ZFuel

[Wade Nelson]

Ok, I'll be the guy to wizz on your parade, since I helped design early digital ECM's for Bosch...

Forget a plug and play L-jet replacement. And you already have the moment you discussed unplugging the AFM and plugging in a MAF. Should you decide you want your new ECM to also replace the (worn out) mechanical advance mechanisms in most Z-car distributors, you're already cutting wires and adding new connectors. Along with adding a fuel pump control (oil pressure based, etc?)

You'd simply have to make TOO MANY compromises to make a new, tunable ECM plug and play on this old warhorse. You're far better off to get out the wirecutters and do it right....

It's time for you to go for broke. Megasquirt.

MegaSquirt - Electronic Fuel Injection Computer by Bowling and Grippo ©2010

Megasquirt is an off the shelf DIGITAL fuel injection system. There's a community of thousands of users, tons of code, laptop monitoring, everything. You can buy a microcontroller, download some code, and be running in a week or two.

And there are already Z car owners running Megasquirt, building harnesses, and writing software. MADKAW here on this forum is apparently involved.

Your first task is to locate some guys actually daily driving some Z's running Megasquirt and road trip out to visit with them.

Because that's all you will do with Megasquirt; wire it in, and write code to optimize it for the size injectors you chose, brand of MAF, 02 sensor, etc. There are hundreds of code snippets already available for you to cut and paste. Tuning software. Everything under the sun for a do-it-yourself ECM guy.

If you want something to help and possibly to SELL TO other Z-car owners, come up with a complete Z- Megasquirt package, controller, optimized code, plugNplay wiring harness, replacement MAF, distributor kit, etc.

[FastWoman]

If you want something to help and possibly to SELL TO other Z-car owners, come up with a complete Z- Megasquirt package, controller, optimized code, plugNplay wiring harness, replacement MAF, distributor kit, etc.

This would work for me as well.

But what I would MOST like, and what I think many Z owners would also like, is an easy drop-in replacement. This would be a solution for someone who is too busy with other stuff to spend days tweaking their Z. It would need to handle I/O with existing equipment, such as the AFM, but it should ideally be expandable/improvable for when life is no longer in the way, so to speak.

And Wade points out something very important: It's not the antiquity of a distributor that's at issue with our ignition, but rather that our advance mechanisms are falling apart. The breaker plates typically get stuck and may randomly stick at various stages of advance. Rebuilt units are apparently getting scarce, and the rebuilding process comes with some risk of damage to the shaft. So to get a Z running "right," usually involves straightening out the ignition somehow. In my case, I was able to do it with a rebuilt OEM distributor. Many people upgrade to a Pertronix distributor/ignition system. I think it would also be feasible to go with a crank sensor and a wasted spark ignition. But what we should NOT do is to rely on the OEM distributor to provide an accurate timing input going into the future, barring some sort of modification like mechanically fixing the breaker plate in place (now there's an idea) and handing over all advance to the new ECU.

Edited July 22, 2013 by FastWoman

[FastWoman]

BTW, Wade, there is nothing wrong with the ECU box or the wires. The engine compartment connectors might be a bit crusty, but they're easily replaceable (another project). Nobody needs a completely new harness. An upgrade from AFM to MAP could simply involve the re-purposing of the AFM wires. That could be accompanied internally (inside the ECU box) by the shifting of a jumper connection or the substituting of an internal connector. That could be achieved in the engine compartment either with a plug-in adapter, for those terrified to make their own connections (such Z owners exist and are probably in the majority) or with a quick snip/rewire job for those comfortable working with wire. The adapter would be appealing even to someone like me, because it would keep the car reconfigurable to stock if needed. I don't like snipping wires, even when I do. And it's not because I'm even the slightest bit shy about electronics.

[superlen]

Hey, why are my cornflakes all soggy?

I agree Megasquirt is a perfectly acceptable solution to many. However, in my case I just like doing things from scratch & I think there are a few others that would like something similar to my concept of a Digital L Jet replacement. And while I will offer units for sale to the Z community, the primary reason for the project is because I personally want to design one.

Superlen

[superlen]

Just wanted to post an update for anyone interested.

