Current best approach for distributor?

Hi Everyone,

I am lucky enough to be inheriting a '93 XJS from a family member here shortly with the caveat that it requires a fair bit of maintenance. It still runs, but has some various leaks and needs most things rubber replaced. None of that looks to be too alarming, but I am wondering about the distributor.

After reading here and elsewhere about the inherent issues with the Marelli cap/rotor, I am wondering if replacing it with like aftermarket is the way to go. I have seen reference to several distributorless builds, but the complexity of these seemed a bit out of reach from the data I can gather. I am wondering if that is partially due to the equipment out at the time as most are from 10 years ago. Would converting the 5.3l v12 to distributorless be significantly easier with the new megasquirt modules out these days? If so what would the be recommended way to approach that project (include injection? Waste spark or other?)? Any current links on the subject would be much appreciated.

Thanks!
Adam

Dear Adam,

If that part works, then perhaps don’t fix it.

One alternative is to fit a Lucas CEI ignition from the earlier v12 car.

Alternatively, look at the msextra.com (and the old photo archives of this) website for cars with megasquirt fully mapped distributorless ignitions to help decide whether it is within your skills base to make such a change. The quality of the installation is ultimately dictated by the builder as there isn’t a wide enough user base to warrant making a kit which would be common to all car. The problem is that the v12 had so many variations with the peripherals bolted onto the front of the engine as time went by. As a minimum it involves fitting a trigger wheel, crank sensor and map sensor onto the engine and wirring up a new loom between them and the coilpacks.

All of the newer MS2extra and MS3 (via MS3X or MS3pro) can run the v12 quite comfortably.

kind regards
Marek

Thanks Marek,

The Cap,Rotor, plugs, wire, and coils all need replaced at this point. Makes it a good time to change if I am going to do it. I still worry about decent parts availability with the Lucas as it is even older. I understand it had its own set of issues as well.

I have indeed been searching through the megasquirt posts, but many are vague on the precise details. I have not seen any that mention the specific coils/plugs in use or how they were ultimately wired with the megasquirt. Also curious about what the ideal setup would be with the current models. Were many of the older builds waste spark due to limitations of earlier megasquirt or with implementation on this specific engine?

Most of my experience is with electronics (though mainly not automotive) and I am fortunate enough to have a friend planning to work on the car with me who is a very experienced mechanic. Fabrication and installation should not be too big of an issue for us if we can get a better understanding of the task at hand.

Looking at your posts it seems you have worked a fair bit with just this type of setup. If you were going to do it again today how would you go about it?

I still worry
about decent parts availability with the Lucas as it is even older.

Don’t. Cheap aftermarket Lucas caps and rotors work fairly well and are
readily available and cheap, while genuine Marelli caps and rotors are scarce
and the aftermarket parts are disaster-prone.

If you’re gonna do anything other than just renew parts, I’d recommend
retrofitting the earlier Lucas CEI system. It’s far easier and cheaper to
maintain, and it doesn’t have a particularly nasty failure mode that the Marelli
has. Yeah, you could go with something aftermarket and/or distributorless,
and those work great, but it will involve more fiddling.

If you happen to have the Teves III ABS system with the electric pump,
you’re going to want to retrofit that with the earlier system, too.

– Kirbert

i have used the Crane XR700 ignition for 22yrs.

altho it is old tech ,it does the job, unless new systems are a LOT better i’m not changing things.

very simple easy to install, parts are very cheap, like my universal coil,was $30.bucks at Advance auto.

when i did the conversion, i did it because the LUCAS was unfamilier to me , and seemed overengineered for just a simple engine.

but i did change to a standalone electronic fueling computer, SDS(simple digital systems), fully adjustable at all rpm and load conditions.

anyway works for me.
Ron

“i have used the Crane XR700 ignition for 22yrs.”

Me too. Well, perhaps not 22 years. Maybe 15.

Ed Sowell

'76 XJ-S coupe, red

http://www.efsowell.us

If I were to do this again, I’d look to plan in some redundancy by making the wiring harness a lot more comprehensive than that needed for the immediate task. I’d also shy away from using second hand components of dubious quality. What has tarnished the Megasquirt name in the past is the idea of it being done on the cheap and with a limited amount of expertise on the part of the builder. This inevitably unravels as you then have to redo your work and all people see is a tortuous upgrade path from something that wasn’t well understood by the original builder to something that still isn’t understood by others, as evidenced by “problems” expressed on the online forums.

An example of “doing it properly” is the peak and hold driver board which runs my 24 injectors. My first effort worked fine but the switching between lpg and petrol was done mechanically using relays, it contained wiring and a small daughterboard in the centre of the case and was a pig to debug when I had an injector problem.

