I think I've mentioned DT/M/P connectors to him numerous times That's what I do for expansion harnesses as well, although i typically only use ~6 of the 8 expansion wires, so I only use a 6 pin connector, they're even more compact.

And don't feel bad, while my car has all DT connectors, and I recently went through and thoroughly cleaned it up, it's still a mess. But I can still preach it.

prowireusa.com has a good configurator for connectors, fyi.

 

seconded for the amazingness of the DT(range) of connectors.
Simple. Reliable. and fairly cheap. (well some are chineesium cheap, )

 

I'll get there someday... let's just say my wiring now is a lot better than my high school days when I was into car audio. So many splices and crimps... I did order some open barrel splice connectors along with those JST-XA connectors for the CAN harness, so I'll hopefully get a feel for those and start incorporating them into future projects. At some point I want to redo the wiring on the old 63 Chevy, which I think would be a good opportunity to design a harness and build it properly. Then I could carry that knowledge/experience over to other projects.

OBDII Bluetooth adapter showed up yesterday. I've pulled it apart and what I'm thinking is to solder the CAN wires straight to the board. I'll probably drill a hole in the side of the housing to pass the wires through. My original plan was to desolder the pins, but I don't have a good setup for desoldering and the pins are what locate the boards in the housing, so I figure I'll leave at least a few attached to the board for packaging reasons.

Any reason this shouldn't work? The wiring seems simple enough (12V, Ground, CANH, CANL). This is from the G4+ manual, but I assume it's the same for G4X considering CAN/OBD aren't standards that change very much (I imagine).





I figure wiring it straight up is going to be simpler than trying to splice into the existing OBDII port on the car. The underside of the dash is a hectic place and I don't really want to hack up that wiring if I can help it.

 

That'll work just fine.

In reality you could go to the dash plug with just the pair of can wires. I was planning on doing that with the Maxx Piggyback so I can put more data on the OBD2 for SoloStorm tracking, but I haven't found (or looked hard) what the pins are to terminate and lock into the OBD2 housing correctly.

 

Yeah I'd splice the CAN into the harness (again with a DTM plug), it already has B+ and GND, might as well use it.

Or buy one of these:

 

I would also tap into the OEM harness, or grab an OBD port like curly shared. That has been my plan at least, repin the OEM port, put a DTM-4 subharness to the Link's CAN 2 port and grab the 12v and GND from the CAN bus too.

 

Yeah, so I kinda started the job before anyone gave any feedback, so it's not ideal, but I think it should work.

Cut a hole in the housing, solder some wires and here's what you have.



Don't forget to twist the CAN-H and CAN-L wires. Honestly I assume this is more of a best practice thing than a necessity, but it's not hard to do.

Now, while you're waiting for your XARP-04V CAN header connector clip things (this is what you need to clip into the Link G4X CAN header), you can cut up your hood.

No going back now... for a second time. Start by making some constellations in your hood.


After that do your best imiation of a slasher movie and hack the thing to bits. It doesn't have to be too precise or clean, but the better job you do here, the less time you need to spend with the bastard files. Side note - from some brief research, it sounds like the name of the bastard file comes from it's coarseness being a mix of coarse and fine, hence being irregular and not fitting in a specific classification. Take that with a grain of salt, but I always wondered where that name came from.




When I added the first vents I used an air powered body/reciprocating saw. That thing would bind up a lot, but also allowed for nice precision when doing rounded cuts. After talking to a buddy who added hood vents to his turbo'd 86 I decided to give this thing a try. $35 at harbor freight and honestly it melts the hood metal. It's not as precise, but it's far from unwieldy. Coupled with a Dremel and some good cutting disks it's a pretty good way to go about it.

It did boil and burn my immaculate paint, but that'll get covered by the vents anyway.




Cut, measure, cut, file, measure, etc...

I ended up picking some distinct points on the hood, such as the points at the back, to do measurements from. I would use those to locate the tape measure and check distances to the rivet holes to make sure I was fairly symmetrical between the two sides. I wasn't too worried about making them perfect, but I wanted them to look decent at a glance if nothing else.



Shockingly, the center panel, which I was the most worried about, ended up being the easiest. After a quick mockup using the stock hood frame as a reference I mocked up the locations for the rivet holes and used some OEM holes in the hood to check the alignment. Pretty spot on.



I didn't realize it when I took the photo shown below, but there's 1 bar in the center vent that is connected to the rest of the panel in 2 spots. Why? No idea, but I had to bust out the Dremel again to clip those when I went to bend the flaps.



