Aviator 902S

A while back (about 26 months ago) I installed a metering pump adapter on my '85 GSL. As previously described, this unit bolts between the metering pump and the mounting surface on the front cover. It acts as a block-off plate to block oil flow from the engine, and instead feeds 2-cycle oil from a separate reservior while still using the stock metering pump. It contains its own drive shaft to link the metering pump drive (located inside the front cover) to the stock pump. I've posted a couple of updates and promised to continue to do so, reporting any news good or bad. Here's the latest:

Last week my metering system quit. At each fill-up I top off the reservior with TCw3 2-cycle oil, and each tank of gas uses about 8 oz of oil from the reservior. But this time I found the reservior still full and the nylon lines from the metering pump to the carb empty. The car ran over 200 miles with no metering oil. Because 2-cycle oil lubes, adheres to and permeates the pores of the metal way better than 4-cycle oil does, the engine is still protected for several hundred miles more without metering oil than it would be with the stock 4-cycle set-up.

I removed the pump and adapter, thinking that either one would prove to be the smoking gun. The adapter checked out fine, with no wear or broken components. Next, I completely dis-assembled the metering pump. Everything was as it should be with nothing broken or seized. The oil flowed out of the line from the reservior just fine, so there was no apparent blockage in that part of the system. I then had my wife crank the engine over while I watched the pump drive to confirm that it was turning. This too checked out ok.

Baffled, I tried to move the pump drive on the engine by hand. There appeared to be some end play, and I wasn't sure if this is normal. Still, it didn't seem to be enough to completely disengage the drive slot in the shaft from the one on the pump or adapter.

I took the opportunity to try the alternative mentioned earlier by another member on this forum --- removing the adapter and instead drilling a hole through the edge of the stock metering pump's mounting flange to the oil inlet hole on the pump. I then tapped the end of the hole and threaded a custom fitting/ tube into the hole so that the oil line from the reservior could be attached to it. Used with a thin aluminum block-off plate and two gaskets, this method would render the adapter obsolete since the modified pump could do the same job. But rather than experiment with my current pump (the one with the new non-leaking Viton seals in it), I picked up another one from the wrecking yard, bead-blasted it clean and then performed the above-mentioned mod on it.

The good news is that this time the pump once again was pumping green 2-cycle oil into the carb. The bad news is that it leaked like a seive. Figuring that old seals were the culprit, I removed the pump for a closer look. What I discovered was that the 1/8" diameter hole I had drilled removed too much material from the 1/4" thick flange. This caused a hairline crack to form as soon as I torqued down the mounting bolts. The 1/8" hole was larger than it needed to be and weakened the flange too much to withstand the torque and vibrations.

So I completely dis-assembled both pumps to clean them and compare parts. This now-broken pump had been pumping oil where the previous one had not, so it stood to reason that maybe the wear on the previous pump was more than what was visually apparent. But when I cleaned both and compared their components I found no measurable wear.

So I went ahead and re-did the mod on the original pump, drilling the hole to 1/16" diameter rather than 1/8" to provide more strength in the flange. I counter-bored the first 3/16ths" of its depth to accommodate the fitting, and then re-inforced the fitting with a very healthy dollup of JB Weld around the outside.

The good news this time is that there were no leaks. The bad news is that the system is once again no longer pumping. This can only mean one thing: The end play in the metering system drive is more serious than I thought. It may have been pulled outward far enough to engage the junkyard pump (for a while anyway) but had retracted end-wise when I installed the modified original, causing engagement to be lost and no oil to be delivered. At least that's my current theory.

Further inspection of the drive shaft inside the adapter revealed what appeared to be very faint wear marks on the very end of the adapter's drive shaft key where it mates into the slot in the end of the metering system drive shaft.

Could the end play be so much that the key was slipping out of the slot but still making enough contact to produce wear marks on the end of the key? If so, could this end play have been caused indirectly by driving a stock pump via a shaft inside the adapter? I don't believe this to be the case since this adapter is in such good shape and its shaft turns so easily that it would present very little if any extra load on the drive system.

It's been so long since I've had one of these engines apart AND inspected the metering shaft that I can't recall if this is normal or not. I'm thinking it's not, but can anybody confirm this? Has anybody had their stock metering system fail due to end play in the metering system drive gear and shaft assembly? I'm running pre-mix in the tank until I've resolved this issue.

RotorMotorDriver

iTrader: (1)

Long read, and unfortunatly, I dont know EXACTLY what youre talking about, but a free bump for ya in the mean time. Im a visual learner, so forgive me if Im wrong (Ive seen the unit youre using however), but are you trying to say that the drive shaft coming out of the engine (looks kinda like a flat head screwdriver, right?) has end play and is disengaging from the pump unit by moving back into the engine?

~T.J.

Rx7carl

IIRC, theres a bit of freeplay in the drive for the OMP. The only retainer is a roll pin that is accessed from inside the front cover. Did you measure how much engagement you have? I mean the junkyard pump works and drives, but your original doesnt. Regardless of your comparisons the evidence shows some kind of failure of your pump, not the drive. How bout putting the guts of the junkyard one into your old housing? Remember on one end of the pump is the 2 notches on the flange to keep the end captive, the other end is flat. Let us know what you find.

Aviator 902S

Exactly. Everything else is as it should be. Oil flows into the pump from the reservior, adapter and pump are both in pristine shape (I dis-assembled the pump down to the last component, then reassembled and checked operation before re-installing), nylon lines are almost new. The only thing I found was the end play in the shaft coming from inside the front cover.

Aviator 902S

Thanks Carl, I couldn't recall the exact setup inside the front cover except that it is a gear that drives the shaft directly. I didn't actually measure the end play, but I did pull out on it and noticed roughly 1/8" to 1/4 " of end play. When I was back at the wrecking yard later (to pick up a signal light lens) I checked the end play on the pump drive of the engine that I took the junk yard pump off. There was very little if any end play on that one.

Unfortunately, the junkyard pump was destroyed when I drilled the oil passage too large (1/8" diameter in a 1/4" wide cast aluminum flange) and the flange cracked when I tightened the mounting bolts down. But prior to installing that pump I pulled the drive out as far as I could before installing the pump. This would explain (I think) why that pump worked, at least during the short time it was in the car.

When I re-installed the original pump I did not bother to pull the shaft outward, so it appears that the shaft is the most probable culprit. When I had both pumps on the bench I dis-assembled them and compared internals. There was no detectible difference so I re-assembled each pump without swapping anything over except the modification fitting.

Right now I'm running premix, but I'll have time this Friday to remove the pump again. This time (just for grins) I'll swap the components over from the junk yard pump and get an actual measurement of shaft end play before re-installing the pump. BTW, can the pump drive gear be viewed by removing the distributor and shining a flashlight down into the hole? I can't recall.

jayroc

When you did the mod to the OMPs, did you use the adapter plate?
Also, have you verified that the OMP adapter is turning with the engine and there's no blockage in the pathway in the adapter? Maybe the adapter itself developed some play?
From your description it seems that is where the problem is...
Can you try your good pump on another engine? Also maybe try and actuate the OMP by hand at idle when you get the best vaccuum...

I remember also that there is some degree of play in that OMP shaft, but how much I forget.. Could be 1\8". You should know if it is enough to cause a problem if it goes in without having to line up the notch..

Rx7carl

I dont recall if you can see it or not, but the gear is bound to its drive gear in the cover so it cant move that far. Also the roll pin keeps it from moving too far in. And butting against the OMP keeps it from moving too far out.

Jay, manifold vacuum wont matter as the tubes are above the throttle plate, but hand actuating the OMP will tell if its working or not.

Aviator 902S

Modding the omp rendered the adapter obsolete. A hole is drilled through the omp mounting flange to the oil inlet hole that normally allows oil from the engine to enter. A fitting is then fabricated and threaded into this hole, and then the oil line from the reservior is attached to the fitting.

I use a very thin (0.020") aluminum block-off plate cut to the same dimensions as the gaskets to keep engine oil from being pumped. It mounts between two stock omp gaskets. BTW, total thickness of all three = < 1/16", so it's not enough to disengage the drive shaft from the omp. Besides, the problem first occured while the adapter was still being used (with only one gasket between the adapter and the front cover), but continued even after I installed only the modified omp.

I did actuate the pump by hand before installing and it worked fine. Unfortunately I do not have another engine on hand to check this pump with. I'll know more by Friday afternoon, after I've pulled the pump off again.

rustnrot

Can you post a pic of the mods you did to the OMP pump to allow flow in from the oil tank. I have two junkyard pumps and would like to mod one.

jayroc

I wasn't sure if you were still using the plate with the modded OMP or not.. Definite pump problem, then. Very strange problem, indeed.. Good luck with it, man.

rx7carl
Ah, I see. I forgot about it being before the throttle plate. I was thinking of something else.. heh

trochoid

iTrader: (6)

Looking at one of the front covers I have off, the OMP shaft can recess about 3/16". Total travel in and out is about 1/2". I will venture a guess that the way the gears drive the shaft, it forces the shaft outward, until it is stopped by the OMP. You may be able to test this by rotating the engine over by hand and checking to see if the shaft is pushed out.

I do not recommend starting the engine to do this because when the shaft is fully extended outward, the gear on the inside will contact the inside of the front cover and either, chew up the cover, or jamb the shaft and start stripping out gears.

Aviator 902S

As much as I'd like to I don't have the digicam or the know-how. Before they changed this forum's format several months back I was fluent in posting pics from photobucket that I'd scanned into the computer, but now they've changed the procedure and my scanner has crapped out.

But I think I can describe the process so that it makes more sense. Go grab yourself a coffee (or beer) and a comfortable chair because this will be a lengthy post.

This mod is an alternative to the Mazda rotary engine oil metering pump (omp) adapter sold by PCV Technologies. It works on the same principle, but eliminates the need for the adapter since the metering pump will now be able to do the job that the adapter used to do, ie: block off oil flow from the oil pan while delivering 2-cycle oil from a separate reservior to the metering pump and on to the induction system. I got this idea from another member on this forum who has done this mod or something similar, but I can't recall who this member was. Whoever you are, thanks!

Also, this example is for a 12A engine. The 13B engines used a different style of omp, but with the exception of the block-off plate (the 13B uses an o-ring instead of a gasket so an alternate means of blocking oil flow from the oil pan would be required) the necessary steps would be very similar. The procedure is as follows:

You'll need to drill an approximately 1/16" diameter hole into the EDGE of the 1/4" thick mounting flange portion of the omp casting. Do not use any larger diameter drill bit because this will weaken the flange and cause it to break off when the omp is tightened down. This hole enters from the LOWER FRONT portion of the flange and angles upward so that it terminates in the factory oil inlet passage where the oil normally enters the omp when it exits the engine via the front cover.

This means that the hole you drill will be PARALLEL to the mounting surface on the front cover when the omp is installed, but angling upward from a point near the bottom of the flange to the oil inlet. The hole has to be drilled this way to facillitate both the drill bit and the home-made fitting that you'll require. Be sure to remove all swarf (shavings) resulting from drilling the hole. Better yet, completely dis-assemble the pump, clean and inspect all components and then re-assemble after all drilling and tapping has been completed.

While the pump is apart it would be wise to replace all three o-rings with new Viton ones. Two-cycle oil bleeds through places that 4-cycle oil won't. If your pump seeps now, it will absolutely gush once you switch to 2-cycle unless these seals are replaced. More on this later.

This new hole you drilled will need to be tapped (threaded) about 1/4" deep to accept the threaded end of the fitting you are going to make. Do not cut these internal threads any deeper than this because a) it isn't necessary and b) doing so will weaken the flange.

For the fitting, unless you can find a supplier of fittings that have an internal diameter of at least 1/16th" and an external diameter of no more than 3/16" you will need to make your own. It's not that difficult at all. I went to a wrecking yard and hacked off a 2" length of emissions system steel tubing. Most of this tubing (usually found on carburated cars) will be too large in external diameter for this purpose, but some of the smaller ones will work. If possible, try to find a piece that is "beaded" on one end so that you can better attach your supply hose onto it with a small hose clamp.

Next, very carefully bend the tube so that there is an approximately 10 degree bend in it. This is necessary to allow clearance from the end of the fitting from the engine mount. Do not exceed this bend amount or you may kink the tube. De-burr both ends of the tube. Next, clamp the tube in a vise and then cut the external threads in one end for about 1/4" of the length of the tube. The die will need to be a suitable size for the diameter of the tube (ie: very small). Thread pitch doesn't matter as long as the size and pitch is the same for both the tube and the hole in the omp flange.

Now mix up a small batch of JB Weld or equivalent. Smear a small amount into the external threads of the tube and then thread the tube into the omp. If you have red Loctite on hand this would be better, but JB Weld is what I used and it should suffice. Be careful not to allow any JB Weld or Loctite to block the passage you drilled. You'll want the end of the tube to be facing slightly upward, forward and to the right-hand side of the car when threaded into its final position. To do this it will be protruding from under the forward portion of the omp when viewed from the right-hand side of the car.

When threading the tube into the omp, do not force the tube around again if it gets tight before it's in the ideal location --- if you do it will break off because the external threads will have weakened it considerably. Instead, back it off until it is pointing where you want it to point. the JB Weld or Loctite will do the rest and it will stay put. Now use the rest of the JB Weld to re-inforce the tube. Gob it on so that the tube is well supported by the pump body. This will better seal against leaks and make the tube more rigid. Be sure the pump body has been thoroughly cleaned with solvent before doing this. Otherwise the JB Weld may not adhere properly and you may end up with an oil leak as a result.

This completes the mod to the actual omp. Now you need to mount a suitable reservior with a length of neoprene fuel hose and a 45 or 90-degree elbow at the omp end. Both the reservior and the hose must be impervious to oil. The fuel hose will be, but the reservior needs to be either metal or a power steering fluid reservior because those washer fluid containers that seem otherwise very convenient will get devoured by the oil in less than a year. Mount the reservior up high (on the upper portion of the Right-hand strut tower works well) to insure proper gravity feed. A short (about 1-1/2") length of fuel hose can then be clamped onto the other side of the elbow to give you something to clamp onto the fitting that you just fabricated.

One more thing: if your engine is a 12A, you'll need to buy two omp gaskets ($3 each at Mazda) and make a block-off shim to block oil flow into the pump from the engine. This can be fabricated from a beer can. Just cut out a 3" by 2" piece of aluminum from the can, flatten it out and then drill a 3/4" diameter hole through the middle of it. (Actually a hole cutter works best for this. Drill bits just destroy the metal). Then place the drive portion of the omp over the hole so that it protrudes through it. Now use a fine-tip felt marker to trace around the outside of the omp onto the aluminum and use a pair of snips to trim the shim to the proper shape. Then it's just a matter of lining up the shim with the omp and drilling the 1/4" mounting holes and de-burring.

Sandwich the shim between the two gaskets, smear on a very thin coat of silicone and then re-install the omp and connect the 2-cycle oil supply line to the pump. Fill the reservior and you're done.

A quick note on the required o-rings:

You'll need two VITON o-rings, one for for under each end cap on your omp. These o-rings are 13/16" external diameter and have a 1/16" cross-section, ie: thickness. The recesses for these o-rings are actually more triangular in shape, but the o-rings will conform to this shape if you coat them with silicone before initially working them into place. The third o-ring is smaller. it's actually a metric size that is *very* slightly larger than 1/4" outside diameter and with a 1/16" cross-section. It fits over the oil feed adjustment shaft. This is the shaft that's rotated by the lever that's actuated by the fuel delivery system. Viton o-rings stand up to heat (there is a LOT of heat in the vicinity of an omp) way better than common neoprene 0-rings and will last much longer. Accept no cheap immitations.

Aviator 902S

Thanks trochoid, this is good info. I'm now wondering of either a) the slot in the end of the drive shaft is worn (I can check this more closely on Friday) or b) the gears are already stripped. Hope not.

rustnrot

Nice "how to". When I do it I will post pics. A couple of Q's though.
1. You say drill a 1/16" hole but go on to say use a 3/16" OD tube. How is that possible, I know I do not understand.
2. What OD tube should I use--do you think one of the "outlet nipples" on an old MOP pump, if removed from the old one and installed in the new one would make a good "inlet nipple"?

Aviator 902S

Oops, seems I forgot to include that one little detail when I mentioned tapping internal threads to a depth of not more than 1/4". The 1/16th" diameter hole extends all the way from the outside edge of the flange to the inside surface of the original oil inlet, but the first 1/4" of this new hole's depth must be drilled out to accept the depth of the external threads cut into the outside diameter of whatever you use for an inlet fitting.

This would still leave sufficient material to provide the necessary strength to prevent the flange from breaking off, as long as you only drill this counter-bore no deeper than 1/4" and no larger than 1/8" in diameter. This 1/8" diameter would leave sufficient material in which to cut the internal threads that would accept a 3/16" diameter threaded tube.

As for using one of the outlet nipples from another omp for your inlet nipple on the one you're modding, it would certainly be large enough in internal diameter to deliver enough oil from the reservior, yet small enough in external diameter to fit nicely into the edge of the flange. It even has a beaded end to better clamp the tube onto it.

But is it long enough? I believe it is about 1" long, of which 1/4" will thread or press into the hole that you drilled. This leaves 3/4", but you'll need to reinforce it with JB Weld to strengthen it and to prevent leaks. This will consume at least another 1/4", leaving you only about 1/2" left on which to attach the clamp. This clamp is about 1/2" wide. Then there's the question of accessing the clamp with a screwdriver to tighten it --- not easy since this short length would leave the clamp tucked up under and behind the omp.

Aviator 902S

Update:
The omp is once again pumping as designed (except for one little mod ), and the problem WAS with the pump and not the drive system.

The first thing I did after removing the omp was to re-check the drive. The slot in the end of the drive appeared to be in good shape, but I found that I could push it in about 1/2 inch from the mounting surface. There is a sleeve around the end of the drive that goes in about 1/4", and the drive shaft itself (with the slot in the end that mates up with the drive key in the omp) slides in an additional 1/4".

This is too far in to engage with the omp drive key, and initially I thought I had confirmed the culprit. But just to be sure I then disconnected the coil high-tension leads and cranked the engine over for about 5 seconds. A second look at the drive revealed that cranking the engine had forced the key outward. This confirmed Trochoid's theory that the engine-driven gear's rotation forces the drive unit outward, securing an effective mesh with the omp drive key. Sounds like a helical gear mesh, but I can't recall.

So I dis-assembled the omp and cleaned it and its components up yet again. I then dis-assembled the broken omp, this time removing the metering cam mechanism that's adjusted by a rod controlled by the intake.

When I compared this metering cam with the one on my modded omp I found the problem:

Upon close inspection it was evident that the cam surface on the original omp was worn down in such a way that it was always in a no-flow mode of operation, even when the udjustment lever was pulled all the way up! The wear wasn't even apparent until compared closely with the cam from the broken omp.

I also noticed another discrepancy: The 1/4+" diameter o-ring that seals the metering cam and lever on my original omp is actually TWO o-rings side-by-side in the same 1/8" wide slot. When I swapped in the Viton o-ring last year I did not remove the metal "cup" that slips over most of this seal and therefore failed to notice that there were two o-rings rather than one. Still, the single Viton seal (in conjuction with the overlooked and worn neoprene one) was effective in stopping the leak.

As if that wasn't enough of a revelation, the same seal on the broken omp from the wreckers was not even a pair of o-rings but rather, a 1/8" wide neoprene seal of the same diameter as the o-rings --- but with "cup" style flat sides. This is obviously a Mazda-specific component, but can be successfully replaced with two Viton o-rings of the same diameter and 1/16" cross-section.

When I reassembled and re-installed the metering pump (this time using the guts from the wrecking yard pump) the green 2-cycle oil began to fill the lines almost immediately, and no leaks are visible at this time. It appears that this modification is a success and renders adapters designed for this purpose obsolete. Thanks everyone for all of your help with this. I'm going to update the previous "how to" spiel to reflect these changes and to better explain the process so that there's no confusion. But I'll do this on a new thread in this forum under the title "how to..."

As for the bad pump, I've cleaned and loosely re-assembled it too, and I'll take it to meets to show other members what to do, what not to do and what to look out for.