superlen

The ceramic header I'm sure keeps a lot of heat from pooling under the intake during a soak. That combined with your rail and FPR is a great solution....and it looks cool. Len

Zed, That's fine, I can start another thread. I think you are right on the money with the manifold mass & I believe someone said that when they switched to headers their hot soak improved. It sounds like Sarah and you both have good arguments/evidence pointing to the injectors themselves being the culprit. That makes my fuel rail argument less convincing...even to myself. I also believe I understood that plastic injector bosses were better than the aluminum ones & that would also support injector vs rail. Hopefully, I can still correct it just by metering more fuel than calculated, for x amount of time. It would be open loop in this case unless someone is running on O2 with the system which is easy to do, but they will have to weld the sensor bung into the exhaust. Lenny

Sarah, It was you that primed forever.... I knew it was one of us on here I was just too lazy to go look it up. I think Captain O may have done this as well. He's probably lurking but he's been busy with suspension duties. I agree on the higher FP. I have an adjustable and will test hot soak at different pressures when I get to that point. I still would like to have a solution for the stock system of course as many won't want to change out the FPR. I'm not convinced it's just the injectors as there isn't much fuel from the rail through the stub and injector body. It seems that after a few shots the vapor would be purged by the liquid fuel pushing it out...like burping the brake line when bleeding the valves. The first crack and the vapor(air in this case) pops out fast. I'm still a believer that a large portion of one of the fuel rails (specifically the rear one) has flashed over and the rear three cylinders are running lean. All conjecture of course. This phenomena really bothers me and piques my interest. Given enough measurements and studying and we will figure it out. To my knowledge, no one has done a qualitative study of all the variables to properly understand the physics going on under the hood of a soaking Z. Soon we will have have the perfect test bed with the HellFire. It will be easy to adjust for different FPRs, prime times, log temperatures on all the injectors, etc. Len

Three possible scenarios I see. 1) Zed's theory on leaky injectors 2) Sarah's intermittent CLT sensor connection 3) Intermittent internal ECU issue Cold start might have been in the running, but you said you had replaced it. I suppose it could still be an electrical issue firing it, but the time/temp elements of this problem don't support that. Let's ignore the CSV for now. Of the three, I think the CLT sensor wiring is where I would look first. The problem could smells of be leaky injectors as well, but that's more of pain to test. I know you replaced the CLT sensor, & you said you replaced connectors, but did you just change the injector connectors or did you do the CLT sensor as well? That's a critical one. Here's an easy test to learn something. It will require you get possibly stranded somewhere so drive around like a crazy teenager and get it good and hot then pull back into your driveway. Now, try and start it and hope it exhibits the problem. Assuming it does, now pull the CLT sensor connector off (Make sure you distinguish between it and the temp sending unit, they look almost identical - I'm sure you figured this out already). Now try and start it. If it was any other problem other than the CLT sensor then your really hard to start car will now turn into an IMPOSSIBLE fully soaked plugs floating in fuel car. This is because you have leaking injectors/bad ECU/.etc PLUS super rich from the CLT sensor being disconnected. If however, you see no change and it's still hard to start but not impossible or at least much worse, that's a good indication that the connection to the ECU from the CLT sensor is suspect. Len

