Official Carb to Fuel Injection Thread. (machinist and fabricator input very welcome)

If I were doing this, I would simply use a Micro Squirt system.
Cheap. Clean. Easy. Tons of support. They even have their own forums where other MS users can offer tips hints and support as well.
And it can be used for just about anything on any size of engine, with almost any number of cycls etc. Turbo? Nitrous? Big bore/stroke? No problems. Everything is infinitely variable so it can accomodate you.

I am going to be using one on my 16 valave 2.4 turbo SRT motor in my carbed 1978 Dodge Omni. Gunning for 400+ FWHP, and at least 11 second 1/4 miles.
 
Freestyle72 said:
Can I just thread the fuel rail and put those fittings on? That might actually be easier/safer than brazing for me. And it would also offer me repeatability if I wanted to do this again, or another bike.
Click to expand...

The wall thickness on the OEM rail is usually too thin to cut threads in. I have done this Carb-FI a few years back. I used aluminium tube for the rail, welded threaded AN bungs on the ends and AN fittings to connect the rail, pump. Had to machine and weld a flange to mate the FP regulator.

Used DTA S60 ECU to run the engine - it had many tuning options and the price reflected that.

You will need to learn to TIG aluminium or make friends with someone who is capable.

Also your target of 20:1 AFR is going to put you close (or right into) lean misfire territory. It will be difficult to tune when running that lean, especically the throttle transients. Maybe try something more conservative like 16-18 AFR steady state.

If you have time, try reading "Internal combustion engines" by Richard Stone. It would have made my life a lot easier if I read it before embarking on a clean sheet fuel/ignition map. It graphically illustrates things like lean misfire limit vs. throttle opening - which is handy to know when you are trying to push limits when tuning the maps.
 
I was just drawing your attention to the fact that your grab bag of parts might have different fittings, and you may not be able to re-use the factory fuel lines from whatever bike.

With carbs its no problem, you just use barbed fittings, but I would hesitate to used barbed fittings on rubber hose at 50 psi with fuel...for testing ok, but it would be sketchy on the bike running day after day. I'm sure people do it, but I wouldn't have a lot of confidence in it...

The AN fittings are typically aluminum, but they're also made in steel and cres. Maybe take a fitting, and turn it down to the ID or bore to the OD of the rail and braze it? Or make a shouldered adapter on the lathe and weld or braze that.

Or, perhaps make a frankenstein fuel rail, braze the pieces together, and use the factory lines.

http://www.holley.com/types/Steel AN to AN Adapters.asp
 
Also, air-fuel ratio has to be slightly rich (usually 13:1 or thereabouts) when the engine approaches full load at any RPM. At a minimum, running lean near full load will cost you horsepower. More likely, running lean at higher revs near full load for an extended time is asking for burned pistons and exhaust valves.

At part throttle cruise, fuel consumption and emissions are more important. Running unnecessarily rich wastes fuel and it can also help to wash oil off of cylinder walls leading to higher piston ring wear. Best fuel consumption on most engines at part load is usually with air/fuel around 16:1 to 17:1 at the outside. Leaner than 18:1 is asking for misfire. Mapping for cold start and warm-up is trickier the leaner you go.

Some engines are capable of running leaner than that, but normally only by using direct-injection (nowadays) or careful placement of injectors and injection timing and airflow within the engine so as to create a stratified-charge effect in the cylinder. The old Honda Civic HF engine, and that of the original Honda Insight (not the current one) were like this. Sequential injection - and a lot of mapping experimentation - is necessary for doing that.
 
Ok so it looks like we are back to brazing the stock line lol. I don't really want to make a custom fuel rail because that is more money, and certainly more effort than I really want to go through for little or no return when compared to just brazing in a little section for the line.

Any other thoughts on how to extend this fuel rail?

BTW I am retarded I have no idea what I was typing when I said 20:1. I meant an extra rich mixtures like 10:1 just to make sure she starts and stuff at first. LOL
 
Remember, you can do anything you want to an engine or a bike, swap this or that in, change it from Carb to FI, supercharge, the possibilities are endless.

However, big projects require big budgets, and it sounds like you have a little budget, if at all.

Cough up the money where it is needed, bodgery equals failure.
 
Ah, here's your project. Let me know if you need any help. I would seriously consider getting a UEGO of your own. A cheaper LSU4 controller. If this is your first time it is extremely helpful in seeing what is happening in real time. Logging while long term riding lets you make up much nicer correction maps.

I'm not certain what you mean about angling the adapter plate, but proper placement of primary injectors for sequential injection is facing the valve opening as much as possible and fairly close.
 
Last edited:
Ok so I have gotten my focus back on this now.
I figure a wideband O2 sensor will do me good whether this conversion is successful or not so I will start with buying one of those.

