EFI conversion on a Gen-1

[RempageR1]

Another update on the project: I`ve been driving with the V1200i last weekend and drove around 900KM. I had the carburetors and fuel pump with me as a backup but luckily I didn't have to use them .

The system worked perfect, with one exception on the day I set off to drive home. Then, after 500m it stumbled and died. After turning the ignition off and on, and starting again it worked as it should. Although I didn't have the logging enabled at that time, it`s likely due to the rather poor signal I get from the crank. As shown before, I get two pulses one of which needs to be filtered out to make just one. This gives me just one pulse per rotation which is not ideal. Therefore I`m looking for a `91 or newer flywheel. This has six lobes so you get six pulses per rotation which is much better. If anyone has one laying around please let me know.

Then there`s the fuel map, which looks something like this:


Nothing strange, except over 80% throttle where it seems to require a huge amount of fuel around 3000 RPM, and then it requires way less around 4500 RPM. If I time how quickly the revs rise, the 100% TPS position is the quickest so something must be right. I`m unsure what happens with the fuel load though.

Another thing is the Vboost. Since the throttle bodies have a way larger diameter, the VBoost is only noticeable with the throttle on 40% or less. So effectively there is little benefit of leaving it on. To address the fuel load 'strangeness' above and to perhaps give it more power, I will over winter create some new manifolds which are larger in diameter and without the Vboost.

Will update on the progress

 

[VMAXXIMUM]

Another update on the project: I`ve been driving with the V1200i last weekend and drove around 900KM. I had the carburetors and fuel pump with me as a backup but luckily I didn't have to use them .

The system worked perfect, with one exception on the day I set off to drive home. Then, after 500m it stumbled and died. After turning the ignition off and on, and starting again it worked as it should. Although I didn't have the logging enabled at that time, it`s likely due to the rather poor signal I get from the crank. As shown before, I get two pulses one of which needs to be filtered out to make just one. This gives me just one pulse per rotation which is not ideal. Therefore I`m looking for a `91 or newer flywheel. This has six lobes so you get six pulses per rotation which is much better. If anyone has one laying around please let me know.

Then there`s the fuel map, which looks something like this:
View attachment 79296

Nothing strange, except over 80% throttle where it seems to require a huge amount of fuel around 3000 RPM, and then it requires way less around 4500 RPM. If I time how quickly the revs rise, the 100% TPS position is the quickest so something must be right. I`m unsure what happens with the fuel load though.

Another thing is the Vboost. Since the throttle bodies have a way larger diameter, the VBoost is only noticeable with the throttle on 40% or less. So effectively there is little benefit of leaving it on. To address the fuel load 'strangeness' above and to perhaps give it more power, I will over winter create some new manifolds which are larger in diameter and without the Vboost.

Will update on the progress

Sean would probably be your best bet for the flywheel.

 

[02GF74]

If you cannot source one, an option would be to weld on the projections, ensuring that they are diametrically opposite to keep the flywheel in balance. A square piece of steel with a hole plug welded on and you could measure up the exact position you want them.

 

[WILLRIDEAGAIN]

Another update on the project: I`ve been driving with the V1200i last weekend and drove around 900KM. I had the carburetors and fuel pump with me as a backup but luckily I didn't have to use them .

The system worked perfect, with one exception on the day I set off to drive home. Then, after 500m it stumbled and died. After turning the ignition off and on, and starting again it worked as it should. Although I didn't have the logging enabled at that time, it`s likely due to the rather poor signal I get from the crank. As shown before, I get two pulses one of which needs to be filtered out to make just one. This gives me just one pulse per rotation which is not ideal. Therefore I`m looking for a `91 or newer flywheel. This has six lobes so you get six pulses per rotation which is much better. If anyone has one laying around please let me know.

