OM603 semi-super pump testing, fitted with Bosch 6.0mm elements
OM603 semi-super pump testing, fitted with Bosch 6.0mm elements
Prior to the pump build & swap, I did dyno runs with a stock rebuilt injection pump, which show the stock fuel cutoff hits sharply around 4750rpm, drastically limiting power above that point. I also did a dyno run with a different maxed-out stock pump, and did testing from 0-100mph with both configurations. Unfortunately the 0-100mph testing data is slightly skewed because in the meantime, I installed LSD and larger brakes on the car, which take a bit of power to spin due to the increased rotational weight. So, the test data shows minimal performance gain between the stock pump turned up 0.5 turns vs 2.0 turns. However this isn't quite true, it did make more power since the car was slightly quicker despite the weight penalty. I'll add the dyno & performance data later, I need to compile it into an easier-to-read format.
Anyway: When maxing out the stock pump, both Casey and myself found that beyond +1.5 turns there were driveability issues, mostly that the engine would not always return to idle, and it was hard to start. So although I did test a +2.0 turns, I went back to approx +1.5 turns as the most which retained normal starting / idling. There was very little, if any, difference in power between +1.5 and +2.0 turns. Although I don't have an intercooler on the car, I do have IAT, EGT, and boost gauges. With the stock pump maxed out, EGT's would reach about 1300°F by approx 100mph (approx 30 seconds of WOT). Turbo is stock KKK, exhaust is stock but the oxidation cat fell off and resonator (middle muffler) was replaced with a straight pipe. Injectors have new #314 nozzles and are recent / balanced / etc.
I got the Bosch spec sheets for the 603.960 and 606.962 pumps, and while it's mostly in ancient Greek and Egyptian hieroglyphics, I did notice some oddities: I found it interesting that the fuel quantity wasn't linear between the two pumps. Spec sheets say ~51cc for 110kw on the 603 (0.464 cc per kw), while the 606 shows 64cc for 130kw (0.492 cc per kw). I don't get why 606 specs more fuel per kw/hp. In general the sheets are fascinating, the 603 sheet appears to indicate that 34cc of fuel is delivered at 2000rpm with no boost, increasing to ~51cc of fuel at 3200rpm with full boost, but that it drops to ~49cc at 4400rpm with full boost! The 606 pump makes more sense... 55cc @ 1000rpm, 60cc @ 2000rpm, and 64cc @ 4400rpm (no boost reference since it's all electronic). I can't figure why the 603 pump spec actually seems to reduce fuel delivery before the peak power RPM. Also of interest is that the 603 pump idle spec is ~6cc, but the 606 idle spec is nearly double that at 12cc. I suspect this could be due to camshaft profile differences, but still, twice the quantity at idle? Strange. The hybrid pump dialed in a ~79cc with 16.5mm of rack, spread was 1.6cc. The ALDA would pull about 25cc with no boost signal. Idle was 6.1cc (in spec), spread of 0.7cc. On paper, it looked great.
Installing the hybrid pump was straightforward, I set IP timing slightly advanced from spec (about 13.5° ATDC via RIV method, spec is 14°). The good news is, the pump starts and idles like stock - absolutely zero issues with returning to idle, or starting (cold or hot). Initial test drives showed much more power with much less throttle travel, which was expected. I had to turn up the transmission vac modulator 2 full turns to get good part-throttle shifts. Part throttle upshifts are way too early, but that's a separate hassle (if I adjust the Bowden cable to make the shifts correct, it causes other problems - more on that later, probably in a separate thread). Boost builds FAST, I mean it would peg the boost gauge to 15psi in a couple of seconds at ~2/3 throttle. There was little to no smoke except at WOT at higher RPM, then it would smoke noticeably. Boost peaked at 15psi, maybe a bit more.
Smoke = lack of air. I tried turning up the wastegate, which is adjustable on the KKK. At first I did 3 turns on the setscrew. But the car made less power, and less part-throttle boost. I mean a LOT less. At 2/3 part throttle, instead of shooting to 15psi like previously, it would slowly climb to 15-18psi. The change was obvious. Peak boost did increase (I saw spikes to 23psi at upshift time) but the part throttle loss was awful. The KKK wastegate operates much differently than the usual Garrett, it's hard to explain, but I think this is what caused the odd behavior. At any rate, I ended up with a final setting of +2.0 turns on the wastegate, this provided more boost than stock, without the loss of part-throttle power. Unfortunately, it's still not enough air. Although peak boost is ~20psi or so, as RPM's climb under load you can clearly see the boost drop off. It drops from 15-18psi back down to about 12-13psi near redline in lower gears, and from 20psi down to 15psi near redline in 3rd gear. Again, I think it's the funky KKK wastegate design, it's not happy operating outside the design specs.