I think the Innovate LC1 is the most appropriate wideband for me because it plays nice with megasquirt. Any difference between the Innovate LC1 and the AEM UEGO? They are both very close in price, but I can't really tell the difference in terms of features etc.

Also should I get a set with a guage?

For widebands with gauges I am looking at these two:
http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&item=370369909145&viewitem=&sspagename=STRK%3AMEWAX%3AIT
http://cgi.ebay.com/ebaymotors/ws/e...0436678037&viewitem=&sspagename=STRK:MEWAX:IT

Withought a guage:
http://cgi.ebay.com/ebaymotors/ws/e...0566137403&viewitem=&sspagename=STRK:MEWAX:IT

So about $60 difference.

What do you guys think? Suggestions?
 
Freestyle72 said:
Ok so I have gotten my focus back on this now.
I figure a wideband O2 sensor will do me good whether this conversion is successful or not so I will start with buying one of those.

I think the Innovate LC1 is the most appropriate wideband for me because it plays nice with megasquirt. Any difference between the Innovate LC1 and the AEM UEGO? They are both very close in price, but I can't really tell the difference in terms of features etc.

Also should I get a set with a guage?

For widebands with gauges I am looking at these two:
http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&item=370369909145&viewitem=&sspagename=STRK%3AMEWAX%3AIT
http://cgi.ebay.com/ebaymotors/ws/e...0436678037&viewitem=&sspagename=STRK:MEWAX:IT

Withought a guage:
http://cgi.ebay.com/ebaymotors/ws/e...0566137403&viewitem=&sspagename=STRK:MEWAX:IT

So about $60 difference.

What do you guys think? Suggestions?
Click to expand...

The sensor on its own, without the gauge, doesn't do you any good unless you are using an EFI controller that has wide-band support built into it. The gauge (or the EFI controller) contains the drivers necessary to make the sensor function, and without one or the other, the sensor is a paperweight. Buy either one of the two kits with the gauge. Having a gauge like this works wonders for setting up either carbs or fuel-injection.

The AEM is the exact one that I have. But for the same price, if the Megasquirt stuff is designed to work with the Innovate gauge and you are planning to use Megasquirt in the future, go that route. One less uncertainty to deal with.

I know that the AEM gauge has a narrow-band lambda sensor emulator output, so that you can give that signal to an automotive-style ECU that is expecting a narrow-band lambda sensor input, and you can program it to give that emulated signal at any air/fuel ratio you care to use ... but it's a feature that I have not played with at all.
 
Brian P said:
The sensor on its own, without the gauge, doesn't do you any good unless you are using an EFI controller that has wide-band support built into it. The gauge (or the EFI controller) contains the drivers necessary to make the sensor function, and without one or the other, the sensor is a paperweight. Buy either one of the two kits with the gauge. Having a gauge like this works wonders for setting up either carbs or fuel-injection.

The AEM is the exact one that I have. But for the same price, if the Megasquirt stuff is designed to work with the Innovate gauge and you are planning to use Megasquirt in the future, go that route. One less uncertainty to deal with.

I know that the AEM gauge has a narrow-band lambda sensor emulator output, so that you can give that signal to an automotive-style ECU that is expecting a narrow-band lambda sensor input, and you can program it to give that emulated signal at any air/fuel ratio you care to use ... but it's a feature that I have not played with at all.
Click to expand...

So what you are saying if I just put the $60 towards my microsquirt I can do that and log the info that way? Might actually be better since I can feed megasquirt from the ignitor box so it knows RPMs vs AFR. Am I thinking correctly about this?
 
I have the LC-1 without the gauge. I'm making up a 4 digit 7 segment LED display for it myself. It's not that hard and the spec for the protocol is on the innovate site. I plan to make it open source.

Unfortunately I thought the connector was a 3.5MM audio plug, so the jacks I ordered wont work. The connector is actually a 2.5mm connector and bare jacks in that size are not that easily available. So I'm trying to figure something out.

The LC-1 has a fully programmable analog output, so you can make it emulate narrow band or just about anything you want.
 
Go with the innovate. The latency is a quarter of the AEM (100ms vs 400ms). Both are decent (unlike some other kits which are basically pretty lights that feed you bad information).

See http://tunertools.com/articles/FordMuscle.pdf

I bought the Innovate w XD-16 for a car last year and it works well. I got it shipped from injected performance in the states. A company in Toronto stocks them but wanted almost $200 more. I went w the XD-16 so I can use one gauge for multiple inputs if I want (and the colour changing led's w mixture help for corner of the eye tuning).
 
TheDirty said:
I have the LC-1 without the gauge. I'm making up a 4 digit 7 segment LED display for it myself. It's not that hard and the spec for the protocol is on the innovate site. I plan to make it open source.

Unfortunately I thought the connector was a 3.5MM audio plug, so the jacks I ordered wont work. The connector is actually a 2.5mm connector and bare jacks in that size are not that easily available. So I'm trying to figure something out.