Then there`s the fuel map, which looks something like this:
View attachment 79296

Nothing strange, except over 80% throttle where it seems to require a huge amount of fuel around 3000 RPM, and then it requires way less around 4500 RPM. If I time how quickly the revs rise, the 100% TPS position is the quickest so something must be right. I`m unsure what happens with the fuel load though.

Another thing is the Vboost. Since the throttle bodies have a way larger diameter, the VBoost is only noticeable with the throttle on 40% or less. So effectively there is little benefit of leaving it on. To address the fuel load 'strangeness' above and to perhaps give it more power, I will over winter create some new manifolds which are larger in diameter and without the Vboost.

Will update on the progress

Keep up the good work

 

[Sidecarjohn]

Admire your commitment to this project. Beyond impressive.
Look forward to further updates in the future.

 

[RempageR1]

Just a small update for today. Modifying the lobes on the old flywheel might work but I wonder how to balance it afterwards. Luckily, after doing some calls I found a good flywheel and pick-up.

And when it came in, it was time to program the Speeduino with the following pattern:


After some days of programming, it seems to work on the bench so the next step was trying to get it to work on the bike.

But before doing that, I had to reprogram the Ignitech with the config used for the 1990+ bike with one pick-up and install the new hardware:


After a job well done, I tried to start the bike tonight and it wouldn't start. After messing with some settings it ran very rough and backfired like crazy. Looking down into the throttle bodies to identify the backfiring cylinder might not be the best way to troubleshoot:


So I compared the wiring instructions for the Ignitech of the 85-89 (four pick-ups) with 90+ (one pick-up) and found out the Ignitech changed the wiring order for the HT coils for both models:
Orange wire: Old: Channel 1 New: Channel 2
Yellow wire: Old: Channel 3 New: Channel 4
Grey wire: Old: Channel 2 New: Channel 1
White wire: Old: Channel 4 New: Channel 3
Funny guys......

So after changing the firing timing in the config, all was well and it starts now with the new flywheel. Next up is testing (and likely) debugging the trigger in the Speeduino config so the Speeduino can work with the Vmax new flywheel.

To be continued....

 

[RempageR1]

Hi there!

As mentioned above, I had written the trigger for the Speeduino and I wanted to share what I did. Basically, the flywheel looks like this:



To make a trigger I have to sync on the upward edge (beginning of a lobe). U1 is TDC of cylinder one, and U2 is 70 degrees further. From thereon, it`s 40, 70, 70, 40 and 70 degrees. However there is a challenge. To sync, I have to use the wider lobe, but if you see that one, you`re already at D1 so you`re too late for TDC of cylinder 1. One option is to wait until you see U1 for the second time, but that would result in a slower trigger sync, one extra rotation of the crank.

Therefore, I`ve written code that can sync on the U1-D1 width, but then treats U2 as the first lobe so you at max have to wait for 1 rotation of the crank before the injectors start firing. Then in config, the lobe 1 is set at 70 degrees after TDC so it is working but also exactly synced. I tested this by configuring the Speeduino with the same basic crank advance and then compare the ignition pulses of Cylinder 1 from the Ignitech and the Speeduino. They exactly match

With that, I have six pulses per revolution instead of one, so happy days. Also with this much better resolution the Speeduino can also be used for ignition in case I decide to go that way .

New flywheel code attached might someone need it some day.

 

[RempageR1]

Hi,

I thought I added a bit more theory around the flywheel lobes since I`ve spent way too much time on that . To understand them, the first thing needed is to understand the firing order. The Vmax is a V-4 which is at 70 degrees, not 90. That makes it fire a bit unevenly. Below, you can see the firing order:

So first, the rear cylinders fire, no. 1 and 3 at 180 degrees apart (half a turn of the crank). Next, there is a pause of 70 degrees (we go from the rear to the front cylinders) and then cylinders 2 and 4 fire. This sequence takes a total of 2 rotations.



