Chipping an HFM ECU ! - Page 5 - PeachParts Mercedes-Benz Forum

duxthe1 · #61 03-22-2025, 10:51 AM

The value 02 is the hysteresis that turns the cam back off. The value 0 is off but doesn't turn it off once it is on. That was a pretty big eureka moment.

I seem to recall the 3.6 intake cam is slightly more aggressive than the others but I don't have specs off the top of my head. Additionally, the advancer is a few degrees different. In my '90 I use an exhaust cam in the intake position which gains 17 degrees of duration so I use the V8/4cyl advancer that advances over a much smaller window.

duxthe1 · #62 03-22-2025, 06:58 PM

I got the coupe on the dyno today! Overall a really good experience. I baselined on the stock 3.2 mapping. Peak wheel horsepower came in at 204. After 25 pulls it gave a very heat soaked 221hp, a gain of 17hp. The best hp gain was at 4400 rpm at +39hp. The best torque gain was at 2900 with +57 ftlbs.

I did the cam on/off and intake on/off runs and both showed clear points where they crossed to suggest the switchpoint. Both were very close to where the stock map switches them. Looking at the graph, the lower 2 lines are hp. The baseline has two significant dips at 3500 and 4400. The first is from the intake switchover, and the second dip is the cam switchover. The large improvements are from pushing the switchpoints of both several hundred RPM later. The dynamics of either switching pushed them a surprising amount later, compared to where the static runs suggested. If anything I expected to tune them in the other direction to account for the delay of them actually mechanically switching. The best gains under the curve came from optimizing them. Pushing the AFR down to approx 13:1 had the least gain, and adding timing had great torque gains down low but didn't affect the peak much. Inverting the airbox lid was only worth 3hp.

Frank Reiner · #63 03-22-2025, 07:27 PM

Quote:

Originally Posted by duxthe1

The intake camshaft has an advancer that is only on or off, and not continuously variable. It's switchpoint is determined by an 8x8 map based on load and RPM. It is retarded at idle and switches to advanced sooner at higher load and later at lower load. At higher RPM and above 30% load the cam switches back retarded through to redline.

I plan to do dyno runs with the cam retarded the whole pull and another with it advanced the whole pull to determine my own optimal switchpoint. Its probably close to where MB set it, but the hard data will tell the tale.

Quote:

Originally Posted by duxthe1

The value 02 is the hysteresis that turns the cam back off. The value 0 is off but doesn't turn it off once it is on. That was a pretty big eureka moment.

I seem to recall the 3.6 intake cam is slightly more aggressive than the others but I don't have specs off the top of my head. Additionally, the advancer is a few degrees different. In my '90 I use an exhaust cam in the intake position which gains 17 degrees of duration so I use the V8/4cyl advancer that advances over a much smaller window.

Does the value "0" = retarded? And the value "01" = advanced?
Does this statement: "The value 0 is off but doesn't turn it off once it is on.", mean that on startup the cam position is retarded, but once it is moved to the advanced position the cam will stay advanced even when the engine is returned to idle? Or does the value "02" move the cam back to the retarded position?

duxthe1 · #64 03-22-2025, 09:03 PM

The map is drawn so that once the cam is on, 01, it will pass through a turn off, 02, at some point on the way back down in RPM.

nico.w114.c36 · #65 03-30-2025, 04:38 PM

hallo duxthe1 i would like to introduce mijself.i am nev here. i read this topic. i have same c36 engine with hfm ecu from siemens vdo from w210 320 engine bj 1996. i would like to ask you if this program or ic tool is for sale to adjust mij ecu myself. i drive my car with stm6 piggy controller between the air mass sensor that works well too. what you do that better. i have on the dyno +- 290 hp on lpg injectie with vialle lpi system that inject liquid into the engine.sorry if my language mistakes . Greetings nico
from holland.

duxthe1 · #66 03-30-2025, 07:16 PM

Greetings Nico! The eprom chip used in the HFM ECUs are one time programmable. Mercedes never intended to rewrite the eprom so they had no reason to use a more expensive rewritable chip. You have to pull the original chip from the board, write a new chip, and replace it on the board to do any tuning. I only do this on the Bosch version ECU. The VDO version is built such that the eprom is virtually inaccessible and requires a ridiculous amount of very difficult soldering work. The likely result of which is destruction of the circuit board. Hence, I currently only tune the Bosch ECU.