I've been working on the software and it's coming along well. The GUI is to a state that I can read/write all the control structures to/from the target hardware & view pulse timing & sensor readings, the communications between the two is in place, and the embedded firmware that runs on the hardware is in place and running as well. For now, I'm using hardware that I already have from another current project & it doesn't have any of the signal processing electronics on it, so I'm just using a potentiometer to simulate the AFM and a signal generator for RPM to test *extremely* basic functionality. I'll hard code the other variables until I have the actual hardware reading real world sensors. The test system has a Freescale Kinetis K60 processor on it which is a bit of overkill for a simple ECU, but it makes an easy test platform.

The next few steps are:

- Hook up one of my stock LJets on the bench and start testing to get a rough transfer function of it's AFM/RPM/IAC ->pulse width.

- Finish the schematic (More realistically, just call it finished as I keep wanting to add more features for later options)

- Layout the pcb

I have been doing a crapload of research on EFI tuning of speed-density and MAF systems. For anyone interested, here are some pretty good books that have a lot of good information:

"How to Tune and Modify Engine Management Systems" by Hartman

"Engine Management - Advanced Tuning" & "Designing & Tuning High Performance Fuel Injection Systems" both by Greg Banish.

When I first started this project in the 90s all I had was a book on the Bosh systems and it also had some great information, but I forget the title and have since loaned/gave the book to a friend. It covered L-Jet great, but of course was very dated in it's overall engine management overview.

Of the above, all are good. Both the Banish books cover basically the same thing & are a little more current than Hartman. Hartman has a little more total information & some of the explanations in Hartman are really good. In particular, Hartman has some great info on intake manifold design/tuning & how it effects VE. (Not that this helps at all with Zfuel, but it's a nice rabbit trail to follow)

One thing I'm anxious to do is see how my calculated pulse widths based on MAF & the Zcars Stock engine/injectors match the pulse widths measured on both the car and the test bench. In theory they should be very close, but bear in mind, the ECMs I have to compare to are 40yrs old, likewise for the AFMs.

More to come, and as always any questions or comments are welcome. Feel free to critique, hassle, praise, call BS, encourage (a lot of that would be good), or ask technical details.

Lenny

[superlen]

Just wanted to post an update for anyone interested.

I've been working on the software and it's coming along well. The GUI is to a state that I can read/write all the control structures to/from the target hardware & view pulse timing & sensor readings, the communications between the two is in place, and the embedded firmware that runs on the hardware is in place and running as well. For now, I'm using hardware that I already have from another current project & it doesn't have any of the signal processing electronics on it, so I'm just using a potentiometer to simulate the AFM and a signal generator for RPM to test *extremely* basic functionality. I'll hard code the other variables until I have the actual hardware reading real world sensors. The test system has a Freescale Kinetis K60 processor on it which is a bit of overkill for a simple ECU, but it makes an easy test platform.

The next few steps are:

- Hook up one of my stock LJets on the bench and start testing to get a rough transfer function of it's AFM/RPM/IAC ->pulse width.

- Finish the schematic (More realistically, just call it finished as I keep wanting to add more features for later options)

- Layout the pcb

I have been doing a crapload of research on EFI tuning of speed-density and MAF systems. For anyone interested, here are some pretty good books that have a lot of good information:

"How to Tune and Modify Engine Management Systems" by Hartman

"Engine Management - Advanced Tuning" & "Designing & Tuning High Performance Fuel Injection Systems" both by Greg Banish.

When I first started this project in the 90s all I had was a book on the Bosh systems and it also had some great information, but I forget the title and have since loaned/gave the book to a friend. It covered L-Jet great, but of course was very dated in it's overall engine management overview.

Of the above, all are good. Both the Banish books cover basically the same thing & are a little more current than Hartman. Hartman has a little more total information & some of the explanations in Hartman are really good. In particular, Hartman has some great info on intake manifold design/tuning & how it effects VE. (Not that this helps at all with Zfuel, but it's a nice rabbit trail to follow)

One thing I'm anxious to do is see how my calculated pulse widths based on MAF & the Zcars Stock engine/injectors match the pulse widths measured on both the car and the test bench. In theory they should be very close, but bear in mind, the ECMs I have to compare to are 40yrs old, likewise for the AFMs.

