New turbo for old wagon - Page 2 - PeachParts Mercedes-Benz Forum

duxthe1 · #16 11-24-2018, 05:00 PM

I'm using the later 603 manifold. I designed an adaptor set that bolts to the head and then allows the manifold to bolt up. There's an old thread somewhere here on the original build. The extra space the adaptors use was sorely missed getting the bigger turbine mounted.

I've got 40 lb/hr injectors, well... at the higher pressure of the MB regulator the math says they should flow 40 lb/hr. At 3 bar where they're rated they were somewhere in the 30's.... 32, 36 maybe. I don't quite remember.

I feed them with a peirburg 800 pump that I run in two stages. First stage runs through a ballast resistor, and as boost comes up a relay bypasses the resistor. Had to do it that way because the 800 was overwhelming the pressure regulator at idle. Chased my tail a fair bit on that problem.

So far I'm having trouble spooling the new turbo. I know my manifold leaks a little by the number 4 runner, but it was never an issue with the old small turbo. I'm also thinking there is more than just that. I plan to pull the diverter valve for the air injection, inspect it, and weld a block off. I never measured it, but I know the old turbo had a ton of backpressure so it wouldn't be a surprise for the diverter to be burned up and bypassing. I was half planning to pull the manifold and address the warpage during the install but it became obvious that everything else was going to be a fight so I left it for the time being.

It's a bit frustrating / disappointing because the .63 turbine should spool pretty damn good on this engine. Looking at datalogs, I can see a small momentary drop in boost between 2-3 psi that shows up consistently, so I don't think it is noise on the signal. I think its a tell tale of an exhaust leak costing me energy to spin the turbine. Given how poorly the downpipe is sealing on the exhaust system at the moment, no other exhaust leak will be detectable until its addressed.

My other suspicion is that the wastegate isn't holding as it should. I didn't buy the cheapest one, but I didn't buy an OE Garrett either. I need to rig a gauge and pressure source and watch it closely to make sure its doing what its supposed to.

So far I've managed to hit 10 PSI. It's pretty pissed off. I've also noticed that torque off idle up to 2K is really good now that it isn't choked with backpressure. Once I get it sorted its going to be a beast.

duxthe1 · #17 11-29-2018, 11:26 PM

I've been working at the details over the last several evenings. I got around to re-welding the downpipe v-band flange. Post turbo leak won't affect the spool but its good to have the biggest exhasut leak fixed. I pulled the diverter and welded it shut. The old oring sealing it to the head was poor so that got replaced too. I've got an egr valve, for show for emissions, but its plumbing had a small leak too.

While fixing exhaust leaks I decided to make an adaptor to pressurize the entire intake to check for boost leaks. Found and fixed a few small things, but nothing was a smoking gun for the slow spool.

This just leaves the exhaust manifold itself. It is warped by the #4 runner and it has a crack between 2 and 3. I priced a new one today, it isn't as bad as I was suspecting, but its significant and the holidays are impending. I didn't check availability when pricing the part. Hopefully they are actually still available.

So after getting all of the easily fixable leaks remedied and playing with the boost control solenoid settings I have hit 13.5 PSI. At that pressure, its healthy. Once boost control kicks in, it can't hold. Too much pre-turbo leak and as it tries to modulate the wastegate it loses its head of steam and falls back to wastegate pressure ~8-9 PSI. Even then it aint slow, but I feel there is still plenty being left on the table.

atypicalguy · #18 11-30-2018, 03:44 AM

Quote:

Originally Posted by duxthe1

I pulled an all nighter Monday, after work and finally in the wee hours, it was alive. Every. Damned. Thing. ... was a fight.

Here's a tip. The Mercedes 5 bolt turbine housing DOES NOT MATCH the common Ford 5 bolt housing. (Which BTW is the only version commonly sold) First hurdle, bolt my wastegate housing to a turbo that has 3 out of 5 bolt holes in the wrong position and 30% of the flange not even matching up.

