Wednesday, May 7, 2014

DTM5: Show Me the Data (With Dyno Video)

by CFlo

If you remember from the last installment of the DTM5 intro saga, I explained how I picked up a highly decent, thoroughly rebuilt and beastily upgraded example of the last hand-built ///M car for pennies on the dollar. There was a pattern - the previous owners dumped money in, drove for a few years, then passed it on to the next guy when it started acting up. The engine tuning was never quite right; high compression and lean air/fuel ratios are a recipe for heartache and an empty bank account. So I got a cheap M-Fünf with some issues. How do we go about fixing it the DT way? As cheaply and efficiently as possible - with data of course, and some help from the pros. It's dyno time!

Lucky for me, right down the 710 freeway in picturesque Wilmington is the go-to shop for testing and tuning in Southern California. Church Automotive Testing has played host to countless gearheads building project cars, dorifto scenesters, legitimate race teams, and automotive "journalists" (I fall mostly into the first category). Church can baseline your car on their Dynapack dynamometer, tune it under load at any condition imaginable, and make sure all your hard-earned bucks and sweat equity don't go poof when your fancy pants high strung engine melts a piston or chucks a valve. Loading the engine at a variety of speeds and conditions while measuring air/fuel ratio from the exhaust is the right way to tune. Measuring output consistently and with the minimum of variables helps immensely in troubleshooting, too. This is how you can be sure you're making the best power possible while steering safely clear of pinging, knocking, detonation, pre-combustion...whatever you want to call it - and the piston melting temperatures, head gasket eating and rod-bending unpredictable "events" that come as consequences.

So because I'd heard that the last guys to touch the car didn't really have a clue, and that the air/fuel ratio was "said to be" going up into never-never-land lean territory, I decided to take the DTM5 down to my friendly local shop for an $80 Baseline session. Shawn Church himself ran the car thrice on his Dynapack, watching air/fuel ratio (AFR), torque, and power while letting the engine rev to 7,000 rpm at full load / wide open throttle. Kind of nerve wracking, when you consider the immediate history of the car prior to my ownership was: molestation by nincompoops, storage at our buddy DWood's parents' house in Colorado for several years, followed by DT Editor-in-Chief Vince having it shipped out to SoCal to either A) sell it to a stranger, B) keep it for himself, or C) pass it on to a friend (he chose C...wisely). Thankfully while under Vince's stewardship the DTM5 was treated to a new 3.0 bar fuel pressure regulator. For those of you who don't speak European, that's the thing that's attached to the fuel rail that's supposed to keep fuel pressure at a steady 43.5 psi. That's the idea, anyways.

See, Vince was operating under the assumption that the mentally challenged mechanics who last molested this beauty were on to something (and begrudgingly, I have to admit that they were) but didn't know how to fix the problem. By boosting the FPR's vacuum reference line with shop air, they increased fuel pressure above its normal ceiling and got AFRs under control - or so they claimed. So maybe the regulator itself was the culprit? Vince made the logical call to go ahead and replace it, and now here I was, tasked with determining if he solved our lean issue. The car sounded mean on the dyno, and power and torque looked promising. The pink print-outs are standard issue except for the color (the shop's printer was running low on blue ink). 313 hp @ 6423 rpm, and 295 lb-ft @ 4562, all as measured at the rear hubs, told me this was indeed a strong S38. Using Church's 25-30 hp rule of thumb for RWD losses on the Dynapack, this engine is making about 340 hp at the flywheel (up from the factory rated 315). Great!

But the AFR curve on the next plot told the rest of the story. Everything was fine until around 3500 rpm where you can see the curve start trending upwards (the higher the number, the leaner the mixture). 12.5 is happy land, but 13.5...14.0...15 point uh-oh...this is going terribly wrong. Around 6400 rpm the curve takes a sharp hook and skyrockets even leaner, very very rapidly. Mr. Church shut down the test from there. His many years of experience told him I should be asking all of the following questions...

  • Are the injectors maxed out at 100% duty cycle, meaning they're too small for this modified S38's airflow requirements and thus need to be replaced with higher-flowing units?
  • Is the fuel pump dying, causing low fuel pressure and inadequate fuel flow?
  • Could it be as simple as a dirty fuel filter, restricting the flow past a certain rate?