I decided that I could do better. The second attempt consists of six identical double sided boards, there is no wiring, the switching is electronic using transistor logic and there is simply no scope for anything to go wrong because it has been designed rather than cobbled together. If there is ever going to be a fault, it’ll be easy to either make up a spare board or swap boards in and out to debug it. Despite fitting into the same sized case, all of the components are reachable and nothing gets disturbed if you have to put a meter across it to test something.

It’s a quantum leap forward in every way.

By a similar token, plan out what will go where long before you go ahead and do it. Here is another example:-

Also, always consider turning the question on its head:- if you simply do the same as someone else and NOT do it differently, you can learn from their mistakes and not have to reinvent the wheel yourself.

kind regards
Marek

This is all awesome info, particularly your photos Marek. Would love to see more of these detail photos in forum threads as the old galleries are pretty hard to search.

Few more questions for those who have done this:

It seems the '93 model XJS had low impedance peak and hold injectors. If I am reading the megasquirt info correctly I should be able to directly run those semi-sequential off the MS3 ultimate, or sequentially if I use the addon board. I assume your (rather beautiful) driver board creation was due to limitations of earlier hardware and the extra injectors of the dual fuel system, correct?

Did you keep stock (or equivalent) injectors and plugs?

What coils did you end up using?

Did you do full sequential or semi? Where did you stick the trigger wheels?

I fully agree that it is better to learn from those that have done before (hence the 100 questions). I will look into the XR700 as well, had not come across much involving that yet.

Thanks all!

The two MSv3.0 board mainboard injector drivers (whether fitted with MS2 or MS3 processors) can each run 6 low impedance injectors in batch fired mode.

To run sequential fueling, you need one output per injector and MS2 will not have enough output lines to do this, so you’d need to fit the MS3 processor rather than the MS2 chip onto the v3.0 board.

The MS3X card can run 8 high impedance injectors so needs additional circuitry to run 12.high impedence injectors.

You can run 12 low impedence injectors from the standard MS3 card header by adding 12 peak and hold drivers directly.

My setup allows for the switching between 12 low impedance injectors and 12 even lower impedance propane injectors. I am using standard injectors and plugs with first generation Ford coil packs.

The ignition is wasted spark using six BIP373 chips directly controlled by MS3. The MS3 also controls idle valves, a plethora of sensors for EGT, coolant, fuel and a million other things that get datalogged on each journey.

For a trigger wheel, I currently run a thirty six minus one wheel cut into the crank damper and a separate cam sensor triggered from the B bank camshaft. The old photo albums detail the trigger wheel setup. You only need a second wheel for a cam sensor if you decide to run sequential fueling.

The advantage of distributorless ignition is that you can easily map the spark advance via software to where you want it for any load or rpm setting, giving accurate control. Keeping an old fashioned distributor gives you no control other than changing the hardware itself, however accurate that may be after all of these years…

You can easily search the old photo galleries by entering a search term like “injection” - because the vast majority of the old forum posts are from carburettored car owners, anything off of the beaten track soon sticks out. You can then follow that users email or user name as the next search term.

kind regards
Marek

Hi Marek,

If I understand correctly the HE 5.3l v12 has low impedance peak and hold injectors, correct? From what I have been reading each of those would require its own driver (cannot be run from MS directly). So how can MSv3.0 run 6 low imp injectors in batch mode? Won’t it still need something to do peak and hold? Do you know of any affordable high imp injectors that work in the v12? It seems like it might be easier to just swap some in if going for a MS install anyway.

For ignition you have each BIP chip wired to two coils right? Any reason to go that route over individual LS1/2 coils run straight off MS? (Other than perhaps fitting them in =/ )

I have just started to look into what is involved to set up the mapping in MS. How hard was the actual programming once the wiring is all set up?

Thanks again for the info!

-Adam

Dear Adam,

Injectors fall into two categories, electrically speaking. Those whose resistance/impedance is lower than 4ohms are considered low impedance and those whose resistance is typically 10-12ohms are considered high impedance.

It takes 4amps to open a low impedance petrol injector, but only 1amp to keep it open. Feeding it 4amps continuously would overheat it, so it is desirable to limit the current once it is flowing fuel.

The Lucas 16CU does this by having two transistors in the driver circuit. Once hits it for 4amps for about 1mSec and then the second transistor pulses a lower current into the injector for remainder of the desired injector opening time.

The 16CU has four driver circuits and each one is connected to three injectors. It is thus batch fired in four banks of three injectors.

The MSv3.0 board has only two injector drivers. These are however more robust than the 16CU drivers and can handle six low impedance injectors apiece. This would also thus be batch fired but in two groups of six injectors. Rather than having separate transistors to handle the opening peak and the hold phases, the MSv3.0 board does this with software, whereby it is told to limit the current to a fixed percentage by flashing on and off very rapidly during the hold phase. It is strong enough to handle six injectors per driver circuit.