Here are videos of the gauges during WOT runs. The first video (9MB) was taken prior to the final wastegate adjustment and shows slightly lower boost levels. The largest video (38MB) is the best, if you have the time to download it, and shows the current boost level:
http://www.w124performance.com/movies/Me...1987_300D/
I can't test past 80mph at the moment due to high EGT's... as you can see in the video, they hit a bit over 1400F by 80-85mph. I'm not comfortable with those temps. I need to get more air in there. My options are:
1) Swap in a 3.5L 0.55-trim Garrett turbo, which I have on the shelf. Hopefully the Garrett wastegate won't cause the weird loss of boost like the KKK. This should (in theory) be a straight swap with minimal mods.
2) Add a small side-mount intercooler. I have a 300ZX IC that should work perfectly, although it's rather small. But it should help at least a little. I can't add a large front-mount unit like Casey did without cutting the bumper (or swapping on an AMG bumper). That is in the long-term plans, but I'm not going to do that anytime soon.
Summary of acceleration times:
0-60 = 10.7 sec, 0-80 = 19.5 sec (stock IP)
0-60 = 9.3 sec, 0-80 = 16.3 sec (maxed-out stock IP)
0-60 = 8.6 sec, 0-80 = 14.9 sec (hybrid IP, but smoking = lack of air)
Comments? Any ideas on how to get EGT's down without headers, big IC's, water injection (already tried that - waste of time), or giant aftermarket turbos?
Additional performance data & dyno graphs will be added eventually - stay tuned.
Dave M.
Boise, ID, USA
1997 E420 - 149kmi (Bugeyes)
1994 E420 - 136kmi (Blondie)
1994 E500 - 116kmi (Q-ship)
1992 500E - 179kmi (Mach 5)
1987 300D - 330kmi (Sportline Stage 2)
Click here for my website!
Several years ago I got the bright idea to try installing the Bosch 6.0mm elements from the 606.962 turbo IP, into my 603.960 IP. In theory, this should have provided about the same power as a "chipped" W210 E300 turbo, or 210-220hp at the crank (170-180hp at the wheels). I picked up a used 606 pump as a donor for the elements and also a spare 603 pump to use as a core. I'll spare the long story as to what transpired between then and now (several years, lots of waiting) but I finally received a completed, working 603 pump with 6mm elements in September 2010. This was largely due to the generous assistance of an MB owner/enthusiast in WA who happened to have a friend that works at a Bosch shop. (I will update this post with the name & contact info for the shop, after confirming that they are willing to have this information posted publicly.)
Prior to the pump build & swap, I did dyno runs with a stock rebuilt injection pump, which show the stock fuel cutoff hits sharply around 4750rpm, drastically limiting power above that point. I also did a dyno run with a different maxed-out stock pump, and did testing from 0-100mph with both configurations. Unfortunately the 0-100mph testing data is slightly skewed because in the meantime, I installed LSD and larger brakes on the car, which take a bit of power to spin due to the increased rotational weight. So, the test data shows minimal performance gain between the stock pump turned up 0.5 turns vs 2.0 turns. However this isn't quite true, it did make more power since the car was slightly quicker despite the weight penalty. I'll add the dyno & performance data later, I need to compile it into an easier-to-read format.
Anyway: When maxing out the stock pump, both Casey and myself found that beyond +1.5 turns there were driveability issues, mostly that the engine would not always return to idle, and it was hard to start. So although I did test a +2.0 turns, I went back to approx +1.5 turns as the most which retained normal starting / idling. There was very little, if any, difference in power between +1.5 and +2.0 turns. Although I don't have an intercooler on the car, I do have IAT, EGT, and boost gauges. With the stock pump maxed out, EGT's would reach about 1300°F by approx 100mph (approx 30 seconds of WOT). Turbo is stock KKK, exhaust is stock but the oxidation cat fell off and resonator (middle muffler) was replaced with a straight pipe. Injectors have new #314 nozzles and are recent / balanced / etc.
