For some reason I had a mental block on wiring up the oil temperature and pressure sensors. Don’t know why since the issues creating all the angst are resolved. Today it was like having pushed off the yucky vegetables until last. Anyhow, some time back the sensors were calibrated, so with the resistance values on-hand the scalings were entered into the dash. Then there was snaking the new wires through the existing harness, a reminder than it isn’t done until it’s done. The wiring behind the dash had to be changed, too, adding the additional inputs I’d failed to plan for.
Race Technology references sensor readings to 5 V, and trouble starts when using cheap sensors like the ones used with pointer gauges. The trouble is that the cheap sensors tend to have very low resistance at one end of their scale. Race Tech provides a +5V voltage reference for the sensors, but they conveniently assume users will use the expensive sensors that have higher resistance (that they just happen to sell.) For that reason they used a weenie voltage reference capable of only outputting 50 ma (0.05 amp) and unfortunately, using just one low-resistance sensor means it may draw 50 ma on its own, leaving nothing for additional sensors. Soooo, higher power precision voltage references were going to be added, capable of supplying 150 ma apiece, but I forgot I’d need perf board or circuit board material. Whining aside, that’ll be picked up this week, the reference wired in, and the issues will be put to rest.
Great news on the ECU front. Joe of Locash Racing very generously tested my Hondata ECU and found that indeed, the engine has to be running for MAP to read. He said A/F read fine, too, but only after the engine’s been running for about 45 seconds. In the 30 seconds I’ve run the engine, A/F has always been frozen, so that explains that. Knowing the ECU is good is a HUGE step forward mentally (and financially). I can now move forward knowing that when something doesn’t work, it’s either wiring or a bad sensor, not the ECU.
There’s developments regarding the Race Technology ECU/Dash Interface. Turns out that RT makes exactly two units: a generic one that works with virtually every ECU out there, and another to work with the one ECU on the planet that’s different – guess which one that is? The Hondata/Honda ECU’s datalog output isn’t true RS-232, so RT made a unit just for this ECU. Unfortunately they apparently never told their U.S. rep, so when my order came in, it was handled the same as virtually all orders, and a generic interface module was programmed for my ECU type. So with that cleared up, the correct unit should be here by the end of next week, and hopefully by then a known-good ECU will also be on-hand.
Interestingly, someone with this exact same problem posted on the RT forum about a year ago. Unfortunately, people complain a lot easier than they offer solutions or complements; once they have their answer they don’t bother to post the solution. I’m trying to do better, posting the problem along with the solution, maybe saving someone some time in the future.
The ECU has been shipped to a guy who specializes in KPro ECUs. He’ll check out the ECU completely to at least get me to a known baseline. As it is, it’s hard to know if it’s the ECU, me, the dash, or a combination. As said before, air/fuel ratio is always stuck, the O2 heater never comes on, and MAP is stuck. It’s possible that I have to run the engine, or run it for a while, to see if these issues resolve themselves. Asking about it on the Hondata forum was frustrating due to the circular reasoning and went something like this:
Me: What’s MAP read when the engine’s off?
Forum: What does it read when it’s running?
Me: I don’t know, I can’t run the engine now, but it appears stuck. I’m going to send it to someone.
Forum: You haven’t even run the engine and you’re saying it’s bad. Try it in a known-good car.
Me: I’d love to but don’t know anyone with one. The reason I think it’s bad is because no one’s answered my question.
Forum: It doesn’t read until the engine’s running, don’t trust the readings until it’s running.
So I still don’t have an answer. Turn on the key, see what it says, how hard is that? Since no one can do that… okay, I’ll figure it out on my own, so, surprise, it’s going out to be tested.
This brings us to the ECU Interface module between the ECU and dash. As mentioned, data is coming out of the ECU at the proper baud rate, so that’s not the problem. The Race Technology rep contacted me today and it’s down to these possibilities: 1. The Interface module expects different data then what’s coming from the ECU, 2. It needs some handshaking it isn’t getting, or 3. It’s broken. I don’t think it’s #2 because it’s supposed to “just work” without any outside help. Assuming it’s not broken, that leaves #1, so looking on the Hondata site, a list was found of the data that’s sent out the datalogging port. Comparing it to what the Race Tech Interface module outputs shows some interesting differences. The Hondata ECU outputs 15 engine parameters, yet the Race Tech Interface outputs only nine parameters… hmmm, what’s with that? So either the Race Tech Interface is tossing out parameters it thinks I don’t want, or it’s rejecting the entire packet because it thinks it’s the wrong length, which, given that nothing’s coming out, seems to be the problem. Awaiting a reply from Race Tech in England…
Heading into the warmest few weeks of the year – very unpleasant to be in the garage. Good news bad news. The good news is that the problem’s been isolated for why ECU data’s not appearing on the dash; the bad news being that it appears to be the ECU Interface module. With it’s hooked up per the manual and feeding the output directly into a PC – nothing. Took it apart and put the scope on it, which showed data’s coming in on the right pin, the board’s powered up, but nothing’s coming out. Oddly, it reflashed okay so it’s at least somewhat functional. Waiting to hear from the U.S. rep.