The LC-1 has a fully programmable analog output, so you can make it emulate narrow band or just about anything you want.
Click to expand...
Leave it to you to pull something like that! Make two and I'll pay for both yours and mine! (depending on cost)

GreyGhost said:
Go with the innovate. The latency is a quarter of the AEM (100ms vs 400ms). Both are decent (unlike some other kits which are basically pretty lights that feed you bad information).

See http://tunertools.com/articles/FordMuscle.pdf

I bought the Innovate w XD-16 for a car last year and it works well. I got it shipped from injected performance in the states. A company in Toronto stocks them but wanted almost $200 more. I went w the XD-16 so I can use one gauge for multiple inputs if I want (and the colour changing led's w mixture help for corner of the eye tuning).
Click to expand...
This is the exact kind of info I was wondering about. Thanks a lot man. Innovate it is.
 
If I were trying to setup and tune a new FI setup I would want a heck of a lot more info than just AF ratio.

I use a Innovate LM2 with a SSI.
2x O2, map,3x rpm, 2x fuel pressure... 32 inputs.
Not a lot more money than a LC1... and it outputs to a SD card or a computer.

I buy Innovate from Bob Perkin's auto 905 565 9947. He has an 'in" at Motorstate and can usually undercut most prices... and he's one of my sponsors.
....oh and Innovate has no stock on LC1.
 
Last edited:
bitzz said:
If I were trying to setup and tune a new FI setup I would want a heck of a lot more info than just AF ratio.

I use a Innovate LM2 with a SSI.
2x O2, map,3x rpm, 2x fuel pressure... 32 inputs.
Not a lot more money than a LC1... and it outputs to a SD card or a computer.

I buy Innovate from Bob Perkin's auto 905 565 9947. He has an 'in" at Motorstate and can usually undercut most prices... and he's one of my sponsors.
....oh and Innovate has no stock on LC1.
Click to expand...

I'll give him a call. If he can come close to the ebay price I would be more than happy to give him my business. If you want you can shoot me a PM with your name if you want me to mention it or something lol.
 
If the Microsquirt has capability to interface with that sensor then you don't NEED a gauge. But there is something to be said for seeing something happening in real time.

If you have a gauge that you can see, then as you are riding along and you feel a hesitation or stumble, you shoot a glance at the gauge and say, in real time, "Aha - we have a problem at 4500 rpm and around an eighth throttle or thereabouts and it needs a little more fuel there".

If you have data logging but no gauge then you are going to be searching through reams of data logs afterwards for that condition. "let's see, it was around 20 minutes into the ride, or maybe 25? dunno ..." but once you DO find that data point, at least you will know the exact throttle position (if it's being logged) instead of guessing at it with the gauge.

If you have BOTH data logging AND a gauge, you are all set.

If it was me, I'd spend the sixty bucks. It will save you more than sixty bucks worth of aggravation and headaches. And once you have a gauge, you can use it on ANY engine, even one with carbs or one with EFI that doesn't have data-logging capability, not just the specific one that you're working on right now.

Obviously some folks are making their own displays, personally I wouldn't be going down that route because (A) my knowledge of component level electronics is so close to zero that one might as well call it zero, and (B) even if I knew something about it, the time that it would take to create something instead of buying something ready made is worth more than sixty bucks worth of my time, and (C) if you have a gauge that's meant for that sensor then you have something that you know is working properly (or not) - there is less uncertainty to deal with.
 
Brian P said:
If you have data logging but no gauge then you are going to be searching through reams of data logs afterwards for that condition. "let's see, it was around 20 minutes into the ride, or maybe 25? dunno ..." but once you DO find that data point, at least you will know the exact throttle position (if it's being logged) instead of guessing at it with the gauge.
Click to expand...
I also like having a display, but it's not hard at all to find issues in datalogs. Some simple Excel macros can graph a full datalog so you can see rich/lean points. I made up some simple programs to show maps with rich/lean trends dividing points up into transients, steady state, and temperature. Quick finds of one off problem points are not hard either to find and troubleshoot. That's all pretty useless though if you are not datalogging all the time.

Maybe I'll make a display with through hole components so that people can make it themselves, but for now, my first few will be SMT parts that generally people aren't going to put together themselves anyway. If it works out and people want them, I'll get a few made. I can understand the hesitation about trusting a display not made by the manufacturer, but it really is pretty simple and it's just decoding serial data sent by the unit. There's not much to fail with it.

I've made two similar displays a few years ago for other aftermarket ECU's. The PCB for this one will be much smaller though, and I figure I'd leave out an enclosure. Any enclosure I can make would be crap and with the bare 4 digit display people can custom mount it in dashes or wherever they want.
 
You must log in or register to reply here.

Back
Top Bottom