By contrast, a high impedance injector simply runs by using a “flat” current profile.

Ignition is a totally different topic.

The simplest setup - as currently fitted - is to have one coil which charges and discharges twelve times per engine cycle and then have a method for distributing the electrical pulse to each spark plug.

The most complicated setup is to have twelve dedicated control lines used by the ECU to charge twelve coils and discharge them sequentially to the relevant spark plugs per engine cycle. The ECU only has a simple ability to send a 5v or 12v low current control pulse out and so a transistor circuit is needed to change this into something which can charge the coils. Some coils have this circuitry built into their design and some don’t. The Ford coil packs I have used are simple standalone items, hence the additional BIP373 ignitor box.

The setup I use is called wasted spark. This uses six control lines from the ECU and feeds six double ended coils. The coils are fired twice per engine cycle. One of the pulses ignites the mixture in the conventional way and the second pulse, which is 360’ out of phase, ignites the mixture in the cylinder which fires 360’ later. Because the coils are double ended, the spark plug in the “wrong” cylinder, 360’ out of phase also fires, but there is no combustible mixture to ignite, hence the term “wasted spark”. I have chosen this setup as the best compromise between using too many of the ECUs resources and having the most functionality.

Programming the MS unit relatively straightforward, as you can start off with a mapping supplied by another user like myself and then adapt it to your own needs. It’d be much more daunting if there were no one to copy, but even then the default settings should be enough to get the engine to start and run, before the autotune features can be used to take over.

You could also consider running just a trigger wheel, a MAP sensor and a Megajolt for ignition and leaving the fuelling to the 16CU for the time being.

kind regards
Marek

Marek,

More great info, I appreciate the time. I have been pouring through this and many other post trying to come up with an ideal setup. It seems that at the point of doing ignition I may as well use a MS3 and hook up injection as well. Here is what I am thinking thus far if I go the MS route (please correct any mistakes I may be making here, as it is still very much a plan in progress):

Looks like a MS3X w/ 3.57 board should work well for this install. Don’t see any major reasons to go with the pro or ultimate (unless anyone can point out something I am missing).

Ignition: I am thinking of three 4-tower wasted spark coils with logic level input. Model 032 905 106B is referenced in the MS manual and looks like it should do the job without messing around with the BIP373 chips. Also avoids needing to mount EDIS modules.

Injection: It seems that the MS3 should run two 6-injector banks of the stock injectors directly using PWM. I can start with batch and if I ever want to go sequential the MS3 should be able to handle it with some modifications.

This setup should mean one 36-1 trigger wheel will be sufficient and extra sensors will be icing on the cake. Another side benefit is most of the critical wiring will be replaced by new harness (as most of the old stuff is in rough shape anyway).

Still some remaining questions: What additional sensors are worth fitting? Are there any available map files that would be a good starting pint for this setup (since yours would have different injection)?

Alternatively, if I chicken out on the pile of work that this will certainly be and decide to just go with an older Lucas distributor where is the best place to source one? Finding caps and rotors is easy, but the distributor itself seems scarce online - even on ebay.

You need to be a little bit more rigorous with terminology before going further.

There are two mainboards currently used for an end user put together MS installation:- the v3.0 board which uses discrete components and comes unbuilt; and, the v3.56 board which is the same (afaik) but is made from surface mount components and is supplied prebuilt.

Whether these would be termed MS1, MS2 or MS3 is simply a question of which processor is fitted onto the board.

There are also a couple of plug and play MS units which are differently spec’d and typically feature the MS3 processor and are sold through companies like DIYAutotune.

The MS3X card is an add-on card for the MS3 processor. It doesn’t support low impedance injectors and can’t even handle twelve injectors at all without an additional diy built piece of circuitry.

Whilst your initial batch fired wasted spark approach looks simple enough, I wouldn’t underestimate the complexity of putting in a new wiring harness into the car. I would conservatively estimate there is over 200 metres of new wiring in my car and making a good harness isn’t easy.

As a minimum you’ll need a coolant sensor and an air temperature sensor to add to your crank sensor.

It will dawn on you that MS can also control the fans, air valves, alternator, warning lights, run closed loop fueling if lambda sensors are fitted, electric water pumps, fuel pumps and generally bring the car into the 21st century.

So long as you can get the engine to fire up at all, then getting it running ought not be a problem. The cranking map will be similar for most v12 engines.

kind regards
Marek

Hi Marek,

I was looking specifically at this: https://www.diyautotune.com/product/megasquirt-iii-ems-system-with-ms3x-expansion-v3-57-assembled/

It seems they are up to 3.57 on hardware rev. for the pre-assembled. This would have the MS3 processor, 3.57 daughter card and the MS3X expansion card. Plenty of input/output for what I need and then some to grow into the other features you mentioned.

