8-6-4 Seleniods not a bolt in...

[63GMC]

I have a 500 on the engine stand and have wanted to try for some variable displacement ability... after a $18 trip to the junkyard I have four seleniod assemblies, valve covers and push rods. However the infamous stock caddy rocker assembly has one bolt ( 7/16?) in the center, while the 8-6-4 assembly has two bolts ( 3/8? well smaller diameter anyways). these two bolts go through each rocker arm and bolt into the head. My first thought would be to use the stock 500 "T" post with a stud that was 7/16 on bottom and necked down to the smaller 8-6-4 diameter with just one selenoid activated pivot point on the top of the "T" post.... my degree is not in mechanical engienerring but I know that the stock caddy rocker arm assmebly is not that stout and worry that releasing the pivot from the center of the "T" post would lead to "side loading" or some other such disastourous (bad) things... any suggestions??

[Al from MTS]

I just took a quick glance at both style heads. My first thought would be to start with a set of castings with no A.I.R. hump down them. #552 or #950.
Then I would look at machining down the height of the stock rocker mount pads. I would install a moly bar straight across all the lowered rocker pads.
Drill and tap new bar to accept the original V4-6-8 rocker assemblies.
Mount sytems styled like this have worked for many aftermarket Cad rockers. They work well.

[63GMC]

Al, should I use the stock 8-6-4 rocker arms for all cylinders?? I dont know about ratios compared to stock 500 but they do have a more positive control with the bolt down the center, would they handle a vt10 or vt15 cam?... I like the moly bar idea.

[Al from MTS]

I would put the whole 368 rocker system on. As far as cam, I've never really looked at the V468 rocker to see if it would have lift issues. It won't take much to figure that out.
If need be, a custom grind cam is no hassle either.

[fleetwood472]

Has anyone been able to get this type of system going yet? I just saw this post and it looks like a very good idea. - Justin

[Nashalac]

Gee, I wonder why this system was discarded from the factory -even with their millions in devolopment. Just a thought guys

[fleetwood472]

I heard that the systems worked great it was just a problem of the computers controlling them. I also heard that GM had a problem with v6 mode making a vibrating feeling but v4 it ran find. With the New modulated displacement system they are using it in only v8 and v4 they dropped v6. What would be cool is rig up v4 mode to a switch and dont use a computer. That way when your cruising on the highway you can switch it into v4 and save some gas. - Justin

[Nashalac]

The system needs to continually drop different cylinders or the ones not used will go 'cold" and not fire properly on restart.

[fleetwood472]

Is that how the original system worked? Do you know where I could find any info on it? Thanks - Justin

[Nashalac]

Would think factory manual would have info. I think that was the issue with set-up was that they dropped same cyls all the time and went cold causing running and emission issues.

[CadillacPat]

