Here's the finished boards for the temperature and pressure module as well. Just need to design a case and print it and should be good to go. Thermocouple amps are MAX31855's, pressure sensors are MPX4250's, and the micro is a Teensy 4.0, for those who might be interested.

 

This project has all sorts of custom fabrication going on, electrical, casting, machining; and then the programming to actually read the data, it's very impressive. Thanks for posting

 

Lol sounds a lot like my project...that's why its never finished. Not casting though. That's some next level ****. I wouldn't trust myself.

 

Well, I dun goofed. I may have spun a main or a rod bearing while testing out the new -16 turbo drain. Pulling the engine in the next few days for some diagnostic works so hopefully things are salvageable.

But, I do have 2 5.3's on the way that will be getting the good ol' ring-gap-and-send-it treatment. I've also sold the Kistler pressure transducer and will be picking up a spark plug pressure transducer in the near future so that it's portable between engines and projects.

 

What made you want to use a -16 turbo drain and how would that make you spin a rod?

 

Long story short, I had different plans when I started this project so I mounted the turbo in a spot that's not great for drainback but was a bit more sturdy than a cantilevered mass hanging off exhaust studs. It's sitting about a foot behind the passenger headlight now. Retrospectively it would have been smarter just to go with a Hooker manifold or something similar and be done with it. With a -10 drain I had oil entering the turbine and compressor housings. Did a complete redo on the drain line and chose -16 because I never wanted to touch it again, lol

But, in testing that the drain was properly flowing, I warmed it up and revved it to about 4500 for 10 ish seconds to see if I could get it to push oil again. When I brought it down to idle there was new tapping coming from the front of the engine. Unfortunately not accessories - I took the belt off and checked again, still there. Hoping that I collapsed a lifter or something simple, but it seems to be ticking at crank speed and not valvetrain speed. Will have it pulled shortly and will start diagnosis.

 

I know this project was born on the idea of reinventing the wheel; however tens of thousands of LS engines have the turbo hanging off the exhaust studs and they seem to be alright

Probably wasn't worth the effort, as you self-identified. But what's done is done

Any pics of how you did the turbo off the body instead?

 

I remember 10+ years ago everyone was saying how in the future in cylinder pressure transducers were going to be cheap enough to be everywhere and everyone was going to be using one. I always though how it would be so simple if they could make one work mounted in an existing spark plug. I guess its one of those things, a lot like camless engines, that everyone thought would be the greatest thing even then turned out to not be that much benefit for the cost.

 

Update on teardown ?
Also, any more pics of the ALuminum injector housing casting process ?
Thanks for sharing the very cool data aquisition project. Great work !!

 

Looks like it's been a little while since the last update, but it's been kind of slow over the winter. I haven't really wanted to do any mechanical stuff out in the single digit temperatures so I've been trying to hammer out the stuff I wont' want to do during summer (coding/electronics design).

Since last posting, she's been running and driving around occasionally. Mild amounts of boost (~12psi) on 91 oct has gone fairly well, but I've tried to avoid beating on it too hard until summer rolls around. I've actually been daily driving it for the last week since my car cracked the radiator end tank. There's still something jacked with what I think is the charge temp model that's causing lambdas to drift pretty badly when ambient temperature changes, but I haven't wanted to get the instrumentation on there yet to validate a new calibration.

Progress since last time:

• New injector drive board (pic below). All drivers are on the same board compared to the last rev. This was my first attempt at reflow soldering and it went ok, but plenty of room left for improvement. I used a thermocouple, a multimeter, the same oven I did my cerakoating in, and some solder paste. It's surprisingly easy to do when you don't sneeze on the board and send components flying while carrying it to the oven. As a result my carpet now has about 20 more resistors embedded in it.
• Code for injector drive board is partially complete. Some design flaws in the board are getting software fixes until I can get another rev made (key dead time variables will need to come in over the data acq CAN since I was being a bit dense/ran out of I/O on the microcontroller (Teensy 4.1)). It's amazing how quickly code gets out of hand, length wise. I'll be well over 1000 lines of code shortly but hopefully it will be for the better. All of the lambda targeting, deadtime, and biasing will be done on the driver board. Cold starts and batch firing (if any) will be done through the original ECU. I've configured it to run in "pass through" mode (fires original injectors on ECU command) until it's completely warm, then I'll be able to use the biasing functions to blend fuels, change lambda targets, and do some individual cylinder biasing.
• Code for the pressure and temperature board is complete. Hardware validation is a WIP.
• Injector adapters are mostly machined (pics below). Getting some of the setup correct was a bit tricky, but I was able to 3d print some "soft jaws" that fit the profile of the cast parts which allowed me to index the parts fairly easily. I still spent quite a bit of time thinking "f*** the engineer that didn't put balancing/indexing/reference features on these stupid parts" while I was machining each, but I guess learning is really the goal of this project so I'll be better for it in case I have to design cast parts in the future. Note to self: no blind holes (hard to get ceramic shell out), put balancing and reference features on parts (lots of messing around to get the 40 deg primary angle and 3.7 deg secondary angles for the injector holes right), and DEFINITELY degas the metal better next time. I ended up with enough hydrogen porosity (pics below) on the rails that I'll have to re-cast or make them out of billet. Metal quality is great for mounts, but I don't 100% trust them to hold pressurized flammable fluids.
• Rails are partially complete but I have to work around an "it fit in CAD" type of issue. It looked great on the computerized manifold I had, but I forgot about the regulator crossover and will have to bend the existing rails to clear unless I pick up a set of billet rails for the stock injectors. Interestingly enough, PLA, when printed at 95% infill, was actually strong enough to make a functional press die for the rails (pic below). I had to anneal the rails before pressing to take the temper out of the spot to be bent, but I didn't have enough press force to get them bent back to flat. I included springback angle guesses in the tool, but we'll see if they were accurate as soon as I can find a buddy with a hydraulic press.
• The instrumented plug showed up and I attempted to make a couple copies of the plug itself. It's a standard plug with a machined adapter silver soldered onto it, or at least it looks like it. The tough part that I'm running into on the indicating system setup is actually supply side constraints - I can't get any of my IC's/chips until the middle of 2022 or early 2023. That might put a bit of a damper on things for a little while.
• Injector wire thermal shielding removed - runs better now. It turns out that adding foil around the plug wire just adds capacitance to the secondary.
• Pre turbo connector cast piece is finally machined and a TMAP is installed. I learned more about how blind holes are annoying on investment casts and actually broke the metal piece once while trying to get the ceramic out, but after a bit of welding it's good to go.

IOU's on pictures:
silentbravo - header heat shields
arthursc2 - body mount turbo
everyone else: general truck pics, cast pre turbo connector piece



Bottom side of partially finished "equal leg length" dual injector adapter

Rail metal porosity

24 injector driver board V1

24 injector driver board V1

Different views of (some partially) machined injector adapters. Three variants needed: 1 equal length leg , 1 non equal length leg, and 1 end design.

Adapters with previous rev rails and previous attempt at adapter casting in 356.

Current rail design with PLA press tool. Test piece seems to be ok, but need more force to get it to fully bend.

Cylinder pressure transducer (left) and my first swag at figuring out how to machine these things (right)

Machining offset injector bosses