Damn, finally looking ice cold. Honestly I think I like this look better than the single side.

Ps - sorry for the photo compression. I'll get a proper photo up soon enough.



So Digikey also came through and got me my XARP-04V connectors sooner than expected. I was quick to get some harnesses built. One with the cheapo OBD2 dongle I showed previously and one for future projects.

Also, because I got roasted by @redursidae I decided to trim up the case a bit and route these out alongside the Denso 76 pin connector



Finally got to use the cool drill trick for twisting wires. I've been wanting to do this since I saw the car audio guys doing it when I was in high school.



@redursidae eat your heart out.



That was a surprisingly full day. A lot of time was spent cutting, filing and measuring on that hood, but I'm happy with the way it came out. The wiring was actually pretty quick, until I tried to use it. I got most of the ECU and sub harnesses reattached and installed only to realize that I had no voltage on any pins.

Ripped everything back out and apart to realize that while my wires passes a quick tug test, that none of them had really penetrated into the XARP-04V connector. Nobody seems to mention the part number of the connector for these harnesses, so I figure I'll go overboard with it. The other harnesses on the board are XARP-08V by the way. The crimp connectors/pins for these are sxa-01t-p0.6 or at least that's what I've been using.



Anyway, I used a pick tool to push all the wires in. I'm using some pretty garbage wire honestly. It's some cheap Amazon stuff, but it's what I had on hand and it's held up decently for the stuff I've used it for previously, so I figure I'll keep going and see how it does.

Got it back together and... nothing... I spent a decent amount of time troubleshooting but wasn't able to get the OBD2/CAN stuff working. I can communicate with the OBD2 dongle via bluetooth, but not to the ECU. I pulled the dongle and can measure voltage on both CANH and CANL and verified that they are coming through on the appropriate pins. Tried it with RealDash and "Car Scanner", which I believe I've used in the past on my Subaru.

Tried a variety of protocols and baud rates but couldn't get anything to work. I ended up clipping the +12V and ground wires and running the dongle off the built-in OBD2 power/ground. Still no luck. Not really sure what would be causing the issue. I definitely should've gotten familiar with the dongle and an app before trying it on the Miata, but I figured I was being paranoid and this would be simple.

I think I'll keep doing some reading and maybe post on the Link forums to try to get some advice. I think it could just be a setting on the Link that's causing the issue, but the OBD2 support there seems pretty straight forward so I'm not sure what I would've missed.

Signed up for the Packwood Prosolo tonight too. Haven't been there in a couple years, but I feel like the car is starting to get sorted enough to be competitive, so it'll be nice to see how it stacks up.

 

Shoot man. I'm sorry you felt roasted by my comment! I did not mean it that way! But nice job on the new method

Can you post a screenshot of your CAN 2 page in PCLink? Does the dongle light up with key on? If CAN 2 is On, then the dongle should at least light up and you should have 12V and ground. If you're not getting voltage then something is up with the wiring.

 

I'm just giving you a hard time, no worries

So what I see on the app I've been using is that the app can connect to the bluetooth on the OBD2 reader, but can't communicate with the ECU. I do see a power LED on the bluetooth dongle and a flashing green light, which I'm assuming just indicates Bluetooth traffic. I'm using CAN1 fwiw.

I've tried connecting with the dongle prior to using the app to connect, and allowed the dongle to handle the connection. I think I had issues with MSDroid when you connected to the bluetooth on it via the phone instead of allowing the app to handle it. No change there.


I did twist the CAN-H/L wire, but didn't add a termination resistor. I assume that would be built into the dongle. Regardless, I doubt that's necessary. I'd assume that would be more of a best practice type thing. IIRC there was a post from Link that said you didn't need to add a termination resistor, but I could be wrong.

I did end up clipping the power and ground pins from the ECU and am using the ones from the OEM OBD2 Port (fitment issue and debugging). I could probably reconnect the ECU provided power/ground, but I don't have any reason to assume that would make any difference.

I'm kicking myself a bit for not figuring out the app and verifying the dongle worked on my Subaru beforehand. I could still do this, but I would need to pull the ECU to disconnected the OBD2 subharness thing. I could also clip the CAN lines from it and splice them into the OBD2 wires from the OEM harness. Either way, I will likely need to pull the dongle/ECU at some point to double check that I didn't break an existing solder joint - on CAN-H/L - when soldering wires to the dongle.

