Brad The Bike Boy: Marelli 5AM "bad software" issue

Saturday, July 4, 2026

Marelli 5AM "bad software" issue

This may be one of those instances of me making something up to try to explain something I've experienced.

I have a customer who used to have a late 999S with a Termi full system - one of those with the hydroformed section around the later style black swingarm. I serviced it a few times, and one thing that I noticed (but never really thought about too much) was that, after I'd run it on the bench to do the tune and check the thermo fan, then shut it off, checked the chain tension, headlight aim, etc and pushed it outside to go for a ride, it would idle funny once restarted half hot - in the region of 50 - 60 degrees on the dash.

An erratic idle you could hear and feel. But during and after the road test, it'd be fine. Just like it had been on the bench after being started from cold. As such, I never really paid too much attention to it.

Then, some time later, I had a very late 2005 or 2006 model year 749 come in with a weird running issue. It would intermittantly run badly at low speed, idle badly, cut out at idle, that sort of thing. It was fine as an all std bike, with the problem starting after a Ducati Performance slip on kit with ECU was fitted at a local dealer. It had been back to them a few times, with no improvement, and had even had a custom ECU remap - again, with no improvement.

While I was going through it all, trying to find something wrong, at one point I restarted it half hot and noted an erratic idle and thought "have I seen this before?" So I hooked it up to the diagnostic tool to see if there was any weirdness going on, and lo and behold - there was.

Everything was as expected while it sat there sounding odd, except that the ignition advance as displayed was all over the place. From memory, if I was expecting 12 degrees, it was showing 4 to 14 sort of thing and varying rapidly. Just like it sounded.

Then, at some point, which I later detirmined had no relevance to the engine temp, or how long it'd been running, or how long it'd been since the ignition was turned on even, it would suddenly smooth out. Like just clicking your fingers and it was beautiful. And, most importantly from my point of view, the ignition advance became completely stable, with no variation displayed at all. I ran through the scenario several times, and every time it was the same thing.

The problem then was - why? And how to fix it? First I had a look at the ECU fitted - DP # 96513904B, Marelli software # 2218PT70.

Interestingly, the part # 96513904B is listed by Ducati as fitting both 2005 - 2006 749 and 2004 - 2006 749S. Which is, of course, nonsense. The 749S has the 6mm valve stems with tapered collets, bigger valves and most importantly, a lot more camshaft duration. But that's the only 749 5AM DP slip on kit ECU I've found.

Anyway, ECU fitment lunacy aside, none of that really helped. The owner had provided the original ECU, so I read the file out of that and it was Ducati # 28640851M, Marelli software # 2219DS32. Having a look through my 749 5AM files I found I had 28640851G / 2218DS27 as well. Which begged the question, why did they stop with the G and go the M?

Comparing the 96513904B and 28640851G, they appeared to be much the same basic software and locations wise. Whereas the 28640851M had variations, in the typical Ducati / Marelli fashion, where everything is just in a slightly different place. So it was somewhat different at least. I had another 2005 MY 749 in recently and it had the 28640851G ECU originally.

Because the fitted DP ECU had had a custom map done as well, I copied the fuel tables, etc, from that ECU and dropped them over the original 28640851M file to make a "not bad" basis, as well as played with the ignition advance, advance at idle and other stuff like I usually do. I flashed that in and set it up, and it was happy again.

So, I'm curious to know if anyone else has experienced this specific scenario. I'm not asking for everyone who has a badly set up 749 or 999 to pipe in, because I know there's thousands of them out there. But if you've noticed an erratic idle on half hot restart that settles noticeably after some time, I'd be curious to know the specifics - model, year, ECU #.

No comments:

Post a Comment

The 851

The 400SS

I bought this 400SS (genuine 93 model Australian delivery! Woo hoo huh.) after seeing it in Ebay for what was possibly an good price. As ever, good was too good to be true. Badly repaired and rusted out fuel tank, badly repainted side panels, lots of km (83,000 of the bastards), etc. You know, the usual. Purchased in October 2012, as of now (Jan 2014) I've nearly finished a major service. The cam timing is done with some of the Madasl adjustable pullies that have been drilled in 4 different settings for easy dyno testing. And I have some 900 headers to try a 40mm dual exhaust set up on it (ala M400). Not a great deal of progress for 15 months, so it's taking on a long term feel. There's an F1 endurance style seat base (more taper at the rear) and front fairing sitting around, which will require frame mods to fit. It's a bit of a mental sticking point as this time. But hey, it's not like I can devalue it much more.

The Mighty KR1S 250

How it looked the rainy night someone came to take it away. Painted all black when I bought it, at some point later I realised i could rub the black off and have much of the original colour scheme underneath. Although the main fairing was originally from a red KR1 I think, and definitely badly repaired more than once, so it stayed black. And the tail panel shown was bought to replace the one I smashed the only time I crashed it and the last time I rode it at a Sandown track day. A sensational little bike, holed pistons and frequently failing stator windings aside. The first bike I rode on a race track and the first bike I had my knee down on. And a great commuter too I found. It's sometimes nice to have a bike that won't wheelie past the city speed limit in first gear.