I got the Bosch spec sheets for the 603.960 and 606.962 pumps, and while it's mostly in ancient Greek and Egyptian hieroglyphics, I did notice some oddities: I found it interesting that the fuel quantity wasn't linear between the two pumps. Spec sheets say ~51cc for 110kw on the 603 (0.464 cc per kw), while the 606 shows 64cc for 130kw (0.492 cc per kw). I don't get why 606 specs more fuel per kw/hp. In general the sheets are fascinating, the 603 sheet appears to indicate that 34cc of fuel is delivered at 2000rpm with no boost, increasing to ~51cc of fuel at 3200rpm with full boost, but that it drops to ~49cc at 4400rpm with full boost! The 606 pump makes more sense... 55cc @ 1000rpm, 60cc @ 2000rpm, and 64cc @ 4400rpm (no boost reference since it's all electronic). I can't figure why the 603 pump spec actually seems to reduce fuel delivery before the peak power RPM. Also of interest is that the 603 pump idle spec is ~6cc, but the 606 idle spec is nearly double that at 12cc. I suspect this could be due to camshaft profile differences, but still, twice the quantity at idle? Strange. The hybrid pump dialed in a ~79cc with 16.5mm of rack, spread was 1.6cc. The ALDA would pull about 25cc with no boost signal. Idle was 6.1cc (in spec), spread of 0.7cc. On paper, it looked great.
Installing the hybrid pump was straightforward, I set IP timing slightly advanced from spec (about 13.5° ATDC via RIV method, spec is 14°). The good news is, the pump starts and idles like stock - absolutely zero issues with returning to idle, or starting (cold or hot). Initial test drives showed much more power with much less throttle travel, which was expected. I had to turn up the transmission vac modulator 2 full turns to get good part-throttle shifts. Part throttle upshifts are way too early, but that's a separate hassle (if I adjust the Bowden cable to make the shifts correct, it causes other problems - more on that later, probably in a separate thread). Boost builds FAST, I mean it would peg the boost gauge to 15psi in a couple of seconds at ~2/3 throttle. There was little to no smoke except at WOT at higher RPM, then it would smoke noticeably. Boost peaked at 15psi, maybe a bit more.
Smoke = lack of air. I tried turning up the wastegate, which is adjustable on the KKK. At first I did 3 turns on the setscrew. But the car made less power, and less part-throttle boost. I mean a LOT less. At 2/3 part throttle, instead of shooting to 15psi like previously, it would slowly climb to 15-18psi. The change was obvious. Peak boost did increase (I saw spikes to 23psi at upshift time) but the part throttle loss was awful. The KKK wastegate operates much differently than the usual Garrett, it's hard to explain, but I think this is what caused the odd behavior. At any rate, I ended up with a final setting of +2.0 turns on the wastegate, this provided more boost than stock, without the loss of part-throttle power. Unfortunately, it's still not enough air. Although peak boost is ~20psi or so, as RPM's climb under load you can clearly see the boost drop off. It drops from 15-18psi back down to about 12-13psi near redline in lower gears, and from 20psi down to 15psi near redline in 3rd gear. Again, I think it's the funky KKK wastegate design, it's not happy operating outside the design specs.
Here are videos of the gauges during WOT runs. The first video (9MB) was taken prior to the final wastegate adjustment and shows slightly lower boost levels. The largest video (38MB) is the best, if you have the time to download it, and shows the current boost level:
http://www.w124performance.com/movies/Me...1987_300D/
I can't test past 80mph at the moment due to high EGT's... as you can see in the video, they hit a bit over 1400F by 80-85mph. I'm not comfortable with those temps. I need to get more air in there. My options are:
1) Swap in a 3.5L 0.55-trim Garrett turbo, which I have on the shelf. Hopefully the Garrett wastegate won't cause the weird loss of boost like the KKK. This should (in theory) be a straight swap with minimal mods.
2) Add a small side-mount intercooler. I have a 300ZX IC that should work perfectly, although it's rather small. But it should help at least a little. I can't add a large front-mount unit like Casey did without cutting the bumper (or swapping on an AMG bumper). That is in the long-term plans, but I'm not going to do that anytime soon.
Summary of acceleration times:
0-60 = 10.7 sec, 0-80 = 19.5 sec (stock IP)
0-60 = 9.3 sec, 0-80 = 16.3 sec (maxed-out stock IP)
0-60 = 8.6 sec, 0-80 = 14.9 sec (hybrid IP, but smoking = lack of air)
Comments? Any ideas on how to get EGT's down without headers, big IC's, water injection (already tried that - waste of time), or giant aftermarket turbos?
Additional performance data & dyno graphs will be added eventually - stay tuned.
Dave M.
Boise, ID, USA
1997 E420 - 149kmi (Bugeyes)
1994 E420 - 136kmi (Blondie)
1994 E500 - 116kmi (Q-ship)
1992 500E - 179kmi (Mach 5)
1987 300D - 330kmi (Sportline Stage 2)
Click here for my website!
1 Guest(s)
1 Guest(s)