These hypotheses all sounded reasonable to me - and I make my living helping professional motorsport teams and engine builders choose the right hard parts so they can reliably make the power they need. Air is more of my specialty (and the S38 didn't seem to have any problem getting enough of the gaseous stuff) so I figured I'd trust an expert when it came to fuel system troubleshooting.

Excruciatingly, I'm going to leave my follow-up and solution to the next post. But only because I'm curious about the collective gearhead wisdom here at DT. So let me know in the comments...what would you do?


Okay, okay, as a reward I will leave you with this dyno video so you can get an idea of the aural pleasures that bellow forth from the posterior of this Bavarian beast.

For further geeking out, see Church's excellent explanation of the Dynapack here: Dyno tech

All of the previous DTM5 and Volvo 242 posts can be found here: Daily Turismo Project Cars


  1. "...Vince was operating under the assumption that the mentally challenged mechanics who last molested this beauty were on to something..." It's quotes like these that keep me coming back day after day. Like a hobo to his favorite trash can.

    1. We are happy to provide the trash to fill your favorite can!

  2. The rule of 'do the simple stuff first' says to replace the fuel filter and the fuel pump, and retest.
    I would also see if I could find specs for the maximum flow of the injectors somewhere on the internet. Has to be there. Or test them, which is quite a straightforward thing. Again, method is on the internet.

    But reading back over the previous story as well, it seems that this thing is still running the original computer, with an aftermarket chip? That is your problem. You are going to spend hours chasing your tail trying to resolve this effectively without a fully adjustable computer in place. I wouldn't build a megasquirt, as someone suggested in a comment on a previous post. Instead, I would get a Link ECU - - because I have used them, and they are straightforward and great.

    1. AW - the original Bosch Motronic DME has been chipped - twice. My bad for not explaining that thoroughly enough. The story is that the 1st chip wasn't conservative enough with ignition timing and caused a gradual head gasket failure. Our friend (and previous DTM5 owner) DWood had the top end rebuilt and a new chip tune installed, but he wasn't able to pass on much info about that chip to me other than it was from Powerchip in the UK. Hence the investigative troubleshooting...

      Thank you for the Link ECU recommendation. My goal with the DTM5 for the moment isn't to keep modifying it but rather to get it fully sorted out as it stands today, and I think the Motronic system hasn't yet outstayed its welcome. Hopefully a full standalone won't be required, but if it is - I'll keep Link in mind.

    2. CFlo -
      My concern is that the motronic, even with the replaced chip, will still have fixed parameters. Because you have custom porting, your head will flow different amounts than other heads at different rpms. The parameters of the motronic and chip are unlikely to ever be optimal. You can probably get it 'near enough' using different injectors and fuel pressures, but is is a very mechanical way to deal with problems with maps. And the mixture and ignition will always be wrong at some place in the rpm range.

      By all means, replace the fuel filter and the fuel pump with higher flow units and retest. But after that, a new ECU is the quickest way to optimal running. And potentially not as expensive as you might imagine. If you can get a second hand Link, it could be very affordable.

      I have a friend who has spent three years on and off, fooling with a modified Alfa V6, trying to get it to run properly on the standard computer by changing injectors and pressures and messing around. Still doesn't run like it should.

      Good luck.

  3. Downdraft Webers would look sweet.


Commenting Commandments:
I. Thou Shalt Not write anything your mother would not appreciate reading.
II. Thou Shalt Not post as anonymous unless you are posting from mobile and have technical issues. Use name/url when posting and pick something Urazmus B Jokin, Ben Dover. Sir Edmund Hillary Clint don't matter. Just pick a nom de plume and stick with it.
III. Honor thy own links by using <a href ="http://www.linkgoeshere"> description of your link </a>
IV. Remember the formatting tricks <i>italics</i> and <b> bold </b>
V. Thou Shalt Not commit spam.
VI. To embed images: use [image src="" width="400px"/]. Limit images to no wider than 400 pixels in width. No more than one image per comment please.