Undaunted, I pressed forward. About 1/4lb of mig wire later, I've added enough meat to the wastegate housing to move the bolt holes and seal the perimeter. My 12" disc grinder was invaluable in that endeavor. Smoothed with a burr, the flow through the modded housing looks pretty good. Tiny sharp shards of steel now everywhere, but mainly in my hands and fingers.

Alright I can bolt the turbo together. Yay! In my 4matic engine bay the .48 turbine housing was a really tight fit. Trial fit the .63 housing, damn. Needs a lot more room. Numerous pie cuts, hammering, and welding (swearing, bleeding, and moments of self doubt) and I can finally lower the engine onto the mount with minimal clearance between the body and turbine housing.

The 4matic hydraulic lines run through the rear corner and one had previously rubbed through on the downpipe. Now its even tighter so I decided to re-bend the 3 lines to run straight down to the motor mount and then straight back. This actually ends up giving more room than I had with the downpipe on the .48 housing. Too easy? Yes. The compression coupling used to fix the previous damage decided to leak. Tightening it revealed that one of the nuts was split and wouldn't tighten. At 2AM no parts to be had, bodge a fix until a new coupler can be acquired.

Time to actually start assembling the turbo onto the car. Lets bolt up the drain flange/adaptor. Hmmmm, the bolts aren't tightening. Oh, they're one thread larger than the feed, or any other Garrett turbo I've ever futzed with. Timeserts were the only fix I could manage and even then I wish the larger threads had left more meat to thread the insert into. At least its the drain. The oil feed threads were correct and match my oil feed line. Can't explain why I still had to tap them when the original bolts started to seize before tight.

Downpipe to exhaust system has a 3" band clamp. It's moved far enough that it won't quite seal despite a nearby flex joint. It is easily the biggest exhaust leak. F*** it! Its far enough under the body that it'll be ok for now. I'll probably get another flange and just drop it off at the exhaust shop.... eventually.

Wastegate. New actuator was expected to be customized to the install. Only by this point I've emptied the spool of mig wire and its after midnight. Make do I guess and get creative. Cut, thread, grind. Eventually it mounts. It does need a boost source to operate the wastegate, only the T04B housing has no bung to thread the nipple into. Not wanting to drill a hole into the thin part of the casting to try and thread the nipple into, I looked to the steel intermediate pipe. Could weld a nut and drill and pipe thread it. No wait, out of mig wire. I have a flange welded to mount an EGR valve, (only for show for emissions) and that weld is the only meat to drill into on this half of the engine's cold side plumbing. Get to drilling, threading, it ain't pretty but the wastegate boost source is done.

The rest of the heat shield and air filter and plumbing went back without too much fuss. My K&N needs cleaned and re-oiled pretty bad. Wasn't happening at 2:30 in the morning. They say specifically not to clean them with compressed air. F*** that too, just blow it out.

With everything all together and running without anything worse than a hyd fluid drip, I loaded the new tune I prepped for it. Pretty close, at least at idle. Start autotuning the VE table on the drive home. Noticed that it only boosts 5-6 PSI. Gonna have to go in and put more pretension on the wastegate. Damn it. It was a fight just to get it where its at. When I go back in and shorten the rod it won't be under the weight of the rest of the project.

So that's where its at. Needs more wastegate spring tension to get to my 7PSI baseline. Needs some tuning, and exhaust pipe work. Need some sleep. Bad.

Not taking no for an answer. I like it.

Sent from my SM-G955U using Tapatalk

duxthe1 · #19 12-03-2018, 08:07 PM

Decided to bite the bullet and order in a new factory manifold. Turns out its no longer available. Damnit.

Next best option is to find a decent used part. Last option is to pull mine and repair it as best possible. I have to pass state emissions by the end of the month so I'm going to focus on getting a solid emissions tune while the search continues.