All,
I have owned two '81 Cads with the V8-6-4 system, and have throughly read the factory service manual. The V8-6-4 system was the first DFI (Digital Fuel Injection) system installed in production automobiles. The same two cylinder "pairs" are shut down-#1 and #4 for V-6 operation and additionally #6 and #7 for V-4 operation (look at the Caddy firing order of 1-5-6-3-4-2-7-8). Although V-6 operation would be similar to the firing order of the older "odd-fire" Buick 3.8 V-6, at the engine speed it occurs, at least 25 mph (one of the requirements for modulated displacement-MD- operation to occur) it is not noticeable.
The theory for MD is that any type of "throttling" of the incoming air into an engine creates a "throttling loss" which reduces an engine's efficiency. The MD system (commonly called V8-6-4 because of the fender ID badges) uses four solenoids (with special rockers), an MD amplifier (MDA, which amplifies the low current computer signal to energize the solenoids), electronic fuel injection and a digital computer to control it all. Each solenoid has a short rod that, when activated, moves a disc that allows a valve actuator to move through a "window". This effectively allows the rockers to move up and down the rocker studs so that no valve movement occurs (the pushrod and pushrod end of the rockers move up and down together; the valve end of the rockers pivot on the valve tip). This causes pistons in the affected cylinders to move up and down with both valves remaining closed. The cylinders don't cool down because no cool air (i.e. "intake charge") goes into them. This allows the engine to transition smoothly from various displacements (V8, V6, V4). As an example, for the Northstar engine function which allows it to run without coolant in an emergency, each cylinder is "powered" (fuel and air are admitted, and then ignited) every OTHER cycle. On the opposite cycle, only AIR is admitted. The admission of cool air on every other power stroke removes enough heat so that the engine can run without coolant. However, power is effectively reduced to about 1/2. Besides controlling solenoids, the computer (ECU) also changes the injector pulse width to compensate for lower fuel usage (older computers strictly used speed density to determine fuel delivery). The vast majority of problems with the system were electrical/computer related. However, the majority of electrical connectors were of the GM "weatherpack" enviromental type. Of the problems that could occur with the system, one was a lean conditon caused by the computer not being able to reconize the solenoids actually operating if a mechanical/electrical malfunction occured to the solenoids themselves (the engine actually running in V8 when the computer "thought" it was in V6 or V4). The computer could diagnose a loss of power in the system, but not if a problem like the connector for certain solenoids becoming disconnected/broken, a mechanical failure of a solenoid or components or a short of certain solenoids. Other reports of unsmooth transition between various modes were reported. In most cases, the simple fix was just to put in an MDA that would not allow MD operation.
For the two cars I owned with the system, the MD function always worked on the first car, but, for the second, the MD function didn't work until I diagnosed a bad MDA. Upon replacing it, the system worked excellent. As for the increased efficiency, in V4 mode, the engine would get 25 mpg on the highway at 55-65 mph (level grade for 65 mph), and 13+ mpg around town. Also, at a constant speed of 45 mph, the MPG indicator would register around 30 mpg.
Now, the big question...Is it worth converting a 472 or 500 to? Yes and no. Al claims his MT3 cam has got 22 mpg on the highway in a 500 powered '76 Sedan DeVille. Is the 3 mpg worth it? To some, yes. With additionally adding the MD's electronic fuel injection, better all around economy would be gained (especially before the engine wamed up and went into "closed-loop). The fuel curve of the ECU is totally wrong for 472/500. However, the computer can be "tricked" into flowing the correct fuel curve for these larger engines by increasing fuel pressure/volume by installing a higher flow/pressure electric fuel pump and adjustable fuel pressure regulator. An aftermarket throttle body would have to also be used with higher flow injectors. For a given injector pulse width signal from the ECU, bigger injectors and increased fuel pressure would allow more fuel flow at all points along the fuel curve.
As was mentioned, installing a switch to shut down the solenoids would be much easier, but, in the conditions which the system was designed to generate the best fuel mileage improvement, city driving, you would be constantly switching the solenoids on and off as you stop and go. Highway driving would be much easier, but you would only gain 3 mpg at most. A much better gain in fuel economy would come from installing an electronic fuel injection system that utilized an O2 sensor to make the enigne run more efficiently when cold and warming up.
As far as the system being dropped, it was only dropped on passenger vehicles; commercial chassis vehicles-limos and ambulances, kept the system through the '84 model year.
Finally, as with anything NEW, some problems are very likely to arrise. The new system, "Displacement On Demand" is supposed to use solenoid valves to control the flow of oil to lifters so they don't "pump-up" to actuate their associated valves. Common sense shows there are possible problems that could occur, the least of which is solenoid valves leaking as they wear or go bad. This could cause the affected valves to partially open which would result in the cylinder "back-firing" through the intake and/or exhaust valve.

[fleetwood472]

Thanks for the info Cadillac Pat. Do you think it would be best to swap the whole system computer and all and just use a 454 throttlebody to run it? - Justin

[Brennan]

I've had an '81 with the MD368 for quite a few years now too -- an interesting engine and control system for its day and reasonably driver friendly, although definately noticeable. Too much for the average Cadillac buyer of the day: vibrations, changes in power delivery and engine sound, plus the nuisance of increased repairs went against what driving a Cadillac was supposed to be all about. I believe the new MD systems use better electronics to control when to shift modes and a more complicated throttling system to mask the 'shifting' sensation when the system changes the number of cylinders. Electrical connectors have been improved too. However, reduction of throttling losses was the name of the game, and it worked: my engine just rolled over 460,000 km (286,000 miles) and still sees a 3-4 mpg average improvement when the system is activated, depending on driving conditions. I turn it off in heavy traffic by flipping a switch I installed that eliminates the ground signal to the computer from the transmission. That makes the computer think the transmission is NOT in 3rd gear, and shuts the MD down. I find the system annoying in heavy traffic, shifting back and forth from 4 to 8 cylinders feels sort of like an overdrive transmission searching from 4th to 3rd and back when you're driving at just the wrong speed.

In any case, if you're planning to engineer yourself a 500cid MD engine, I would suggest you use all of the mechanical components you possibly can from the 368 engine, use a manual switch to go to 4 cylinders while cruising on the highway only, and upgrade to a completely new FI system that is capable of adjusting A/F mixture on the fly that will compensate for the approximately 50% reduction in air flow through the engine when you close those 4 cylinders off. The reason I suggest this is that I believe it would be too difficult to trick the old 1981 ECU into working well enough on that larger engine. You might get it to run, but it's unlikely you would see any improvement in economy... what you gain by reducing throttling losses would probably be lost in the 'trickery' required to get the thing to work. FYI, maximum speed in my '81 Fleetwood Brougham Coupe on level ground while using only 4 cylinders (even at WOT while forcing the engine to stay on 4 cylinders... not really any change, which figures) is about 70 mph. (maybe with a slight tail wind)

As an aside, I had an interesting conversation with two Eaton engineers (on speakerphone at the same time!) a few years ago that were involved in the design of the valve selector assemblies. I was noticing increased valvetrain noise on my engine and managed to contact them with a few questions on the valve selector assemblies. They were more than slightly defensive about their gizmos, and made it abundantly clear that any problems I was having was most likely NOT because of the Eaton valve selectors, and most probably because of some other Cadillac-engineered issue. Interesting conversation, to say the least. As far as I know, new valve selectors are no longer available from GM or Eaton... has anyone heard of a stash of new ones that might be for sale?

[fleetwood472]

Hey Brennan,

Thanks for the info. Can you give me any contact numbers of those engineers? I would like to ask them some questions and get some measurements and info on those systems for swapping them to a 472/500 I figure it would be worth asking them for a little information. If you could post it or private message it to me or even email it to me. My email address is [email protected] Thanks - Justin

[Big-Daddy]

Would it be possible to run the entire MD system as a piggy-back setup where another computer controls the FI sequence? The MD and DFI computer[s] are present, hooked up, and receiving input, but are disabled from controling anything EXCEPT the valve actuators.

How about installing a TPS switch, like from an old switch pitch tranny, to boot the system back into V8 mode every time full throtle was applied [at any speed]?

Something else, if cylinder cooling was an issue on the original setup, why didn't they engineer the original system to keep the exhaust valve partially open on the dead cylinders? Seems to me that sucking hot exhaust into the dead ones would keep them hot and ready, as well as keeping them from using HP to compress useless air???????????

[Nashalac]

To answer part of the question, I think the system keeps both valves closed and if so there would be no air to "compress". Would create a vacuum on down stroke and return to zero after upstroke.

[Clay]

The only time the solenoids can activate is when the piston is compressing or cylinder fireing. This the only time that both valves are closed. If there is the least bit of valve opening pressure it won't activate because the extra pressure of the fingers (that go through the windows when the disc is turned)on the disc will over ride the solenoid turning force. This is according the the manual. Now if a straight switch is used I'm not sure what voltage should be used. CadillacPat states that the solenoids are activated by the signal from the MD module and then amplified but to what voltage I'm not sure. Perhaps CadillacPat might know because of his owning of this engine and system. :?: If the system runs on a lower voltage and 12 volts is used the disc's would turn when the valve is open causing it snap shut at weird times. And maybe causing damage to the valve or snapping the head off.
Clay

[CadillacPat]