More to come, and as always any questions or comments are welcome. Feel free to critique, hassle, praise, call BS, encourage (a lot of that would be good), or ask technical details.

Lenny

[LeonV]

Very cool project! I haven't really been following as I don't have a 280Z but sounds like you've put forth a TON of effort and research. If I owned a 280Z or ZX with stock EFI, I'd be pretty damn excited!

Are you designing and coding the tuning software/interface as well?

[superlen]

Leon,

Thanks.

Yes, I'm writing the Windows Tuning software interface as well. That interface pretty much goes hand in hand with the firmware as it makes a handy real time debug interface, and in reality the guts of firmware that will ultimately run on the embedded board, have been written up on on the windows platform as well. The two processes just communicate directly through dummy function calls as opposed to the usb link. When simulating I just have a screen that lets me set things like RPM, MAP, IAT, CLT, ect. with sliders and up/down arrows and such.

For now, the communication protocol is just a derivative on one that I use all the time for numerous projects. I'll publish it in case someone wants to write their own software interface. In the future (after the basic system is done) I will add in OBD2 so ANY tuning software can at least pull a majority of the sensors and display them. I'm not sure if OBD2 specifies a format for maps ( I don't think it does - someone please correct me or point me in the right direction if you know), or if there is a default communications protocol for this. I mostly thought the OBD2 would be nice for someone if they already had a digital dash type app that they liked using (plus I wanted to play with can some more).

I saw in your tag line you have 2 260's. My first z when I was 14 as a 74-260. I miss it, but not those flat top carbs. grrrr... You need to take one of your 260's and add a Zfuel to it, then tune tune tune.

Lenny

[madkaw]

Somehow I missed this thread until now. This is very interesting and waayyy over my head as far as understanding code and all that stuff..

I wish you luck. I have purchased the MS3X for my car being it was the closest to plug and play-and it's a far cry from it. But, I wanted an Standalone system that I didn't have to actually break into the ECU and make changes. This will be primarily hooking up wires and tuning.

I would also love to see dyno comparisons between the stock EFI and something like you are developing or MS.

[superlen]

Madkaw,

This one's primary goal is 100% plug and play....and then if you want you can start fiddling with it and tune it till the cows (kaws in this case) come home. I too would like to see the dyno diff & will have a few for sure before this is over with. Realistically, I don't think there will be a huge performance gain with the stock motor. At least if your ECU and AFM are both in spec. However, if you change one of our EFIs to even a mild cam you're basically screwed in getting everything back to perfect. The MS you have or Zfuel will be able to handle all that and more.

I'm working on a screen shot to upload. I'll try that in just a bit.

Lenny

[madkaw]

Plug n play would be huge - optimistically watching!

[LeonV]

Leon,

Thanks.

Yes, I'm writing the Windows Tuning software interface as well. That interface pretty much goes hand in hand with the firmware as it makes a handy real time debug interface, and in reality the guts of firmware that will ultimately run on the embedded board, have been written up on on the windows platform as well. The two processes just communicate directly through dummy function calls as opposed to the usb link. When simulating I just have a screen that lets me set things like RPM, MAP, IAT, CLT, ect. with sliders and up/down arrows and such.

For now, the communication protocol is just a derivative on one that I use all the time for numerous projects. I'll publish it in case someone wants to write their own software interface. In the future (after the basic system is done) I will add in OBD2 so ANY tuning software can at least pull a majority of the sensors and display them. I'm not sure if OBD2 specifies a format for maps ( I don't think it does - someone please correct me or point me in the right direction if you know), or if there is a default communications protocol for this. I mostly thought the OBD2 would be nice for someone if they already had a digital dash type app that they liked using (plus I wanted to play with can some more).

I saw in your tag line you have 2 260's. My first z when I was 14 as a 74-260. I miss it, but not those flat top carbs. grrrr... You need to take one of your 260's and add a Zfuel to it, then tune tune tune.

Lenny

Well keep up the awesome work!

Fortunately, both the 260s have triple Webers.