Over the weekend I flashed the latest firmware into the Megasquirt. I've raised my max RPM with the new turbo but I'm hitting a hard cut at 5100 RPM. Hoping it was firmware related because I was running an old beta version. Turns out it isn't firmware related and I can't find anything in the settings to be causing it. I doubt that my crank position signal is dropping out this consistently. May need to look at it to rule it out if nothing else.

sm. pelle · #20 12-04-2018, 12:48 AM

Boost cant be spiking?
About manifold u haven't considered doing yourself? I'm right now between doing myself log style or using diesel manifold, last one is laying in garage so it's tempting.

duxthe1 · #21 12-04-2018, 01:38 AM

Not a boost spike. I'm barely reaching my target with the exhaust leak, overboost still has a lot of headroom. I've never ran more than 5K rpm since using the megasquirt until now so I'm thinking my signal conditioning may just need to be tweaked to handle 6K +. The 60-2 flywheel signal is surely exercising the circuitry. Some composite logging at high rpm should point me in the right direction. I hope. I haven't messed with it since the initial setup because its always been rock solid.

I like the easy quotient of the factory turbodiesel manifold, but that appears to be evaporating quickly. My adaptor set already includes stainless flanges to start from if I do decide to fab my own. Gonna focus on emissions first. The inspectors really don't know what they're looking at but the factory part is certainly less conspicuous than some nest of welded stainless.

sm. pelle · #22 12-04-2018, 01:59 AM

I have 36-1 wheel and haven't got any issues, almost

. Limiter is 7500.

Few times fuse on over voltage relay have been blown when I have hitted limiter, but it's so rarely that I haven't got motivation to look into that. That can't be your issue, car would be dark, but maybe worth looking into that area.

97 SL320 · #23 12-04-2018, 09:28 AM

Have you looked at the crank sensor voltage / scope trace at high RPM? Are you using the stock inductive sensor? Inductive sensors produce more voltage at high RPM and perhaps you are over driving the input.

Spacing the sensor out a bit would reduce voltage but the risk is losing steady signal during cranking.

As for the 60-2 and 36-1 wheels / RPM. The 36-1 should produce less voltage ( ? ) over the 60-2 at the same RPM giving the 36 more head room.

duxthe1 · #24 12-04-2018, 08:16 PM

I'm not using the stock 103 crank sensor, but a later model MB crank sensor. The stock sensor only reads 3 large teeth per rev and I didn't trust it to read 58 small teeth per rev. So I used a later model sensor that originally read a 60-2 wheel. They fit the hole and line up the screw but have to be spaced out about 4mm to get the air gap set.

I hadn't scoped it since it just worked from day one. Tuner Studio (the tuning software) has really helpful high speed logging tools so I fired them up and got a look see at what was going on. The trigger log showed the classic 2nd smaller trigger spike indicative of crank sensor polarity wrong. I opened up my harness and swapped the leads for the crank sensor and voila, she'll eagerly rev to the 6200 limit. Too easy? Yes. It wont idle below 1000.

Had to dig the MS out of the dash and open it up and reset the VR conditioner potentiometers to the basic position, all the way clockwise. Now it idles and revs no problem-o. Its kinda funny that I've run a 5000 RPM rev limit all these years with it wrong when it wouldn't have revved past 5100 anyway.

97 SL320 · #25 12-04-2018, 11:20 PM

Quote:

Originally Posted by duxthe1

The trigger log showed the classic 2nd smaller trigger spike indicative of crank sensor polarity wrong.

I've never run across this given I don't scope a sensor at high RPM. I'm betting when this second blip gets a high enough voltage and crosses a sensing / noise filter, the ECU thinks the engine is now 2X RPM. This would explain the consistent RPM cut off.

When a CW distributor gets a CCW pickup coil, the sine wave is inverted , timing is altered and sparks occur when the rotor is out of phase. Did you recheck your ignition timing with a light? I'd expect it to have moved from the original build.

duxthe1 · #26 12-05-2018, 12:36 AM

Yeah, timing was altered. I've been running my missing tooth at 55 degrees since I built the MS. With the polarity corrected and the timing locked at a fixed value, I had to change my missing tooth angle to 60 degrees to get the commanded timing to match the timing light.

