1982 CBX EFI PROJECT

Hey, what projects are you planning or preparing for? CBX, other motos, workshop, WHATEVAH!
EMS
ICOA Member
ICOA Member
Posts: 10151
Joined: Wed Mar 05, 2003 7:55 am
Location: North East OH, ICOA 3904

Re: 1982 CBX EFI PROJECT

Post by EMS »

In the CBX, you get the signal from the ignition pulser. It involves making a notched disk for the pick-up. But this is the cleanest solution.

User avatar
mlynch001
ICOA Member
ICOA Member
Posts: 20
Joined: Tue Dec 13, 2016 10:42 am
Location: Arkansas
Location: Dardanelle, Arkansas

Re: 1982 CBX EFI PROJECT

Post by mlynch001 »

Syscrush wrote:
mlynch001 wrote:Please Feel free to ask anything, I will help as am able.
That means a lot to me, thanks.
I have zero experience with shower type injectors, and most of the OEM injection systems use an injector below the throttle plate. I would be concerned with the space required and the overall difficulty of packing all that gear inside the stock air box.
Thanks. Something else that occurred to me is that the distance from the airbox to the throttle is different for the inboard and outboard cylinders, which would probably pose a problem for shower type.
My ECM does not support Sequential operation, so that was what drove this decision. I have not considered the tach drive, however, I believe that the tach drive does not operate at the correct speed to use it as a cam sensor. I am CERTAINLY not suggesting that using the cam tach drive is not possible, only that I did not do the research necessary to answer that question. IF you could figure out a simple way to mount a cam wheel only, then, along with the proper ECM, you could conceivably build full sequential injection. Full sequential is really not of any benefit except at idle and low RPM and is mainly used by the OEM's to meet strict emissions standards. The batch fire method simplifies the system and still delivers great performance. Since the OEM Ignition and advance unit was not being used and really not very desirable, this was the obvious place to mount the trigger wheel and sensor. The setup that I selected was really not that difficult to build, mostly straight Lathe work and small amount of milling involved.
Your trigger wheel is a work of beauty. What I was thinking of when I asked about using the tach drive is actually machining the camshaft at the tach drive gear to make it into a pickup wheel. This might require welding it up first, which I assume is possible given the long history of welding up and regrinding cam lobes. The other approach I was considering is a 2-piece wheel that clamps on to the cam such that a sensor using the tach drive mount points would be able to read it - but that obviously has more numerous and more concerning failure modes than having a toothed wheel section machined into the cam itself. One of my concerns is if a hall effect sensor could even work reliably on the small gear teeth required (I'm guessing ~1mm) on a shaft spinning at up to 5000 RPM. I'm assuming that an optical pickup is not going to be possible in the oil-soaked cam galley environment.

Another option would be to build something off the end of the cam and use a custom machined cam cap that passes a shaft through to a sensor arrangement. I don't like the idea of a wart on the top of the cylinder head, nor am I eager to have to manage sealing the oil around that shaft. Maybe a custom cam cap with a viewing window and an optical sensor out on the dry side, and an optical wheel mounted on or machined into the end of the cam, with some kind of oil wiper arrangement to help it work. I don't know, it's more work and more complexity in a lot less space than with your approach to the crank angle sensor.

My goal is to run closed loop with a catalytic converter - I've been assuming that would require sequential operation. Do you think that it would be possible to get there with batch?
I think that batch fire will be possible as long as your injectors are properly sized for good idle characteristics. As I have said before, all early cars used batch fire with catalytic converters with success. Not sure which converter you are using or even why you want one, the converter characteristics will dictate whether sequential is necessary. Again, I am not an expert in this field. The problem of a "Cam Mounted" solution is one of space. On the CBX, It is not possible to use a large enough wheel on the cam to provide the necessary signal for single wheel operation, in the space available. In my case, I did not want any modifications to the cam cover or camshafts, so this eliminated most any cam based sensors as "too complex" for me. For full sequential you can use a single cam wheel, provided the wheel is large enough and has enough teeth or slots to provide a usable signal from the Hall effect or the optical sensor. Being exposed to oil would not be a problem for many available sensors, there are such sensors out there. Typical automotive practice is to use a crank wheel with a large number of teeth or slots and a separate cam phase detector with one tooth to provide the phase information. The Single cam wheel does work, I have one of those running on my FIAT Spider 2000. The down side of the single wheel is slightly less accuracy and the timing is affected by cam belt stretch and other factors in the cam drive. The up side is "simplicity"! My ECM does not support sequential, so this was a moot point for me.