I agree on just a few injectors might be the culprit. I have individual control over all 6 so I can richen up individuals.....if we just know which ones to do that with. I agree that's not a trivial thing to figure out and changing physics out in the engine bay with a better fuel rail is a much better total solution. I'm a fan of the FPR on the end & a priming pulse on the fuel pump to flush new cooler fuel on key "on". Also the higher FP will definitely help. People have tried a long prime pulse and it didn't help. (Zed, was that you? - I know you and I have discussed this some) I think that the FPR in the middle of the rail is bypassing more fuel down one of the lines & not the other so you may still have 3 out 6 cylinders running lean. I haven't tested this at all, it's just an idea. For the bone stock users, I'm hoping that we can quantify which cylinders need additional fuel & how much they need after a certain soak. Crossing fingers that these numbers will be somewhat universal with a stock setup. I have the following variables which I can use to base tables off of. - Last air temp at shutdown. - Last CLT temp at shutdown. - Last runtime (sec from start to key off) - Last every other variable of course, but I don't know what use they could be. - Current air temp at startup - Current CLT at startup What I don't have which would be super nice, but I didn't think of until after I did my board layout was a battery backup for a real time clock. This would give me "Soak Time". which as you can imagine would be really damn handy. Right now, I don't have that. I do have a nifty idea about how to come up with it (or at least a rough idea of it using a big RC time constant on a spare ADC channel). I'll add the real time clock on the next layout for the Hybrid version. Possibly with just the last and current temps, I can come up with something that will help. Soak time would make it easier/better though. Len

Chas is right. Your FPR is off the ZX but it will work fine as it regulates to the same 36PSI the Z does. One thing to watch for is that the fuel rail under pressure can blow the hose off the FPR barb if not tight & the FPR is mounted soild. The early Z had opposing barbs so this was less likely with all the short hoses connecting it. I've never seen this happen & I would think that if it was so loose that it could occur you would notice the massive fuel leak before hand, but nevertheless I've seen it mentioned before when people are running a ZX FPR in place of the Z. ????? I offered to do that earlier in this thread with my first post. I have some Turbo AFMs already coming to me now, but I still have several NA afms and if it helps you out I'll swap one with you. Just PM me. Len

Rossiz, We discussed this some on the original Zfuel thread (I think, it may have been on another thread but we were collectively brainstorming on hotstart), but the answer is Yes I'm pretty sure it can. It's untested of course, but that is a problem that plagues me as well so it's near the top of having a proper solution for. The simple solution is to richen the mixture after a hot-soak condition to combat the vaporized fuel in the injector/lines from sitting. It does sound simple, but it's trickier than it seems. The question is how much do you richen, for how long, & under what conditions do you do this? I will have a configurable table in the user interface that will allow one to adjust all these parameters. Of course, I'll figure out what the numbers should be with some experimentation and have it already setup so in a perfect world you wouldn't need to bother with understanding these numbers or adjust them . However, if a customer has a car that behave a little bit different they can tweak these numbers some to dial it in. Where you live, header choice, fuel choice, driving style, ect all change the way a particular car hot soaks so it may well be a pesky thing to get right. At the very least, we should be able to make it better if not perfect. Len

It's definitely bigger. Hopefully tomorrow there will be a turbo AFM in that pile & I can give you some feedback between the two. Len

+1 on pulling both together, although you would still have them "stickified" lying on your garage floor. It's easiser to pry them apart when they are on the floor though using any/all the methods already posted. You can also try to wedge a small screwdriver in between the two pieces near the dowels. Just a tad of wedge on one side, then the other, then back. Continue tapping all around with deadblow while the wedge is exerting some pressure. Sooner or later something will give. Note: You have to be careful with method as you can get an incredible amount of force in one concentrated area. I've done it many times & luckily have never broke anything (other than the scredriver) but I'm always worried when doing it. Len

Phar, Thanks for the offer on the AFM. I actually have a lead on 15 local as well as some ecus, so hopefully that will get me set up with a few to test, but the more I have the better chance I have at finding the "uh ohs" that we didn't plan for. I'll PM you. The CAN bus is mostly just for me to play around with CAN some more. At my company we build a product that uses CAN now, but we are using an interface chipset, not using our own internal CAN firmware. I was just going to use the HellFire as a good test bench for that. I agree with you for our cars, it's probably not going to be way useful or even a great selling point. In the future that might change though. The Virtual dash is something I have thought about and from my limited research, the people who are in to that sort of thing are just running small embedded pcs in the car. Given that, my HellTune GUI would just run native on it & they could just configure up their own person virtual dash through it as well as have access to all the other features. I think this would be way cool and is what I have earmarked for my own personal Z. BTW, I just downloaded the upgrade to my development system today that allows me to create IOS and Android native apps. I don't know when I'll have time to play with it though. Thanks again for the offer on the AFM and the great feedback. Lenny

Sarah, The pigtail will be a USB Female B connector like this: I envision just having a short 10" or so cable that comes out of the ECU and normally just tucks away behind the kick panel. When you want to connect to the laptop, you just fish it out and then plug a normal USB A - B cable from your laptop to it. When you're done, just stuff the 10" pigtail back behind the panel.