 

to check if you have correct termination resistors, just ohm between can L-H and if you have 60 ohm or in the ballpark, you're good.
If it's 120, then you need to add a term resistor.

have you double checked you haven't swapped can L/H?

And for easy removal you could get one of these
connector kit

Or, one of these.

 

Alright, so the dongle is powered. Yes, verify the CAN wiring and also lower the bit rate. The dongles usually work between 128kb/s and 256kb/s.

 

I did verify that CAN H/L are not swapped. The app I've been using is supposed to auto detect the protocol version that's being used, but I have manually tried all of the 15765 options that it offers. Those are basically 500 or 256 KB, long or standard format. None of those have worked so far.

I'll dig into it later this week most likely. Once it's up and working I really doubt I'm going to pull the dongle back out ever.

 

You sure you don’t still need to transmit generic dash with obd selected?

 

That might be it Curly. I'll give that a shot. When I opened the menu up it was setup to transmit user streams, so my initial testing was with that.

In at least one of the posts/videos I came across the person had it working without any Channel's configured, so I assumed it would work without that.

 

It should work without generic dash, but let us know what you find. I do think you should lower the bit rate to 250kb/s

 

In other news, I took the car over to the local hill and beat on it for a bit last night. I was able to get the car up to 105 C within about 2 minutes of climbing. At that point I hit one of my protections (ended up being MAP limit which lowered with higher coolant temps).

30 C (85 F) ambient temps, higher boost and going uphill is a stress test, but that's quicker than I'd hoped to be seeing those temps. I would post some log data, but I think the log was lost when my PC decided update last night.

Things I'm considering right now in no particular order. PS - I know I'm beating a dead horse here, so feel free to skip if you're tired of me talking about cooling upgrades that may or may not happen in the next 1-2 business years.
1. Pull the radiator and attempt to clean/straighten the fins on the AC condenser (and the radiator while it's out I suppose). I don't think I've ever cleaned the condenser and frankly don't remember what kind of shape it's in.
2. Swap from the Mishimoto Radiator to the SuperMiata unit. I want to do more research/reading on these two units simply because I'd be pretty bummed to spend the money to "upgrade" and not see the performance difference I'd expected/hoped for. I feel like I've done that with some other parts recently.
3. Add an oil cooler. I'm realizing that this will probably need to happen at some point anyway. At some point I want to move to a GTX2860 and run more power, so I figure I'll need something then, if not now.
4. Pull AC. I still need to experiment more to see if I can even use AC when putting around town without overheating. I've been able to use it a bit when roadtripping. At this point, why have it in the car if I can't use it, it's hurting cooling performance (presumably) and it's extra weight? I'd like to keep it if possible, but I'm wondering if I'm being delusional.

I'll rewatch it again I'm sure, but I found this video to have some good topics. Mostly stuff I already have some idea about, but interesting to hear a pro talk about it.

When I hit boost cut it was way more abrupt than I expected. It was setup as a soft cut but felt like a hard cut. That led me to go back through some of my protections and redo/correct some issues I hadn't realized were present. For example, I had setup a General Purpose RPM Limiter for oil pressure, but for some reason that was turned off? I also had the CEL set to illuminate when coolant was above or below certain set points, but only had one of those rules being used in the logic.

The condition below was originally set to Cond 1 or 2, instead of 1 or 2 or 3. I'd see the CEL on warmup, so I assumed it was working (I thought I lowered the hot temp at one point to verify the logic, but I guess not).


I've got most of the protections set to start pulling boost/timing around 110 C (230 F) and pretty aggressively ramp down as things approach 120 C (250 F). I wish Link gave a bit more granularity on some of these tables. Some of the choices on table values they use are strange IMO, but they probably know better than I do. For example, I doubt many people care about tuning boost control around 200 C (392 F...). I'd much rather have some more granularity in the 80-130 C region.


So hopefully that experimentation pays off when I actually need it. Part of the reason I want to get the CAN/OBD2 stuff working is so I can monitor this stuff as we roadtrip around in the hotter months. I remember sweating while we climbed some hills on the way to Montana last year, coolant temp slowly ticking up with very few places to pull off. While ignorance is bliss, I'd like to be able to tell if I'm going to need to pull over. The CEL is better than nothing, but far from ideal.

 

If you change your table from 2D to 3D, then narrow the table to something small like TPS 0 and 100%, you can edit your axis and delete the high temps, and add some more resolution, or granularity as you call it. That being said, just looked at ECT trim and looks like you can't change that one to 3D.

zero that table out and activate GP boost trim, and make your axis ECT instead of RPM, recreate the above table but with your choice of break points. Press the "init axis" to get basic axis starting points.