It is possible to do real time tuning using a very early version ECU and an eprom emulator. I used this route to develop the HFM ECU to the point of being able to define the relevant mapping and modify it on the dyno. This is not a system that I can sell and support. It is much simpler and easier for me to offer tuned ECUs outright and / or offer custom tuning in special cases.

nico.w114.c36 · #67 03-31-2025, 02:07 PM

hello dux too bad. l know bosch easy to tune. that been here for a long time.l had the siemens system of m104 lying around and was cheap to buy they at a scrapyard. for bosch they know the prices here. but anyway l have removed chip from my ecu and soldered an adapter for easy replacement.and l put a copy chipin my ecu and l tried to read the number 256 in the entire file. and we can't do anything with that.you say the same too bad for such a nice system. l now use a stm6 from prefectpower. the box is still for sale on the internet but the company no longer exists. if the software searches i have it.and i use that to adjust the air mass meter signal which works well i have increased the voltage +- 0,50volt.and them i use the vialle lpi gas injectie systeem to fine-tune even more power because the octane of gas is 103. you notice that on the dyno

duxthe1 · #68 04-26-2025, 11:29 AM

I've been daily driving the project coupe a lot lately with the later ECU. I'm not able to tune on the fly or data trace the late ECU but it does have a larger suite of live values available through the scan tool. Probably the most important one is the per cylinder ignition timing knock retard. I've been using these values to mainly tune a more aggressive part throttle timing map. What most surprised me is how often the engine is running below the mapped ignition timing due to the knock threshold. Significant portions are mapped well into the knock threshold and the knock sensing is constantly holding it just at the edge of knock. I always used to think of knock sensing as a last ditch approach to prevent damage in case of bad fuel or lean mixture but the HFM uses knock sensing as a significant part of its main ignition timing strategy.

Look at the stock 3.6L WOT ignition timing map, attached. Below 5k RPM the values are quite low. There's big torque available from mapping that more aggressively but look at the massive jump between 5k and 6k RPM. That is an additional 12 degrees! With pump gas at operating temps and above freezing ambient temps, you're not going to actually get those values. It'll end up with at least 4-5 degrees of retard as its revved out.

Quote:

Originally Posted by nico.w114.c36

the octane of gas is 103. you notice that on the dyno

When you look at the stock WOT map it's easy to see why. When I was on the dyno, it was easy to get gains in the lower RPMs. The engine wants more timing and wasn't mapped for it. It was much harder to get gains at high RPM because its already mapped past the knock limit. Adding fuel gains a little knock resistance that will allow a bit more timing up top. That will gain a few peak HP, but at 10.5:1 compression, the motor really wants better fuel. I'm committed to working on an E85 setup for exactly this reason.

Another thing I have been working on, is a website. I am now offering tuning for these HFM ECUs. I can confidently add horsepower and torque over a stock HFM. If anyone wants to check out the site, or contact me about tuning the address is classicMBstreetfighters.com

duxthe1 · #69 05-10-2025, 09:52 PM

Package in the mail today! Larger injectors, soon to be installed and then tuned for E85.

duxthe1 · #70 05-16-2025, 11:37 PM

Having run out the majority of the current tank of fuel, I put the Coupe up on the lift and drained the rest of the fuel. A jerry can's worth of E85 was dumped in. Old injectors were pulled and the new ones installed. They did go right in though not a perfect swap. The originals were keyed and would only clip in with the proper orientation. The replacements were not keyed and had to be aligned manually. The groove for the clip was also a smidge higher on the body but not enough to be an issue. Lastly the groove did not go entirely around the injector. It was cut into both sides and luckily the clip was able to grab enough of it. All back together, a few key cycles to prime the system, and she fires right up and settles into a nice idle. Short trip to the closest gas station to fill the tank and further dilute any residual petrol.