If I do place the connector there, I'll supply a template for cutting if someone so chooses. Really all that would be needed would be a hole. I'm not sure if there enough clearance to plug in the cable and have it turn before it hits the back of the panel if someone doesn't want to add a hole. As much as I dislike a pigtail cable coming out the back of the ECU, that may well still be the best solution. Len

Unfair, on the Mustang. It's the same way here, you work on everyone else's first then yours....at midnight...with the windows blacked out so they don't know you're in there. YES on the stock NA ECU, stock NA injectors, & stock NA AFM. If all three of those match & nothing else is wrong of course, it should run your L28 engine. Zed, I had forgot that the turbo AFM might be a bigger bore. (I don't know what they are, but the NA AFM bore where it connects to the throttle body boot is definitely 2.75" - I just turned an adapter on the lathe yesterday). I understand why people try to swap them around, but I'm not a believer of needing a bigger TB or AFM unless you have done some serious head work on our Z's. The main limitation is the head, not the intake system.... I think. This was the adapter I was making yesterday to mount this AFM on a flowbench to pull some numbers. BTW, I am going tomorrow or Tuesday to work a deal on maybe 12-15 AFMs. Hopefully some will be turbo. If they are, I can do a good apples to oranges comparison for us. Len

BTW, I was in Fort Wayne a few weeks back it was -12 when I was there. Your weather sucks. I'm fighting weather here in Arkansas too. I'm trying to get a flow bench finished up to do some testing and it's another blizzard going on. It was 66 yesterday, 34 now with ice/sleet/snow. Like you said, it makes it hard to stay out in it. My shop is heated, but I still would rather have a nice spring day to work in. Len

I just saw you posted on the guage. cool. Poor fuel delivery could explain it and the fact it's been sitting supports that. However, it's not the only thing.The AFM responds to load not rpm. Of course they are closely related, but when you're loading up the engine the AFM flap is near full swing (even at a lowly 1500rpms). This full swing position is where I think the two AFMs curve will depart from each other. Racing the engine in neutral will of course move the flap, but 4k rpm in neutral might not have the same fuel requirement as trying to accelerate on the road at 1500rpms. Think about this: 1. The AFM for a NA only has to respond to X amount of airflow. say 180cfm max. 2. The Turbo AFM has to respond to all that extra air, maybe 250cfm. X+70. 3. If volts out/cfm between the two AFMS are the same, then the ECU won't care & life is good. 4. However, I'm 99% certain that Bosch had to change the full scale volts out on the Turbo AFM, or the extra airflow would bottom out the NA AFM. On my NA under full load, there is hardly ANY flap movement left on the AFM. If I added 50% more air flowing through the engine, the flap would stop at full scale before the engine reached it's peak and it would for sure lean out at the top end. That's what leads me to believe that changed the scale on the TAFM. 5. Assuming that to be true, then the spring force on your TAFM is stiffer, which gives you a lower volts/cfm reading than the NAFM at any given cfm. It's even worse at higher flows because they designed it to be logarithmic. The NA ECU you have is metering LESS fuel at every point along the load curve. It's worse at the higher loads (usually associated with higher RPMS) & has probably been accounted for near idle by adjusting the idle air bypass screw so it doesn't show up at light loads. Anyhow, that's what I think, but I'm not real swift on these Z ecus and such. Len

