When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
When I was building my harness and looking for ways to build an AI system this winter, I didn't find much info, and now I'm finding a lot. So I know I'm not breaking new ground here, but hoping to help others with an easy and fairly inexpensive setup (just under $300) that is still somewhat advanced. I wasn't absolutely sure it would work, but it seems like it's working pretty well. The map is somewhat of a guess so far and I haven't done any boost tuning past 10psi yet.
I know the AEM setup is very easy plug and play, but I wanted more control over how much was injected and when. I had an FJO kit for several years starting in 09, but eventually the driver failed. I kept the solenoid, pump, and filter and tried using the Syvecs to run it, but I didn't do any research beforehand, so I didn't think about the solenoid being too high amp to use without some protection for the ECU and fried that circuit. I had replaced the FJO solenoid with a Holley solenoid, thinking maybe that was the issue, and that didn't fix anything. I'm not sure the amperage on the FJO solenoid actually, but the holley is 30a and it looks to be exactly the same. WAY too much with no protection. I had started installing a Snow Performance kit at one point a few years ago too, so I had that pump for a while but the Syvecs only controlled that setup for maybe 5 minutes before it fried.
A couple caveats to this is I'm using the nozzle from the FJO kit, and I'm using my washer fluid tank, so something will have to be sourced for those, and I'm still using the filter as well. I bought this specific pump because it has a built-in pressure relief and I wanted to keep it fairly simple and not have to wire in a pressure sensor, but I might end up doing that anyway at some point since I have a couple more inputs available. I bought this solenoid because it's less than one amp, so I figured maybe I wouldn't even need to use a driver and diode protected relay, but I did just in case.
The frequency for the solenoid should be between 20-30hz. I had the best results so far setting mine at 20hz. At 25hz, it wouldn't start injecting until about 70% duty cycle. At 30hz it started around 30%, and 20hz starts around 13%. I haven't tried any other values in that range yet, but I might do that just to see if I can get it down closer to 1%.
The pump is mounted on the frame rail across from the spark plugs. The relay and driver are mounted where the airbag used to be. I'm not using an enable switch, just 12v switched power. I didn't connect the indicator lights either, but I might do that later.
I got antsy and tested different frequencies tonight, including dropping below 20hz. It seems as though the chosen frequency is pretty close to the duty cycle it starts injecting, 30hz at 30%, etc. I set the frequency as low as 10hz and 15hz. 15hz works, but it cycles slower, which anyone reading this probably already knows lol.
If you're pulsing the solenoid, be sure the driver box has provision inside to handle the back-EMF voltage spike that happens each time the driver disconnects power from the solenoid and the magnetic field collapses. It's hard to say without knowing the FJO circuit design, but getting that wrong could have damaged the FJO driver.
I too am also running a prometh solenoid and that amp box.
found some interesting things when flow testing nozzle
for me i get no spray until i open solenoid at 40% at 30hz
30 hz40%-75cc50- 12560- 15070- 19080- 22590- 295100- 320
using a 20gph nozzle, have my haltech controlling it
And you tried lower frequencies? Do you get ANY fluid coming out before 40%? Are you using a ProMeth nozzle or what? I was looking for options originally but decided to just keep using the FJO nozzle. It atomizes really well.
I don't understand any of these numbers lol "75cc50- 12560- 15070- 19080- 22590- 295100- 320"
Originally Posted by scotty305
If you're pulsing the solenoid, be sure the driver box has provision inside to handle the back-EMF voltage spike that happens each time the driver disconnects power from the solenoid and the magnetic field collapses. It's hard to say without knowing the FJO circuit design, but getting that wrong could have damaged the FJO driver.
When my FJO driver failed, I was using all FJO components still. I didn't even know it wasn't working for a while until I noticed my fluid level wasn't dropping. I'm only guessing, but Rodney from ProMeth said SIRHC uses their solenoids and pumps quite a bit. So I'd say it's good, but I wouldn't really know, I guess I'll find out lol.
because here’s what their website says about the solenoid control method that may explain the variances:
When using PWM to control a solenoid there is a limit to how quickly the solenoid is able to open and close. Below a certain duty cycle the solenoid will never open, and above a certain duty cycle the solenoid will never close. While the opening time is primarily determined by battery voltage, the closing time can vary greatly depending on the hardware being used to control the solenoid.
which they go on to say their solenoid controller helps to control the closing time, but you still having the opening time variance and overall solenoid variance as well.
.
ya I spoke with the guy who made it, and we ended up at my current frequency
below 40 nothing comes out, if I changed frequency it would open soon but would max out faster. so basically just shifts the open/close time from sooner to later
I'm currently using a 20gph nozzle from snow performance
Last edited by AlexG13B; Jul 25, 2023 at 09:38 PM.
because here’s what their website says about the solenoid control method that may explain the variances:
which they go on to say their solenoid controller helps to control the closing time, but you still having the opening time variance and overall solenoid variance as well.
.
I've always used a solenoid, and this specific nozzle, which I really like because it atomizes so well, and I worry that other nozzles wouldn't. Also, I wanted very specific control over injection, like a fuel map, and I know I can get it this way. I'm not sure whether I could by just modulating the pump, I don't remember honestly. I have a pretty wide range now, so I'm good with it. I've watched how it injects so I know exactly what it's doing, when, and how much.
Originally Posted by AlexG13B
ya I spoke with the guy who made it, and we ended up at my current frequency
below 40 nothing comes out, if I changed frequency it wouldnopen soon but the would max out faster. so basically just shift the open/close time from sooner to later
I'm currently using a 20gph nozzle from snow performance
Interesting that there would be such a difference between our solenoids.
reading further USRT wants to use the WMI pump driver along with their direct electronic injector nozzles, so it was just my lack of understanding. I get where you’re coming from now.
.
reading further USRT wants to use the WMI pump driver along with their direct electronic injector nozzles, so it was just my lack of understanding. I get where you’re coming from now.
.
Well, you've read further into it than I have. I knew what I wanted to do and saw the wiring diagram and bought the parts I needed, and that's it lol.
I heard different things from different people regarding what I needed to make this work. I have a solid state relay from Haltech that I didn't use, but most likely could have. I decided to try it this way since there was a wiring diagram, and then once Rodney told me about SIRHC using his parts, I figured I was on the right track.
I was going to contact USRT and get their specific recommendations. Thinking about switching from an AEM system to their electronic Race Valve injectors, Hammer high P pump, filter with sensor, and a SIHRC Cortex controller instead. Possibly the BigUpper pre-compressor injector too since I’m shooting for response with a smaller turbo that may need some additional charge cooling on the upper end.
.
This may be a little off topic, but I would like to also have AI for my build with a Haltech ECU. I am almost convinced that the R3 model, given the price/features, would be a good choice. I am wondering with the PDM if I can run the power line to the AI pump and be able to adjust the voltage depending on boost level and therefore have some ability to adjust the flow.
I was going to contact USRT and get their specific recommendations. Thinking about switching from an AEM system to their electronic Race Valve injectors, Hammer high P pump, filter with sensor, and a SIHRC Cortex controller instead. Possibly the BigUpper pre-compressor injector too since I’m shooting for response with a smaller turbo that may need some additional charge cooling on the upper end.
.
I haven't looked at their selection much, but I'm guessing that's more advanced than the AEM system haha.
Originally Posted by iceman4357
This may be a little off topic, but I would like to also have AI for my build with a Haltech ECU. I am almost convinced that the R3 model, given the price/features, would be a good choice. I am wondering with the PDM if I can run the power line to the AI pump and be able to adjust the voltage depending on boost level and therefore have some ability to adjust the flow.
I was almost convinced on that ECU as well. It was very tempting. But the cost vs my needs and the fact that it's probably too big to fit where the stock ECU is were reasons not to. But that ECU should be able to do anything you want. I don't know for sure if it can do that, but I would imagine it's that capable. Haltech has some great support though.
I've been meaning to bring the car home some time and cruise with you guys. I'm not from StL but the area and I've hung out with Matt Trebacz some when he's come out to Colorado.
I haven't looked at their selection much, but I'm guessing that's more advanced than the AEM system haha.
I was almost convinced on that ECU as well. It was very tempting. But the cost vs my needs and the fact that it's probably too big to fit where the stock ECU is were reasons not to. But that ECU should be able to do anything you want. I don't know for sure if it can do that, but I would imagine it's that capable. Haltech has some great support though.
I've been meaning to bring the car home some time and cruise with you guys. I'm not from StL but the area and I've hung out with Matt Trebacz some when he's come out to Colorado.
I looked at the measurements and the R3 is slightly bigger. I was looking at the IRP mount for the 2500 with CAN bus and hope there will be enough space. It looks like there is quite a bit of space available.
I took agree it's probably more capability than I might need, but the 2500 plus the can bus and wide band is pretty much the same cost.
Should mention that you need to repin the Patch harness to suit the R3, there is no off the shelf patch harness to go with the R3 at the moment to plug into an RX7
Should mention that you need to repin the Patch harness to suit the R3, there is no off the shelf patch harness to go with the R3 at the moment to plug into an RX7
Correct. I spoke with Ry-Wire and they said they could try, but would lose functionality. I briefly spoke with Chris Ludwig and think a custom harness would be the route to take full functionality.