 

Smart idea Curly. I didn't think about using another table for that. I don't know why they limit some tables and allow tons of flexibility on others. I guess that's just how they allocate the memory on the board. I'll poke around at that.

Also, made a bit of progress on the CAN side of things. The app on my phone is now reporting that it can talk to the ECU which I believe since it's automatically detecting the correct protocol and bitrate. Only issue now is that it doesn't actually get any values. I'm assuming that's a setup issue on the app or the link, so I'll poke around with that.

Pretty sure the issue was some of my soldering. I've got things cobbled together for now, but I didn't realize how easy it was to pull the OBD2 port from the car and add the pins, so I'll likely go that route in the end.

 

Current state of the OBD2/CAN project. CANH/L are just jammed in there with no pins. I slapped some electrical tape on the back to make sure they don't pull out.





I was a bit fixated on the cooling system yesterday, trying to figure out anything that might be causing the system to run less efficiently than it should. I ended up deciding to pull the radiator to inspect it. I remember a friend telling me his car was having cooling issues and it turned out to be his radiator was so full of crud that it was barely flowing any air through it.

First step jack up your car. Second step, throw Wingman's latest video on in the background.


I'd also wanted to get under the car to look around on the drivers side front. I've heard some noises recently that seemed like they were coming from there. Nothing super concerning, but enough to make me want to look around. I didn't find anything there, but I did peek and see a ton of oil on the front of the pan. I've got a leak somewhere on the front of the engine. Maybe that'll be a winter project. Shockingly, the trans is still completely dry and clean. The old trans always had a little drip where it met up with the engine.



While we're on the topic of cooling. I realized there's a pretty sizeable gap between the bellypan and my ducting. I'm not super versed in aerodynamics, but I was curious if this will cause a lot of high pressure to build up in the bay and negate my ducting. The coroplast goes right up to the bottom of the bumper. Should I have the belly pan sealed to the coroplast/bumper? I assume so, but am curious to get people's thoughts.



Got the radiator out and...it looks like a radiator. Nothing much to say. I don't think I've had it out of the car since I put the engine in. There were some bent fins and a bit of gunk and grime, but nothing egregious.


This is what a radiator looks like. The M stands for "Man, I should've bought a different radiator"




So the radiator got the hose again, along with some passes from the shopvac brush attachment. I didn't have a great way to wash the AC condenser, but I did what I could with the shop vac. It didn't seem too dirty, but I figured giving it a quick clean wouldn't hurt.

Found a PS line that had rubbed through the insulation. I ended up trimming the fan shroud a bit more to give this some breathing room.


There are some big gaps where the fans meet the radiator. Previously I had these sealed up with tape, but a lot of sand/rocks had accumulated there so I removed that to clean it up. I ended up replacing it with some foam weather stripping. I assume that's beneficial for fan performance, but I'm curious why those gaps are there. I'm guessing they result in smaller gaps with the OEM radiator design?



Cut yourself in the garage? Just glue yourself back together.


I had this stuff laying around for backpacking but have found it really useful in the garage. Seals up cuts pretty quickly so you don't just bleed through bandaids for the remainder of the project.

It was too late to bleed the system yesterday, but I got it all back together and threw some coolant in it. I don't have high hopes for this to make much of a difference though, mostly just doing my due diligence.

So, I'm curious what people think. At this point do I get a better radiator, or look at an oil cooler setup? I don't know what my oil temps are right now, but I definitely see lower pressures when I finish a run.

I think the price would be comparable for either the SuperMiata radiator or and oil cooler, unless I can source either of them used.
The radiator would be an easier install, but I'm not sure how big of a difference it'll make. The oil cooler would be more of a project, but I assume would be a bigger performance increase. I say that in the sense that my radiator might be a 7/10 right now and an upgrade might get me to a 9/10, while my oil cooling setup is basically a 0/10 now and adding a cooler could get me to a 9/10 there.

Ultimately my goal is to maintain coolant temps. There are some highway hills around here that I can't maintain temps on (just doing 55 or so) and I remember watching my temps tick up when we climbed Lolo pass on the way to Montana last year. Both of those are highway cruising, not pushing the car.

 

I wouldn't be taking any action with seeing 105C (221F) coolant temps on a sustained high load climb. Seems to be well within a non-concerning temp range to me. Get north of 240F and then I'd start questioning your setup... my .02