Good looking car. It's hard not to love a red Z. As for the AFM, I don't know if the turbo *can* work with the NA ECU or not, but I KNOW the NA AFM will work with it. My instinct is that by tweaking the wheel in the Turbo AFM one can get a Turbo AFM to act close enough to a NA AFM that the car might run, but it won't be perfect. I'm guessing that's whats been done by the PO in this case. If you think it's fuel related get a gauge. Fuel pressure is a critical number to know when diagnosing EFI and without that number, you're hindered. Len

Ok, I got excited when I saw the potential for swapping out AFMs and glossed over a lot of later thread. Now I'm catching up, and have more to say. The wrong AFM could easily explain the wall you're hitting. The AFM votage out vs airflow going into the engine is modeled in the stock ECU. It expects it to be dead nuts on. The ECU will only dump the correct amount of fuel to match how much air it THINKS is going in the engine. This curve doesn't have to be wrong at every RPM and most likely at idle it isn't!! Both a NA and Turbo are going to have a good chance at being close enough in airflow that the slight curve difference won't matter. However, at higher RPMS the curves start to diverge. With a mismatched AFM in there the NA ECU around 2500 rpm begins getting something different that it wants and starts metering the wrong amount of fuel (I'm guessing lean) and you hit the wall. This EXACT same thing happened to me two weeks ago testing the HellFire. Remember I can change my curve easily through the laptop. I didn't have any baseline curve at all so I just typed in a set of numbers that I thought might be close. My engine idled (rich at first, but I adjusted and got it right)...then I started revving it up. It would fall flat around 2000rpm & I even heard a backfire. Aha! I bumped the numbers up to meter more fuel and voila I could rev to 4-5k no problem. You ECU/AFM mismatch could easily be doing the same thing. The TPS also warrants a look AZS (As Zed Said - I have to type that a lot so I need some shorthand) and several other things could be the issue as well. Just don't assume the AFM isn't the problem because it idles nicely. Len

Zed, I haven't looked at the ZX turbo wiring or AFM so I'm just speculating here, but I think its possible that those L-JETs fed the Turbo AFM with a regulated voltage, not VBAT hence the need for only three wires VREG, GND, & wiper. That would eliminate the need for the reference resistor. Once I get a turbo AFM and take a quick look at the schematics I can confirm. I do this with HellFire, btw. I feed the AFM potentiometer with it's own cleanly regulated 5V and read the signal back on the wiper. I also read the reference resistor but do nothing with the reading. The extra wire going to the AFM that was used for the ref can now be used as an analog input for ANY other sensor such as a wideband oxygen sensor without hacking up the harness too badly. Len

Unfair +1 on what the others have said. It looks like the PO grafted a NA setup onto your turbo motor, except for the AFM. I have several NA AFMS. I would love to swap one out with your turbo AFM. I am in the last stages of releasing a digital LJET ECU replacement & I'm currently testing as many AFMs as I can to make sure I have a good stock calibration curve on them. I have been looking for a turbo AFM to compare against & your posts' timing is very good. Here is a build thread on the HellFire (http://www.classiczcars.com/forums/fuel-injection-s30/49964-zfuel.html), which incidentally would solve your AFM problem as well. It's no problem to run a NA motor with a turbo AFM if you have the right calibration curve. The stock ECU of course is completely non changeable for anything. Of course we don't have that cure yet, I still need a turbo AFM hint hint. PM me if you are interested in swapping out. Later if you go full turbo we can swap back no problem. I don't need to keep the AFM forever. Lenny

I ordered some more samples in of potential USB connections. We'll see if you need to break out the jig saw or not. Lenny

Sarah, I agree, that would look good. I have some kick panels with that hole too, but I'm not sure if all do. I should do some more research. I also considered putting the USB jack there at one time, but some other mechanical internals were in the way. I may still have a solution that would allow this. I think that would be a clean install and easy for the user to jack into. Logo, P/N, S/N info will be on that side of the box too, but I haven't decided where. Nissan's thoughts on this were a good idea as it is nice to be able to determine the P/N of the ecu without pulling the panel. For the HellFire it wouldn't matter so much. I think having a USB jack with a small sticker/logo/pn right beside it would be ideal. Len