With this problem behind me and emissions in front of me I decided to turn table switching back on. I can set up an additional 16x16 volumetric efficiency table and switch to it based on various parameters. I set mine up to switch based on manifold pressure giving me a map for on boost and a map for off boost essentially making it a 32x32 table. The big emissions benefit is having good resolution in heavy vacuum where decel transients tend to be very rich. Having smaller cell resolution (now 5kpa x 300 rpm) across the whole map doesn't hurt either.

97 SL320 · #27 12-06-2018, 11:59 AM

Thanks for the update. What is causing the second blip? ( other than reverse polarity. )

I'd expect there to be a blip on both sides of the 0 volt trace. Does your SW display both sides of the 0 V line? If so, did the blip move below 0V or go away?

duxthe1 · #28 12-06-2018, 10:05 PM

Honestly I don't know. The high speed logger I used was a post filter tooth logger and not a raw data logger. The first tooth after the missing teeth shows up as nearly twice the amplitude of the others. A second somewhat shorter tooth accompanying it is the telltale of reverse polarity. Corrected there is only the one larger amplitude tooth.

duxthe1 · #29 12-28-2018, 12:45 AM

Many thanks to user 97 SL320, the unicorn manifold has been found, pulled, and shipped.

It came today so prep began for its installation in the near future.

As you can see in the first pic there is a very rough casting flash between the 1-2 and 2-3 runners. Given that this is where the current manifold is cracked the first step was to grind out the flashing hopefully eliminating a point where a crack can start. It isn't exquisitely detailed, just rounded and smoothed in the corner.

The next step was checking the flange for flatness. It could stand some significant improvement so I spent about an hour between a flat file, 12" disc grinder and a sandpaper flat plane. I went from 0.017" warpage down to 0.005". I'd like to have perfect but ended up chasing it and determined that 0.005" was as good as I can get it.

After bead blasting the remaining sealing surfaces, I traced the T3 gasket onto the manifold. My every instinct says port match that sucker because there is a fair bit meat in the way. My inner coward is afraid of releasing some internal stresses removing that much material opening the door to cracks in this otherwise perfect piece. I think I'll wait until I unbolt the turbo to see how this gasket lines up on the turbo before grinding anything.

duxthe1 · #30 12-28-2018, 10:15 PM

Since it was a slow day at work I managed to get the manifold installed. I also installed a new motor mount for the right side. My previous issue of turbine clearance was mostly a severely collapsed mount issue. With all of the spacers removed and the new mount installed it sits about 1/2" proud of its previous location. Should have done that a long time ago.

Removal of the manifold was not a big deal but a manifold stud broke. To my surprise it drilled quite easily and didn't take a ton of fight to get it out. The main crack was worse than I thought and had separated about 50% around the runner. Another smaller crack in the turbo mount flange wasn't nearly as bad.

The big find was the manifold gasket was cracked in two for the number four runner. I know that cylinder / runner had sealing issues but I didn't see that coming. Given that the manifold has a coupler between cyls 4 and 5, I speculated that the split was caused by heat related movement of that coupler and the rear section of manifold. In an attempt to prevent a re-occurrence I cut the gasket between 4 and 5.

With the turbo out I pulled the wastegate housing off of the turbine housing to doublecheck the wastegate puck's location relative to the opening in the housing. Having been run a bit there was a visible witness mark showing that the puck did fully cover the opening but was offset so that one side only had 1-2mm to spare. I reamed out the bolt holes to move the wastegate housing down about 2-3mm centering the puck. While I was at it there was also a nice outline of how the two housing mated and I welded in some more material where the contact was sparse.

I reassembled everything with a generous coat of Permatex Ultra Copper RTV. The packaging actually specifies use on exhaust manifolds and turbo flanges, so full speed ahead. What a pleasure to drive it without the persistent smell of pre-catalyzed exhaust.

The spool is definitely improved. Not much datalogging so far but at least a 500 RPM improvement in spool. Potentially more with tuning. It has no problem hitting the overboost cutout (temporarily set a 15 PSI) and needs some tuning of the boost control to get it's behavior under control.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode