Radium Fuel Rail and why I got one
#1
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
After advice I got here on my hot restart problem (and advice from DIYautotune) I have been researching converting to a return system on my MSM. I like some of what I read, but I did not want to rig something up with an NA 1.8 fuel rail. Until I found a local company that made a rail with a provision to install the stock damper, and plumb it how I want to. I did some searching, I found a few mentions of the Radium fuel rail but nobody using it yet. So I am giving it a shot, because I don't want a old NA rail or FM's wonky dead ended duel feed system.
The plan is to run a lines like:
fuel pump
stock filter
Radium fuel rail (with stock damper)
Aeromotive A1000 FPR on firewall (left over from a old project)
Evap live
stock return line on fuel pump cover
remove 60psi in tank regulator.
Can I get any advice on running 60psi static like stock (with FF fuel injectors) or running a vacuum line to the 1:1 AFPR. Is going vacuum referenced even necessary?
The plan is to run a lines like:
fuel pump
stock filter
Radium fuel rail (with stock damper)
Aeromotive A1000 FPR on firewall (left over from a old project)
Evap live
stock return line on fuel pump cover
remove 60psi in tank regulator.
Can I get any advice on running 60psi static like stock (with FF fuel injectors) or running a vacuum line to the 1:1 AFPR. Is going vacuum referenced even necessary?
#2
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
I have one, it's a solid unit.
Also have their manifold pressure referenced fuel pulse damper, maybe look into one of those instead of re-using the stock one. Also went with a Turbosmart FPR and a flex fuel sensor in the return line.
Set up the regulator to be boost/vacuum referenced. Makes things easier.
Also have their manifold pressure referenced fuel pulse damper, maybe look into one of those instead of re-using the stock one. Also went with a Turbosmart FPR and a flex fuel sensor in the return line.
Set up the regulator to be boost/vacuum referenced. Makes things easier.
#4
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
That's a great unit, very well built, and your plan is a good one, but it won't completely solve the hot restart issue like you think and like they say (IMO). I've tried it. It definitely helps though, since the issue is hot fuel, and the faster you can cycle it through the better
Arca_ex: I saw they have a pulse damper that will screw directly onto the rail. Not sure it will fit a stock intake manifold, I am sure it would work with a Skunk2, which would give me another reason to get one and sell my welded up VTCS mani... maybe one day.
All in all I am very happy with this rail, you have to buy a fitting kit separately for another $50. Which will work for a 100% stock install. But that is not why I bought it (because the stock rail flows enough for my IHI and FF injectors) So I only need 1/2 of the install kit. I still think this is a great option for NB owners looking to convert to a return style fuel system... oh and pimping a local company, because buying local produce/meat/car parts is what the trendy people do... because reasons
#5
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
That's a great unit, very well built, and your plan is a good one, but it won't completely solve the hot restart issue like you think and like they say (IMO). I've tried it. It definitely helps though, since the issue is hot fuel, and the faster you can cycle it through the better
![](https://www.miataturbo.net/images/smilies/smile.gif)
I have plans for solving this by compensating for the altered dead time in software, either by adding a circuit to monitor the current draw of the injector (and thus directly measure the dead time) or by proxying it with a thermocouple on the fuel rail.
--Ian
#7
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
I think the issue is actually hot injector coil wires, but cycling the hot fuel out faster lets you cool down the injectors faster. ![](https://www.miataturbo.net/images/smilies/smile.gif)
I have plans for solving this by compensating for the altered dead time in software, either by adding a circuit to monitor the current draw of the injector (and thus directly measure the dead time) or by proxying it with a thermocouple on the fuel rail.
--Ian
![](https://www.miataturbo.net/images/smilies/smile.gif)
I have plans for solving this by compensating for the altered dead time in software, either by adding a circuit to monitor the current draw of the injector (and thus directly measure the dead time) or by proxying it with a thermocouple on the fuel rail.
--Ian
#9
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
As I understand it, the injector works by using electricity going through a coil of copper wire to make an electromagnet to open the injector. Copper conducts electricity pretty well, but it still has non-zero resistance, and that resistance means that it takes a measurable amount of time for the magnetic field to build up in the coil. This is what the dead time is. The higher the resistance, the longer that time is, and as the temperature of the wires goes up, so does the resistance. So when the car is running normally, it's pumping fuel through the injectors, air through the manifold, and ambient air through the engine bay (either by moving or with fans). All of those keep the injector temperature pretty stable, so you don't notice any changes in dead time. When you park the car for 15 minutes, all of that stops, it heat soaks, the temperature of the injectors soars, the dead time goes up by a bunch, and all of your idle fuel values are totally wrong.