Brennan,
You are right in that the valvetrain for the MD system is a little noisy. The extra noise can even be heard with the hood shut. Along with the problems some had with their cars, I will have to agree that this is not in the true Cadillac tradition of a quiet, dependable car, and not the "standard of the world" as Cadillac advertises. However, when compared to the HT4100 that replaced it, the MD engine was a true marvel of engineering and fuel economy. While the HT4100 would get slightly better mileage in cold start/short trip driving (being a smaller engine that required less fuel to warm up), the MD engine had WAY more power and torque and at least didn't blow a head gasket, break cranks at a whim or wear cam/lifters on low mileage engines. My one well-running HT4100 (out of four) would only attain 65 mph on a good grade, and that was at WOT with the TH200R4 downshifting into 3rd. My MD engine would easily bury the speedo on the same grade. Granted, part of the problem with the HT4100 blowing head gaskets was due to improper maintenance (not adding the GM cooling system supplement which was actually "Barr's Leak"), still, many NEW engines developed head coolant leaks well before the first scheduled cooling system maintenance.
As far as the Eaton engineers blaming Cadillac; I don't think Cadillac was to blame. It is just inherent in the system that it would make a slight clicking.

[CadillacPat]

Clay,
You are correct about when the solenoids actuating the valves. But, it makes no difference when they do. Anything trapped in the cylinders will just keep compressing and vacuuming. For the voltage question, the MDA (Modulated Displacement Amplifier) just boosts the CURRENT to energize the solenoids; the ECU is a low-current unit.

[fleetwood472]

Is the valvetrain noisy all the time or just when switching between v8 v6 and v4? I wonder what would make it that noisy. - Justin

[shiftless]

Probably because there is a slight amount of play in the selector mechanism which causes a clicking.

[CadillacPat]

fleetwood472,
Besides the V8-6-4 feature of the MD engine, and the Digital Fuel Injection ECU, it also idles slower than most engines (the later two, along with a smaller displacement and lighter weight being responsible for the engine's better fuel economy). Even with over 200,000 miles, and well worn rings and worn out valve seals, my 368 still ran smoother than any of my 472/500's (although older, my '73 500 has less miles; I will have an update when I get my 30,000 mile 472 in my '82 Coupe). Like you mention, I also believe the clicking is due to the inherent design of the "play" designed into the solenoid actuation assembly. While you can hear the noise at idle, it is not objectionable. The clicking is more "continuous" sounding, as it is for four cylinders, not like an engine with one bad lifter that clicks. And, at anything above idle, the engine runs quieter and very smooth. Also, I don't mind the switching of cylinders-if you drive at a steady speed, especially with the cruise control on, the engine usually stays in one cylinder mode. While some may complain of the system switching cylinders, you don't hear them complaining of a full size Caddy getting 25 mpg (and, the power difference between the 368 and HT4100 is like the difference between the 368 and 472/500). Also, at slower speeds, like 45 mph, the engine gets close to 30 mpg. Finally, although 472/500's have low peak rpm ratings, the 368 is the lowest at 1600 rpm (472/500's are 2600 rpm). With a 2.41 rear axle, and such a low torque peak, although the engine only makes about half the torque of a 472/500, it got much better mileage at the time. I mention "at the time", because that was old school camshaft technology. Modern quick opening and closing lobe ramps allow 472/500's to get very close to the mileage of the 368 at highway speeds. But, in cold weather, low speed or stop and go driving, the 368 does much better. I wonder if a modern cam design in a 368 would allow it to get even more mileage...

[1979lee]

great info on these engines , i saw 2 at te wrecking/junkyard today complete , again excellent info on these engines

[Asakha]

CadillacPat (or anyone else) : Do you know if the TH325-4L would work well with the V4-6-8 ?
I own a 1981 Cadillac Seville Elegante, the engine works great, but it seems the last owner didn't maintenance the car at all... The transmission isn't shifting right anymore and since I'll have to rebuilt it, I was wondering if getting an overdrive + the V4 would be working (would have a good gas milleage at least !).

And is anyone knows if it's possible to get rid of the V6 mode, and just have V4-8 ?