Initial impressions is that it is a tad down on power in lower revs. It has lost the crispness in exhaust note that it has when its running at the knock limit on petrol. The HFM's O2 correction was showing that it was running a bit on the lean side. I swapped in my emulator ECU and added a flat 10% across the fuel maps. O2 corrections are pretty close now. I'll start working on part throttle ignition timing and see if I can get that smidge back with a little on top. I'll have to plan another dyno session soon to dial in the WOT. It already feels pretty good at WOT in the higher revs, as I'm suspecting its able to take advantage of the timing that was being pulled from knock retard on premium petrol.

Frank Reiner · #71 05-17-2025, 06:57 PM

Once above about 50% load, Lambda of .85-.80 may be required with E85.

duxthe1 · #72 05-18-2025, 02:02 PM

I have been targeting .87 but its showing me that it may want more. Even after swapping injectors, I still may not have enough fuel delivery to get that rich at the torque peak.

duxthe1 · #73 06-10-2025, 01:18 AM

I've had a set of 298cc injectors to install for about 2 weeks now but its been a crazy rainy spring so I've put off installing them. My tuning ECU is open to the elements and I'm not keen to drive around with it when its raining. Well, the forecast looked good enough for the next few days so I installed the larger injectors.

The previous pink 260cc injectors ran quite well but were maxxing out sooner than expected. On cold start they took a slight bit longer to fire up, but not terrible. I didn't tune the cranking enrichment for them knowing they would come back out. Will have to see how the 298s do with the cold starts over the next several mornings. The 298s are a better fit in the rail. The clip grooves line up perfect and they even have an alignment tab like the stock injector. Its not quite as wide so there is a smidge of axial wiggle but overall a better fit than the pink 260s.

So far I've just roughed in the fuel tables. I can hit pretty rich (high .7s lambda) WOT AFRs with a bit of headroom now. Its kinda looking like the 298s are less linear than the 260's in that they seem to be leaner at lower pulsewidths. This could possibly be due to a significant difference in injector dead time compared to the stock 200cc injector. Somewhere in the binary file there is a value that describes the injector dead time. Finding and defining a single value is much, much, much harder than digging out the maps.

After I get the part throttle fuel maps sorted, I'll try to get an hr on the dyno to get the WOT dialed. In the meantime I've started working on a setup to run the HFM module on the bench. A fully simulated environment is going to be difficult to pull off, but I think it will pay dividends in what it may allow me to figure out beyond just emulating on a running engine.

duxthe1 · #74 10-21-2025, 01:48 AM

Its been a while since I updated this post. I got the coupe on the dyno again last weekend to get the E85 tune dialed in. I've been driving around all summer with a pretty rich E85 street tune. I have been waiting for similar weather as when I went in the spring and dynoed on gasoline. With the ole gal strapped down, I began leaning out the fuel. Little to no change so rinse and repeat. Over the course of about 10 pulls I leaned it more than a full point of Lambda and only picked up 6 hp. This setup seems to care very little about AFR. The best run of the day ended up being 15hp better at peak than the best pull on gasoline. Above 3500 there was a consistent 10 to 15 hp gain. At torque peak there was only a 4lbft gain but again over 3500 there are again solid gains of 10 to 15 lbft. By 6k RPM the gains are pushing +20 for hp and torque.

Over the summer I was averaging 13.5 mpg on E85 city driving. Compared to gasoline that ends up being right around 1.5 cents per mile cheaper on E85. On a long highway trip, at or near 4k RPM with the low gearing, it pulled down 19.5 MPG. The figures are accounting to the speedo / gearing discrepancy. Now that I'm running leaner after the dyno visit, I am curious what the next tank will average out to.

Frank Reiner · #75 Today, 02:30 PM

Quote:

Originally Posted by duxthe1

Over the course of about 10 pulls I leaned it more than a full point of Lambda and only picked up 6 hp. This setup seems to care very little about AFR. The best run of the day ended up being 15hp better at peak than the best pull on gasoline. Above 3500 there was a consistent 10 to 15 hp gain. At torque peak there was only a 4lbft gain but again over 3500 there are again solid gains of 10 to 15 lbft. By 6k RPM the gains are pushing +20 for hp and torque.

By any chance do you have an EGT (or 2) installed? High percentages of alcohols tend to exhibit elevated EGTs when leaned.

Is the dyno that you used an inertia type or eddy current?