Another page on these @#%!!!!! 996 eproms. Some more information on why, as I think I got a little carried away on the first.

The two problems with them are:

1. Misfire on roll on @ 3,500 RPM

2. "Steppiness" at low speed

The misfire on roll on does not effect all bikes. Its cause is quite elusive though, as I have seen it with almost every eprom available, including standard. I have never been able to find any real problem with bikes that have it, but if a bike does it with the std eprom, you might as well just start banging your head against the wall. The warranty guys have said I should try changing all sorts of things with a known good bike, but as they don’t pay for any problem solving ( just the final fix ) I haven’t got too far. I am confident it has nothing to do with cam settings, or the inaccuracy of. I feel it is just an extreme combination of mass production variances, but I could always be wrong. I have been recently told that swapping pretty much everything from a good bike will not solve the problem.

A thought about this came to me one day, though, and I haven’t been able to come to any real answer yet. The first 851 of 1988 came with dual injectors on both Strada and Superbike models, using the Weber P7 ECU. I’ve never had much to do with these models ( Tricolour ), but all the dual injector bikes except 996 Strada and 996S ( the only dual injector, 16M ECU engines ) have used a staged single/dual injector operation. This staging runs one injector for low injector opening times, and then brings the second in at a preset level. This is just the way it’s always been, but I just started wondering why.

Why would Weber go to all the trouble of writing this into the software? Personally, I think they knew then about this problem and found a way around it, if not the actual cause. The reason may have got lost over time though, and the engineer in charge of the 996 development just couldn’t see the point, so they just ran two injectors all the time. Maybe, maybe not. Duane tells me that the later SP and SPS models ran both nearly all the time, so my conspiracy theory may be wrong.

Running the road bikes on a single injector is not a real issue anyway. The fuel maps in a Weber ECU have a maximum value of 17 milliseconds. Any number higher than that is simply read as 17, and used accordingly. In a single injector 16M ECU, the number is read from the map, modified by the environmental trims and then sent to the injector driver, opening one injector. In the dual injector 16M ECU, the number is read from the map, modified by the environmental trims, halved by the software and then sent to the injector drivers, opening both injectors.

The real benefit of this is when you need 17 milliseconds of injector duration above 7,000 RPM. From 7,000 RPM on, the time for 2 complete revolutions is less than 17 milliseconds. At 12,000 RPM, there is only 10 milliseconds available. With 2 injectors running up to 8.5 milliseconds each, most combinations short of Superbike are covered. On a street bike, single injector with a realistic rev limit is just fine, as the maximum fuel requirement comes at max torque, not max power. The fuel requirement per cycle usually drops as the revs rise beyond 9,000 or so. And you can raise the fuel pressure if required. Raising the fuel pressure to 4.5 bar from the std 3 bar will give an overall increase of 22.5% to fuel delivery

One of the wacky things of this is that running an original eprom with software from a single injector 16M ECU ( say a 916 ) in a dual injector 16M ECU bike will make it run with twice as much fuel as intended, up to a maximum of 17 milliseconds on both injectors. Which is what I thought the original motorvation for running dual injectors was. Alternatively, if you disconnect one of the injectors while using the single injector software eprom, it will run just fine.

This has been proven by the fact that you can run a 996 quite happily using a 916 eprom with one injector disconnected. The most appropriate seems to be the SENNA eprom, so called as it was supplied to suit the Termi carbon mufflers in the 916 SENNA kit, and also available as a spare part. This eprom is lean for a 996, about 15% at the top end, but for many close enough is good enough. Funnily enough, it is still rich in the cruise area, even on a 996.

It seems the people at Ducati Performance heard how appropriate close enough is. The 996 POWER KITS, as the box containing the Termi mufflers, an eprom and a cleaning cloth are called, were supplied with an eprom during 2001 that was a direct copy of the SENNA eprom, apart from a raised rev limit. The problem with that, of course, is that with both injectors opening for the time required for a single, the bikes tend to run a little like a fuel pump. Why the people at Ducati Performance sold them as a dual injector eprom for the 996 is something they only know.

How did we find this out, you ask. Simple, really. We sent one to Duane at FIM and asked "what’s wrong with this eprom ". Ducati, in the usual state of denial, tried to fob the problem off onto DHL, claiming their scanning equipment had damaged the eproms. Needless to say, it did nothing to improve my rather low opinion of Ducati Performance eproms, although I am assured they may be getting better. I must say the DP 996S eprom I’ve been toying with this week has been ok.

Anyway, enough bashing of the official product. I’m sure someone will really go me for that.

Onto the second problem. Hmmmm, more official product bashing. The steppiness at low speed is caused by something most wacky in the 996 software. A little background first. All fuel injection systems have modifying algorithms that change the injector pulse duration based on various environmental trims. One of these is the rate of change of throttle position. For those used to carburettors, think accelerator pump. This simply richens the mixture on rapid throttle opening.

Now back to the 996. The software used in the 996 is unique to it and the following 996S. Duane has known there is something weird about this all along, but exactly what was the mystery. Anyway, one day at work I got a call from a rather excited Duane. He had been playing with the 996 software looking for something else altogether when he found the wacky thing. When he opened the throttle on his simulator box quickly, the fuel didn’t increase above its map value. What did increase was the ignition advance, right off the scale.

Why? Why indeed. I can only imagine that someone writing the software at Weber got it wrong, without really knowing it. Given the way wacky spark map in the standard 996S eprom, it’s obvious nobody at Ducati worked it out either. Not that it will worry them, as the 996 is now a dead model series. The new 998, with the Testastretta engine, is using the 5.9M ECU and a single tower style injector like the 748R.

What this software causes is a massive increase in the ignition advance, which then reduces to the map base setting over a period of a second or so. Riding the 996, this is obvious ( with hindsight ) in the way the bike will surge forward on a small throttle opening, then stop it’s gentle acceleration unexpectedly. Opening the throttle a little more will reproduce the sensation. As I have said before, this drives me absolutely nuts. It takes about 15 seconds on a standard eprom 996 to piss me off.

It also manifests itself in another sensation that struck me very early as ‘not like the others’. The ‘others’ are the single injector 16M ECU bikes that came before – 916, 748, ST2. This is a harshness on full throttle acceleration. It varied from bike to bike, but one bike we did the cam timing on bought real complaints from the owner. I could never work out why, but a large increase in ignition timing coinciding with WOT would be a good enough reason. The extra cylinder pressure our cam timing specs generates would exaggerate this. The thing I can’t work out is why they never pinged.

Whether or not this has any effect on the first problem, I have no idea. I don’t believe it does.

The solution to all these problems is quite easy. Duane has just done a 996 single injector eprom based on the 916 SENNA software. No steppiness, no harshness on acceleration, just lovely, smooth drive from zero to hero. I’ve even done a bike with the cams dialed. Beautiful. Pity it has taken us 3 years to get here. At least I can offer some solutions to guys who have had problems from the start.

Of course, it all might go belly up in 6 months or so. Time will tell.

I’m hoping we can run the 996S on a single injector too. Looking at the std fuel map, it will be fine. How much more fuel is required with some Termis fitted will be the deciding factor.

I’m quite convinced the single injector is the key to this problem, although Doug Lofgren has done a 916 based eprom for the 996 running both injectors as per std. He has had good results with it so far, and the guys at BCM are trying it. Will be very interesting to see how it goes.

[Top Of Page]

Home | Blog | Facebook | Service Enquiry | Products | Reports | The Dyno | Disclaimer | Contact Us