User avatar
mlynch001
ICOA Member
ICOA Member
Posts: 20
Joined: Tue Dec 13, 2016 10:42 am
Location: Arkansas
Location: Dardanelle, Arkansas

Re: 1982 CBX EFI PROJECT

Post by mlynch001 »

EMS wrote:In the CBX, you get the signal from the ignition pulser. It involves making a notched disk for the pick-up. But this is the cleanest solution.
Exactly. I have nothing hanging off the end of the crank and ZERO modifications to the basic engine. In other words, nothing that cannot be put back to stock. The existing ignition drive is the obvious place to obtain the RPM data. It would be quite possible to add a cam sensor to provide the phase information, that would be beyond the capabilities of my chosen ECM. If I had found an ECM that would support sequential, that would fit into the allotted space and at a price that was acceptable for my budget, then I might have gone that direction instead.

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

mlynch001 wrote:The problem of a "Cam Mounted" solution is one of space. On the CBX, It is not possible to use a large enough wheel on the cam to provide the necessary signal for single wheel operation, in the space available. In my case, I did not want any modifications to the cam cover or camshafts, so this eliminated most any cam based sensors as "too complex" for me. For full sequential you can use a single cam wheel, provided the wheel is large enough and has enough teeth or slots to provide a usable signal from the Hall effect or the optical sensor. Being exposed to oil would not be a problem for many available sensors, there are such sensors out there. Typical automotive practice is to use a crank wheel with a large number of teeth or slots and a separate cam phase detector with one tooth to provide the phase information. The Single cam wheel does work, I have one of those running on my FIAT Spider 2000. The down side of the single wheel is slightly less accuracy and the timing is affected by cam belt stretch and other factors in the cam drive. The up side is "simplicity"! My ECM does not support sequential, so this was a moot point for me.
This is super helpful, thanks!

It seems to me that if one were willing to modify the cam cover, it might even be possible to pick up the cam phase signal with a hall effect sensor on one of the cam lobes. It wouldn't be very accurate, but it should be enough to just provide context for the signal from the crank sensor.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

EMS
ICOA Member
ICOA Member
Posts: 10151
Joined: Wed Mar 05, 2003 7:55 am
Location: North East OH, ICOA 3904

Re: 1982 CBX EFI PROJECT

Post by EMS »

EMS wrote:In the CBX, you get the signal from the ignition pulser. It involves making a notched disk for the pick-up. But this is the cleanest solution.


7719

BiKenG
New Member & Happy To Be Here
New Member & Happy To Be Here
Posts: 24
Joined: Sat May 19, 2018 5:50 am
Location: Surrey, UK
Location: Surrey, UK

Re: 1982 CBX EFI PROJECT

Post by BiKenG »

Syscrush wrote:...
This is super helpful, thanks!

It seems to me that if one were willing to modify the cam cover, it might even be possible to pick up the cam phase signal with a hall effect sensor on one of the cam lobes. It wouldn't be very accurate, but it should be enough to just provide context for the signal from the crank sensor.
That is how it appears to me. The CaM Position sensor doesn't need to be accurate at all. As you so succinctly put it, the cam pulse just provides the context and the CranK Position sensor it used as the actual trigger. In truth, the injection pulse doesn't need to be that accurate in any case.

I was thinking the tacho drive gear on the cam could be modified to be a suitable 'rotor' (grind down some and/or weld an additional lump at the right place) and the tacho drive assembly replaced with something that holds the sensor. In this way, just that one camshaft would be modified and they are relatively easy to replace if 'back to standard' was ever required. Although if the internals of the Tacho were replaced with electronics, the new CMP sensor could be used to drive that and no-one need be the wiser :-)

Throttle bodies are however an entirely different sort of problem and I had considered using the original carb bodies. But in the end, I think that's messy and it could well be easier to manufacture new bodies and maybe utilise the original butterflies and shafts etc. However...

What is the advantage of using separate butterflies? Why not one single large butterfly in the intake that then splits to each intake. Won't cause any problem with atomisation of the fuel with such a large actual intake as the fuel will be squirted into the individual inlets just like any other system with a butterfly in each. Big advantage as I see it would be the complete absence of any synchronisation requirement and I suspect the entire manifold could be made smaller and neater without all those linkages running across.

What are the disadvantages of such an arrangement?

It is obviously possible as it has been used on some vehicles (Gold Wing even?), so why is it not more common as mechanically its far simpler. If the one single intake is large enough, there'll be no restriction over and above the size of the individual intake ports, so no difference there. Does it perhaps cause pulse issues when one cylinder is sucking and basically connected to all the other cylinder intakes, instead of the butterflies restricting that interconnection. Does that seriously impact on power potential? Torque?

Anyone sufficiently familiar with FI system design to comment on this?

