as the idle instability/cycling that results in a regular car if there’s a vacuum leak that raises the idle to 2000, but it’s in gear rather than at idle.

That overrun cutoff is supposed to resume fuelling at something like 1500 rpm in order to transition smoothly into an idle. Seems to me there must be something else malfunctioning if the engine is stalling as described.

I thought I’d get my two cents in. When I converted my 91 over stalling between shifts was the normal PITA problem. I sent the ECU over to AJ6. Problem solved . Around $400 with S&H. Your right , not cheap, but should be figured in before, on the change over cost budget.

I just had my 1988 XJS coupe upgraded to the 5 speed manual (Doug Nash-Richmond 5 speed Street) and to the later 6.0 liter V-12. Same problem everyone else has discussed, but needed it to be ready for a five day southwestern car rally, just concluded. Made the temporary adjustments that resolved the idle issue, but still stalled out too often. The ECU will now be sent to Roger at AJ6 for the necessary mods. That is the only real solution to the manual trans conversion issue, and needs to be built in to the budget for any such conversion.

Hi Larry, thank you for the advice. I talked with AJ6 and they said they could send me a patch to apply. They did some reprogramming on my computer board a couple of years ago and so were very helpful.