Below is the crank again and on the right, the Ignitech configuration for the bike. Channel 1 is cylinder 1, Channel 2 is cylinder 2 and so forth and the flywheel rotates counter-clockwise.

As you can see, Ignitech uses the same lobe numbering as I did. So after sync (the wide lobe is seen), the first lobe is 1 which fires channel 2 (see start advance). This is cylinder 2 (front left). The next one is channel and cylinder 3 on lobe 3, then cylinder 4 on lobe 4 and finally cylinder 1 at lobe 6. Since the setup is wasted spark (we have 1 spark at TDC befire firing stroke, and 1 spark at TDC before inlet stroke), the forward cylinders will fire at the firing stroke with the rear ones on inlet stroke or vice versa.
Finally, there is a security feature built into the flywheel. To ensure the spark is not advanced too much, it may never fire before a certain lobe. So for channel 1, with 0 degrees of advance, it would fire at the beginning of lobe 6. Maximum advance is the end of lobe 5, giving a maximum advance of 70 degrees. Channel 2 would fire at 0 degrees at lobe 1 and the maximum advance is at the beginning of lobe 6, again 70 degrees. Please note the '2nd Edge' is not checked here or the max advance would be around 25 degrees.

So to summarize:
Lobe 1 is the firing point without advance for cylinder 2
Lobe 2 is the max advance for cylinder 3.
Lobe 3 is the firing point without advance for cylinder 3 and max advance for cylinder 4.
Lobe 4 is the firing point without advance for cylinder 4
Lobe 5 is the max advance for cylinder 1.
Lobe 6 is the firing point without advance for cylinder 1 and max advance for cylinder 2.

Now you know

 

[one2dmax]

I'm gonna have to save all this info somewhere lol!

 

[Fire-medic]

Thank-you for allowing the open-code, as I'm sure that it will benefit anyone who wants to try to use the findings.

 

[DreamV4]

I hope you can make money selling the bike to compensate your effort. What is your estimate to convert somebody's bike to FI? How much would you charge?

 

[redneksoldier]

I had been planning to study the flywheel triggers in depth the next time I had the cover off. Mostly just to make sense of the program. Thank you for taking the time to figure it out and share your findings. That graphic makes it very easy to understand.

 

[RempageR1]

Thanks all for the support! Much appreciated.

I hope you can make money selling the bike to compensate your effort. What is your estimate to convert somebody's bike to FI? How much would you charge?

It`s still all just prototype today, perhaps I will make it into a proper kit one day. But before that day comes it needs to be bullet-proof. And going from prototype to bulletproof is a lot of work. Also, I would need someone to fabricate some parts to make it look more professional. After that, I would need someone to build the parts in the US and distribute since shipment from the Netherlands would be costly.

I`ve learned a lot over this summer, one of the things is that the O2 sensor in de exhaust doesn't work. It`s too far away from the front cylinders so it has too much latency. That results in the Speeduino compensating too late so the mixture oscillates from lean to rich and so on. I`ll probably relocate the O2 sensor one of the rear cylinders` downpipes. It`s not ideal to measure off of just one cylinder but it will likely work better than what I have now. I`ve spend quite some time modifying the original exhaust but you live and learn.... Also if it works on a downpipe it`s an easier installation for other people that do not run original mufflers so that`s a positive thing

This winter, I`m planning to create new manifolds, something like this:





If I can make it look half as cool as that them I`m very happy. More details on the dutch guy that did that build can be found here:
https://www.saarloos.net/megasquirt-projects/yamaha-vmax-msefi.html
Now a question for you guys: As you know, my throttle bodes are bigger than the original and they are also bigger than the inlet diameter on the cylinder heads. Would it be better to make something that is tapered? Full diameter until the cylinder head? Or reduce the diameter immediately? I hope someone can steer me in the right direction on this one. Thanks!