Sarah, Are you sure you are ready for...beeeep beeep beeep ? Here's a bit more about my plan for the codes. - They will be implemented from the get go and you will retrieve them from the HellTune UI. The ones available to start with will be simple obvious ones like you mentioned above. As the firmware matures, more codes will be added. I will assign numbers that match a common mnf a mentioned previously. Some codes will just be plain unique to HellFire such as "AFM showing X number of dead spots/hr" & I'll either map them to a general purpose MAF code or just have a unique one for us. - HellTune will show the code along with the descriptive String such as.. "Hey you forgot to clean your CLT sensor and it sucks!. I suggest you get off the couch and fix it". - The User will be able to ignore Any/All codes via a setup window in HellTune with check boxes for each code. - When the bluetooth interface is ready, the codes would then be available via your phone. - When the CAN bus interface is added, you can use off the shelf generic OBD readers to read the codes/sensors. BTW, I have one that's already Bluetooth and I read the codes on my Daytona via an app on my Android phone. - You can assign an output for the idiot light, but I've also toyed with adding a small led through the case. For now, I'm NOT going to do that as I think most of us will have the ECU covered by the panel and even if not, no one looks there. - I've considered adding an internal beeper, but it's not present currently. Len

If you are pulling parts and don't want to save the AFM for yourself, I would be interested in it to test with for my HellFire project. I'm looking for several AFMs to measure/test/compare to each other. Lenny

I have diagnostic routines out the wazoo as that's the first code I write when I begin testing a new system out. I have error codes just like a normal ECU that can be set and if you have an idiot light connected to one of the outputs it can come on like the wifes minivan. I'll most likely adopt the same codes as used in my 06 dodge daytona. That way when I write the CAN bus code, the HellFire can be read by an off the shelf scanner if someone cares to do that. They will just have to tell it, it's an 06 Daytona. Two of the main diagnostics I think are important are the AFM & the CLT. Earlier in this thread there is discussion about diagnosing the AFM. It's no problem for me to find any dead spots on the AFM and within reason just "fix" them. As the AFM wiper moves around during normal driving, I'll be constantly doing and FFT on the incoming data stream looking for the telltale signal of a dead spot. Once I see a continuous problem in one area, I pop a code so the user knows what's going on and then determine what the value should be (as best as possible using surrounding data points ) & use that for fuel calcs. I don't know how well it will work until I test it, but my gut instinct is that it will work great. There is one caveat. If the dead spot happens to be right in the sweet spot of your daily cruise and the AFM wiper just hangs out in that one dead spot for a long time without periodically bouncing out of it, the algorithm will fall apart. Sadly, the area where your wiper hangs out the most IS the area the carbon traces wear the most of course. There will also be a dedicated "TEST MY AFM" section in HellTune where the user can see the realtime graphed output of the AFM and can see the overall wiper health on their AFM. I am in the process now of trying to gather up as many AFMs and get a down and dirty flow bench set up so I can to test this out. If anyone has a suspect AFM or extra ones lying around collecting dust, please PM me. The more AFMs I can get my hands on & check out, the better the baseline numbers HellFire can start with and the less user tweaking will be required. Currently I don't have any plans for a dedicated oil temp, but if someone wanted to add one, they could by attaching it to one of the aux inputs & configuring a gauge in the HellTune UI to read it. The HellFire ECU however wouldn't do anything with the data, just report it to the user via the serial link. I think it might be cool too, but fuel rail pressure would be more exciting to me. Also, just a quick update. I have both the HellTune UI software and the firmware in the ECU going through some big changes, particularly the GUI trying to clean it up well enough for others to see. There's nothing much to show and tell this week, but rest assured I'm writing code, testing sensors, and generally making forward progress every day. Len

Steve, Yes. I took a look at your other posting after my first response here. Thanks for the Kudos on HellFire. I don't have ignition programmed yet, but my personal end goal is what you're installing now with trigger wheel, COP, IAC, ect. with a super clean smooth intake manifold. Len