Two injector batches look to be handled with PWM on the “main” MS3 daughterboard and the MS3X has the 6 logic level ignition outputs for wasted spark.

Certainly this will not be as simple as a couple paragraphs here, lots of wire and small spaces to deal with. Exactly why I am still keeping the Lucas distributor open as an option. I plan to tear down the engine to tackle the gaskets, do some general clean up, and get a feel for what I am looking at before I order any of this kit.

I just got caught up on the list today and saw that finding a distributor gear is turning up nothing. Over the years I have accumulated a several distributors and have used various parts for getting the best setup for a 74 XJ-12 and my SIII E Type. I’m currently in the process of converting my E Type to MegaSquirt. I went whole hog and got the MS3 Pro since its water resistant and I wanted to mount it on the firewall, there’s really no room inside with factory A/C on a SIII.

Something else to consider since you have a 93, I am using the intake manifolds from a 95 expressly because they will accept the newer EV14 high impedance injectors. These are readily available on ebay, spray MUCH better than the old ones and won’t require a peak and hold driver box. I know the 94 through 96 manifolds work for this, but I’m not sure about the 93, Something you might want to look into should you eventually go the MegaSquirt route

Anyway, I may be able to help out on the distributor gear if you are still looking. I have noticed that it seems when the gears are fitted to the shaft, the drilling for the rollpin is off center and unique, so they are not always interchangeable. I’d have to dig through some storage boxes buried in the garage to locate all my old distributor parts. They range from SIII E Type to the later GM style pickup. I know I have 2 of the latest and one of them is currently in my E Type, but there should be a shaft and its matching gear out there somewhere. If yours is apart, post a picture of the shaft and I’ll have something to look at to see if its a match.

Hi Steve,

Thanks for chiming in. If I change out the distributor I would need the while Lucas assembly, not just a gear. From my understanding the main problem with the Marrelli is really between rotor and cap, not so much the gear. Do you have any of the complete Lucas distributors around?

The manifolds is an interesting idea, though I am not sure if the 6.0L would easily bolt on to the 5.3L. My car is not currently in my garage (in process of shipping). Once I get it apart I was going to look more into the injectors. Might even be possible to find a totally unrelated high Z injector that will fit in the current manifold. Even if the flow rate is a bit higher the MS should easily compensate when tuning. Even with the current low Z I should be able to match the factory bank to bank setup.

-Adam

Hi Adam,

Most, if not all, are in pieces. I found it was cheaper to pick up an entire distributor than to get any specific parts. What I did was to use the best out of each. I set mine up on an old Sun distributor machine so that I could get the advance curve I wanted by picking and choosing parts. You’d be surprised how many variations there are on these distributors. What I have left over could probably be put together to form a complete distributor, but then it comes down to what parts to use. Most are the early OPUS and one was converted to Allison(that was the name before Crane bought them out) and as I mentioned before, only 2 were of the later CEI type. My current setup uses the CEI pickup and it wasn’t without problems. For quite a while I was chasing a problem where the engine was breaking up around 4000 RPM. Eventually it came down to the pickup. Once I changed that out it would rev to 6500 very smoothly. That means I don’t have a CEI pickup to spare. For a complete distributor, all I could offer would be an earlier unit.

As to the manifolds, I can confirm that the 95 manifolds bolt up to the pre HE 71 heads no problem other than cutting out the useless aluminum bars between the ports for access to the angled plugs, which wouldn’t be a problem for an HE motor with its straight plugs.

The injectors I chose were from late model Fords which were being converted to LPG. They flow much more than is needed for the V12 with normal gas but I also am installing a flex fuel sensor so that I can run E85 for the higher octane. I’ve already experimented with E85 with carbs and the power increase is AWESOME! The E85 burns at about 10 to 1 vs the normal 14 to 1 so injectors with a larger flow rate are needed. The beauty of the EV14 injectors is that they can handle very narrow pulse width for a really nice idle even though they are rated for a relatively large flow rate. They also put out a very fine mist vs. a squirt of fuel.

1 Like

I have noticed that it seems when the gears are fitted
to the shaft, the drilling for the rollpin is off center and unique,
so they are not always interchangeable.

British engineering!

You get roll pins? I recalled tapered pins, tack welded in place.

– Kirbert

I could be wrong but don’t you have the plug angle backwards pre-he were straight and he plugs were angled

You guys are really selling me on keeping a distributor =) More and more electronic sounds better.

Where did you find the newer intake manifolds? Mid-90s Jags (or any Jags) are not common here, particularly in the junkyards…