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

BiKenG wrote:I was thinking the tacho drive gear on the cam could be modified to be a suitable 'rotor' (grind down some and/or weld an additional lump at the right place) and the tacho drive assembly replaced with something that holds the sensor. In this way, just that one camshaft would be modified and they are relatively easy to replace if 'back to standard' was ever required.
Tach drive was my first thought, too. But from my research so far, I've decided to start with batch and see how it rides. If I end up unhappy with the tunability, I'll add a cam phase sensor there or on a cam lobe and convert to sequential.
What is the advantage of using separate butterflies? Why not one single large butterfly in the intake that then splits to each intake. Won't cause any problem with atomisation of the fuel with such a large actual intake as the fuel will be squirted into the individual inlets just like any other system with a butterfly in each. Big advantage as I see it would be the complete absence of any synchronisation requirement and I suspect the entire manifold could be made smaller and neater without all those linkages running across.
The CBX that I saw converted to EFI was done exactly this way. I posted some details here. It's definitely simpler and more tidy. For me, it's problematic because I want to keep much closer to the stock looks, and I also want to retain the stock intake tracts and airbox, which I think are important given the oddness of the intakes on the CBX head.

But others have had success this way and it lets you use an off-the-shelf TB with integrated TPS, so I can absolutely see the appeal.

I have not totally ruled it out, but I think that mlynch001's results are so great that I wish to try to replicate them.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

BiKenG
New Member & Happy To Be Here
New Member & Happy To Be Here
Posts: 24
Joined: Sat May 19, 2018 5:50 am
Location: Surrey, UK
Location: Surrey, UK

Re: 1982 CBX EFI PROJECT

Post by BiKenG »

I am working on a particular project that may have a clearance problem for the original carb bank, so the idea of a slim manifold with maybe just the injectors for each cylinder and a single butterfly is possibly the only solution. However, another bike will have the standard clearance, although I'm still not sure it's worth the effort of modifying the original carbs rather than finding replacement TBs. I've heard an early Kawasaki had TBs that can be banked together, but I would like to find out the port diameters and spacing on triples that are out there. Maybe, just maybe, there's one that would slide straight in. Make the angled linkage in the middle and you're there.

In any case, I reckon a tacho drive sensor is the best solution. I followed a thread elsewhere about injecting an RC30 and that required not only welding a lump (short length of sawn-off bolt I seem to recall) onto a cam, but also having to do some major work on the cam cover to fit and hold the sensor. On the CBX the cam cover is already done for us.I would have thought it relatively easy to grind down the tacho drive gear on the cam and then weld a nice 'lump' in the right place. It's not like we'd have to find the space to squeeze it into. The cam is ready made with the location and space to do this and if anyone's really worried about 'keeping it stock', then I'd suggest modifying a cam like this is a lot easier and less 'non-stock' than butchering a set of carbs and the new CMP sensor could in any case still be used to drive an electronic tacho and no need to replace the cam.

I want to go the EFI route for 2 main reasons. Firstly, it's easier to adjust fuelling (and be more accurate) by hooking up a laptop than by pulling the complex carb bank out for every adjustment. Secondly, you can simply turn the bike off and walk away, for months, and it'll still fire up perfectly. No carbs to drain and then have to refill to start, or you forget and have to dismantle the carbs to clean them from all the dried out fuel.

Since I don't ride any one bike on a regular basis, that last is a real incentive for me. Back in the day I remember all the messing about with draining CBX carbs etc, but more recently I had a 2000 FireBlade (with EFI) that sat unused for more than 2 years and I simply turned it on, pressed the button and it fired up immediately and ran perfectly. I LOVE EFI. :D

I still have some unanswered fundamental questions though. What are the (dis)advantages of single TB vs. individual butterflies vs. single shower type injector vs. individual injectors and what about the orientation of the injectors? Obviously a fully downdraft intake is impossible on the CBX, but given the horizontal intakes, is it best to have injectors on the top and firing down (most common in my experience) or on the bottom firing up (as the OP's carb mod)? I can see an advantage of the latter, but why is it not more common? Does it indeed make any difference? These are things I'd like to understand better and would love to discuss it with an injection expert. No disrespect to to others here, but we're all just speculating and I'd like to get to the truth of it. What works best and why.

It is of course highly likely that compromises have to be made in any case to suit the available space etc, but it would be good to know what we should be aiming for and why.

I'm also considering injecting a VFR750, but that should be easier by using 800 TBs and ECU etc, but also a CX650 which has the same horizontal intake as the CBX, so similar considerations.

So much to do. :?