 

[02GF74]

Now a question for you guys: As you know, my throttle bodes are bigger than the original and they are also bigger than the inlet diameter on the cylinder heads. Would it be better to make something that is tapered? Full diameter until the cylinder head? Or reduce the diameter immediately? I hope someone can steer me in the right direction on this one. Thanks!

Tapered.

Air wants to flow in s straight line unimpeded, so steps resulting from a change in diameter are to be avoided.

My friend fitted throttle bodies on a Ford crossflow engine in a Caterham.

You can buy manifolds that consist of straight pipes cut and welded, not ideal.

What he did was make a wooden former that was tapered as well as curved. Around the former he bashed a piece of aluminium sheet (I don't know the thickness) then had someone weld it up and weld the flanges on.

CNC would do it too but that would be another step up in cost.

Depending on the location of the inlet ports and the throttle bodies you may need to make 4 different pieces, or get lucky with making 2 pairs of 2.

 

[DreamV4]

Parts would be very cheap if I could use parts from Toyota Camry 4 cyl (about same power). I just remembered something interesting: 86 Chevy Celebrity used electrically heated grill under carburetor. If such grill is used, you could use 2 central injector set up from 95 (?) 3.6 Chevy Caprice

 

[RempageR1]

Tapered.

Air wants to flow in s straight line unimpeded, so steps resulting from a change in diameter are to be avoided.

My friend fitted throttle bodies on a Ford crossflow engine in a Caterham.

You can buy manifolds that consist of straight pipes cut and welded, not ideal.

What he did was make a wooden former that was tapered as well as curved. Around the former he bashed a piece of aluminium sheet (I don't know the thickness) then had someone weld it up and weld the flanges on.

CNC would do it too but that would be another step up in cost.

Depending on the location of the inlet ports and the throttle bodies you may need to make 4 different pieces, or get lucky with making 2 pairs of 2.

Thank you! Yes it makes perfect sense. It might also cause turbulence if you immediately go from one to the other diameter.

As metalworks it not my strong point, I don't think it would look the part if I fabricated it myself. However there are lots of reducers availabe online so I might spend some winter evenings looking for the right ones:

 

[RempageR1]

Parts would be very cheap if I could use parts from Toyota Camry 4 cyl (about same power). I just remembered something interesting: 86 Chevy Celebrity used electrically heated grill under carburetor. If such grill is used, you could use 2 central injector set up from 95 (?) 3.6 Chevy Caprice

I`m sorry to say I`m not following you. I have googled the cars you refer to but could not find any part that looked compatible and usable for what I`m doing. Do you have some example picture perhaps?

 

[Parminio]

It`s too far away from the front cylinders so it has too much latency. That results in the Speeduino compensating too late so the mixture oscillates from lean to rich and so on. I`ll probably relocate the O2 sensor one of the rear cylinders` downpipes.

Just an observation here: many cars have at least 2 O2 sensors. Typically there's one before the catalytic and one after. Some cars have up to 5 of them.

I'm wondering if your system might benefit from that sort of setup; at least two sensors. One at or near the bottom of a front downpipe and one closer to the rear.

 

[DreamV4]

Just an observation here: many cars have at least 2 O2 sensors. Typically there's one before the catalytic and one after. Some cars have up to 5 of them.

I'm wondering if your system might benefit from that sort of setup; at least two sensors. One at or near the bottom of a front downpipe and one closer to the rear.

Upstream O2 sensor runs fuel delivery, downstream sensor only checks catalytic converted. Most cars have separate fuel management for each cylinder head.
Some Grand Cherokees I6 have separate fuel management for front 3 and rear 3 cylinders, they have 2 upstream and 2 downstream o2 sensors.

 

[RempageR1]

The Speeduino could handle 2 O2 controllers in code but i`m not sure if the signals could be combined. With a four-in-line with one manifold is much easier to install the O2 sensor.

I could run 1 O2 off of the rear right cylinder and run 1 in the exhaust to see if there isn't a large deviation between that one cylinder and all four. But let`s not get carried away