High end EV14 injectors are carefully matched and controlled to produce a dead time that's consistent for the set of injectors and is relatively short, but they can't eliminate the temperature dependence because that's inherent to the physics of copper. OTOH, the fact that they allow for very large injectors actually makes this problem worse, because large injectors have very short idle pulses, which means that any error in the dead time is an even larger percentage error in the total idle fuel.
Cooling down the injectors faster will alleviate the problem, and going to a return fuel system should help with that by transporting heat out of the rail and into the fuel that's going back to the tank. (This transport is why OEMs all moved to non-return systems about 20 years ago, because it increases evaporative emissions). It'll still happen on hot restart, but it will fix itself faster. You can also alleviate the problem by going with a manifold-referenced fuel pressure regulator, because that reduces the fuel pressure at idle, meaning longer pulses are required, meaning that the dead time change is a smaller percentage error.
IMHO the best fix is to measure the change in the dead time (either directly or proxy via fuel rail/injector temperature), feed that into the megasquirt, and have it incorporate that into the fuel calculations. AFAIK theres no feature in the MS code to do this right now, but it's a pretty straightforward feature to add if you're willing/able to hack the code.
--Ian
#13
Elite Member
iTrader: (2)
Join Date: Aug 2016
Location: New Fucking Jersey
Posts: 3,890
Total Cats: 143
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Maybe that should be the Begi fuel line slogan...
"Begi fuel lines... insurance is just one check away!"
#14
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
I think the issue is actually hot injector coil wires, but cycling the hot fuel out faster lets you cool down the injectors faster. ![](https://www.miataturbo.net/images/smilies/smile.gif)
I have plans for solving this by compensating for the altered dead time in software, either by adding a circuit to monitor the current draw of the injector (and thus directly measure the dead time) or by proxying it with a thermocouple on the fuel rail.
--Ian
![](https://www.miataturbo.net/images/smilies/smile.gif)
I have plans for solving this by compensating for the altered dead time in software, either by adding a circuit to monitor the current draw of the injector (and thus directly measure the dead time) or by proxying it with a thermocouple on the fuel rail.
--Ian
With a S/W solution, you can either alter dead time, or somehow utilize fake voltage correction. But how will you determine when the correction is needed?
I wonder if running the fan(s) after shut-down would keep the injector coils from over-heating? Anyone tried that?
Realistically, how often do we have hot restart, and when we do, it lasts about 20 seconds or a couple of rev-ups. It's not perfection, but it is not something that kills an engine or robs power.
For me, starting rich and letting EGO control AFR at idle has worked pretty well. Need extra ASE and ASE Time (cycles) from about 125 to 180F CLT.
#17
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
It seems likely that stock ECUs have some kind of mitigation in their software, but I'm not sure what it is. The narrowband sensor probably helps, as DNMakinson says, and there may be compensation tables as well. With a lot of empirical tuning & testing, you could build a table to look up a calculated dead time adjustment factor using things like the length of time since last run, ambient temperature, coolant temperature when shut down, etc. OEMs do a lot of this kind of tweaking.
--Ian
#19
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Stock injectors are a quarter of the size of my id1000s, so the idle pulse is 4x longer, which means that any fixed-time error is only 1/4 as significant. This is a problem associated with having a small number of large injectors.
It seems likely that stock ECUs have some kind of mitigation in their software, but I'm not sure what it is. The narrowband sensor probably helps, as DNMakinson says, and there may be compensation tables as well. With a lot of empirical tuning & testing, you could build a table to look up a calculated dead time adjustment factor using things like the length of time since last run, ambient temperature, coolant temperature when shut down, etc. OEMs do a lot of this kind of tweaking.
--Ian
It seems likely that stock ECUs have some kind of mitigation in their software, but I'm not sure what it is. The narrowband sensor probably helps, as DNMakinson says, and there may be compensation tables as well. With a lot of empirical tuning & testing, you could build a table to look up a calculated dead time adjustment factor using things like the length of time since last run, ambient temperature, coolant temperature when shut down, etc. OEMs do a lot of this kind of tweaking.
--Ian
#20
Senior Member
iTrader: (1)
Join Date: Dec 2010
Location: Farmington Hills, MI
Posts: 1,218
Total Cats: 175
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
ORB is a threaded port standard. So a Male SAE-8 ORB to Male AN-8 would be a fitting with male threads on one side (to connect to the female ORB fitting on the fuel rail) and a male AN flare fitting to connect a hose to. Or a Male SAE-8 ORB to Barb fitting would just be a hose barb adapter to thread into the fuel rail.
8AN ORB Fittings
See here, ORB is referring to the threads on the left side of the adapters in the pictures.
8AN ORB Fittings
See here, ORB is referring to the threads on the left side of the adapters in the pictures.