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

I suspect that the advantage of injectors on the bottom only applies to ITBs - you should get the best atomization injecting into the area by the edge of the butterfly that's downstream of the airflow. With a single TB, I don't see how it could matter.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

BiKenG
New Member & Happy To Be Here
New Member & Happy To Be Here
Posts: 24
Joined: Sat May 19, 2018 5:50 am
Location: Surrey, UK
Location: Surrey, UK

Re: 1982 CBX EFI PROJECT

Post by BiKenG »

Syscrush wrote:I suspect that the advantage of injectors on the bottom only applies to ITBs - you should get the best atomization injecting into the area by the edge of the butterfly that's downstream of the airflow. With a single TB, I don't see how it could matter.
Well it seems to me that injecting from the bottom, you're less likely to simply spray the opposite side of the inlet since natural forces (i.e. gravity) will be attempting to keep it away from that upper side. So gravity is working against the tendency for this to occur when spraying not completely in line with the flow. Also, if any does condense on the upper side, it is more likely to run down and get picked up by the gas flow rather than simply pool at the bottom.

I realise this may be a very minor effect and also could be affected by the direction the inlet takes downstream of the injector. If it curves, then that may also affect the optimum location of the injector. However, on the CBX they're fairly straight so that probably won't be significant.

I am interested in a bottom injector location as it provide max clearance above the manifold which is important for one of my projects.

However, I'm not a fuel injection expert designer. I wish I knew someone who is.

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

BiKenG wrote:Well it seems to me that injecting from the bottom, you're less likely to simply spray the opposite side of the inlet since natural forces (i.e. gravity) will be attempting to keep it away from that upper side. So gravity is working against the tendency for this to occur when spraying not completely in line with the flow. Also, if any does condense on the upper side, it is more likely to run down and get picked up by the gas flow rather than simply pool at the bottom.
I'm no expert, but I would be absolutely shocked if gravity was relevant when we're talking about finely atomized droplets of gasoline blasted under pressure into an intake that's flowing air at ~100mph (which I believe is a typical speed for air in the intake tract) and travelling just a few inches to the cylinder.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

I came across this not long ago, and wanted to share it here because I think it's an approach that others doing EFI conversions on the CBX may want to consider.

It's a circuit design for using multiple MAP sensors for ITB engines:

https://github.com/jharvey/MultiMAP

More info can be found in this incredible Ferrari build thread in the post from user "mke" with timestamp "1/18/19 12:22 p.m." - sorry, I can't see any way to link to a specific post there.

What the circuit does is use one MAP sensor and one diode per throttle body, inputting into an op-amp and some other components. In this configuration, the output of the circuit will correspond to the MAP sensor that has the strongest vacuum (lowest absolute pressure) at any given moment. To quote the post linked above:
The idea is the when you have say 4 vacuum hoses on an ITB setup going to a canister as is usually done 1 is always pulling vacuum but 3 are leaking vacuum and you end up with a signal that is about 1/3 to 1/2 what you'd see on a plenum setup
This approach should aid with tuning off-idle throttle response.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

User avatar
Gearheadgregg
ICOA Member
ICOA Member
Posts: 366
Joined: Tue Sep 30, 2014 9:47 am
Location: Rhode Island
Location: Rhode Island

Re: 1982 CBX EFI PROJECT

Post by Gearheadgregg »

Be nice to Fab this to our CV carbs > http://roadstercycle.com/yamaha_vmax_fu ... cv_car.htm

User avatar
Syscrush
ICOA Member
ICOA Member
Posts: 1714
Joined: Fri Sep 14, 2012 2:29 pm
Location: Toronto, ON
Location: Toronto, ON

Re: 1982 CBX EFI PROJECT

Post by Syscrush »

Gearheadgregg wrote:Be nice to Fab this to our CV carbs > http://roadstercycle.com/yamaha_vmax_fu ... cv_car.htm
That is a very interesting approach! I've contacted that shop to see if they're willing and able to try the CBX.
Phil in Toronto
A cool guy deserves a cool bike, a dork needs a cool bike...
Pics of Perry, my '79.

BiKenG
New Member & Happy To Be Here
New Member & Happy To Be Here
Posts: 24
Joined: Sat May 19, 2018 5:50 am
Location: Surrey, UK
Location: Surrey, UK

Re: 1982 CBX EFI PROJECT

Post by BiKenG »

Gearheadgregg wrote:Be nice to Fab this to our CV carbs > http://roadstercycle.com/yamaha_vmax_fu ... cv_car.htm
I looked into using that when planning to inject a Valkyrie. In the end the project never got off the ground, but I think it's feasible. It does seem to make sense when you effectively have the throttle bodies made for the job and just need to swap jets for injectors. I may have a go myself and make the injector holders etc.

I would certainly be very interested in the outcome if anyone were to try it.

However, purpose designed throttle bodies have to actually provide a better solution. Seems to me we need some custom throttle bodies made to accept the spindles, butterflies and linkages of the original CBX carbs. How hard would it be to get those manufactured?

Post Reply

Return to “Project Threads”