Regarding how the dash seemed to only take a portion of its configuration was answered (sort of) by Race Technology, that until the value changes the first time, the dash won’t display it or the title. In other depressing electronics news, there’s growing suspicion that the used engine computer I bought has problems. Ever since Day 1, Air/Fuel has always been frozen, even after checking and rechecking the wiring – it’s as if the ECU’s monitoring circuit is bad. Connecting everything up recently shows that MAP is stuck reading -28.9″ all the time regardless what voltage is fed into the ECU. Thoroughly checking out the wiring shows no problems… crap.
Left to ponder the above, attention turned to the short list of stuff left to do. Fabricated the spacers for the tops of the front uprights. mounted the steering rack, installed the various C-clips on the spherical bearings, put in the keepers for the bottom shock mounts. There’s still the hi/low boost switch and scaling the oil temp and pressure sensors, but it’s just too warm.
So I get an e-mail out of the blue from the U.S. Race Technology rep. He said that someone in Australia told him he had an unhappy customer – me.
Since this is the same rep I bought the units from originally, calling him was on the list of things to do this week; I just hadn’t gotten around to it due to work being nuts. So he and I had a very constructive conversation and he provided great support. So all three parts of the puzzle will have the latest code flashed into them, then we’ll try again. Normally I’d have flashed the lasted code into them first thing, but the manual made a big deal about not doing that unless instructed… well, I got instructed. If something ends up being seriously wrong with anything, he said he’s swap it out. That’s good support, so I’m hoping for the best.
Before his call I’d been scanning their Knowledge Base, printing out all the pertinent documents, so we’ll see how it goes.
Worked all day to get the Race Technology flat-dash talking to the data-logger and ECU, and – fail. The dash configuration doesn’t get applied about 70% of the time the panel’s powered-up. It’s not that it didn’t get into the unit, because if it’s turned on and off a few times, sometimes it’s fine, but turn it off and back on again, it’s configured differently. Even worse, sometimes it’ll bring up only <em>part</em> of the new configuration.
In addition, the dash isn’t reading anything from the ECU, even though the quirky and vague instructions have been read and reread. Data is indeed coming out of the Hondata-modified ECU at the right baud rate, but even with the correct boxes checked in the configuration setup, it’s a no-go. I’ll read through their FAQ after I cool off, as different document’s say different things. It’s like a treasure hunt, piecing together what should be one coherent document, yet it’s seemingly scattered all about…. Not real happy right now, downright grumpy in fact. I really hope the problem is with me.
It’s amazing how long it takes to bolt on suspension: checking and finding the correct grip length of each bolt, making rod-end spacers (the real time-sink), setting the rod-ends at the correct depth in each arm, mocking it up to double-check camber, then tightening everything down – with Nyloc nuts that take a while to tighten. The good news is that the suspension’s all in! And, the rear brake rotors did in fact space the wheels out to where they were supposed to be.
Various bolts are on order, including the ones for the steering rack. Technically, Midlana could be sat down on her wheels. However, as much of a milestone that would be, I’m going to hold off since there’s still a number of things to do – lots of odds and ends in the electrical system. Having it up at “working height” is nice. Also, it’s good to be able to crawl under the car, so, no video of me jumping and down on the suspension just yet.
Progress continues, albeit at what feels like a glacial pace. Mentally, the project’s at something of a tipping point. All the big hard stuff is done, and many of the little bits are done too, so larger sub-assemblies are starting to come together.
Today the right side rear suspension was completed and the axles were installed (they still fit!) What’s frustrating right now is whenever there’s a sense of progress, it gets humbled by seeing what’s still left to do. For example, the rear suspension’s in, but I noted that the oil pressure and temperature sensors have yet to be interfaced to the ECU… which hasn’t yet been interfaced to the dash either, and on and on. So while progress is happening, there’s still many things on the pile.
Found out today that the rear uprights are in a slightly different location from where I designed them to be. Not off by enough to warrant doing it all over again (good!) but it’s very annoying that it happens at all. I triple-check everything suspension-related and this sort of thing really bugs me. It’s not a big deal to the design, but it’s a real drag mentally – the feeling that something that was thought finished and correct might be wrong.
That aside, part of the reason for going with 275mm rear tires was to confirm they do indeed fit, so that when builders ask how large a tire will fit, well, now we know. There’s about 3/4″ clearance between the tire and chassis which is barely enough… though… hey… I just realized something… the brake rotors aren’t installed. They add about 1/4″ offset, which helps correct the above-mentioned unexplained goof… somewhat.
Speaking of the axles; with them in-place, it was possible to finally test the clutch effort and feel. It’s not as short a throw as expected, and pedal pressure’s pretty firm, but I’ll manage. Oh, and the picture with the tire in it, yes, it’s spinning, but no, the engine’s not turning it, I just gave it a whirl before snapping the shot 😉