Jeff20B

I'm sure the oil pump worked for many miles after it broke. I just didn't want to run it in the rebuild. I threw a really low mile oil pump in the engine.

Lynn E. Hanover

The complete lack of tooling marks on all Mazda, and most everyone elses pump ring and rotors suggest to me that they are made of powdered metal. The almost glossy finish, and the failures that look like fatigue failures where a fatigue failure is unlikely.



This type of part has been in use for years. The dies are very expensive, but once in place, parts can be manufactured at machine gun speeds. If you need a million pumps, you don't tie up gear hobbing equipment whittling pieces from bar stock at a 4 an hour rate.



Take a piece of the broken pump to anyone with a good old anvil. Put a rag over it

to control the fragments, and wear eye protection. Swat the piece with a hand sledge. If it is still one piece with a big dent in it, it was cut from bar stock. If it is in more than one piece, probably several pieces, it was powdered metal.



Not all bad mojo. The parts are cheap to replace. They are stiff (and a little brittle) and last a long time. If nothing but oil goes through them they go 100,000 easy with no trouble. They are cheap enough that they should be replaced on rebuild.



All of the early style pumps probably cavitate a bit at high speed. The pressure seems to sag after a long high RPM run. I blame it on the two section phased rotor design. That puts the front segment filling through the rear segment, and a bit out of time (phased) with the rear segment. Also the partition is left with 4 sharp square edges that destroy flow in any system.



A windage tray (it isn't really, but you know what I mean) will help remove air from the spent oil, and improve oiling. So 85 PSI oil that is mostly oil is better for the bearings than 85 PSI oil with some large percentage of air entrained in it.



I had the idea of adding another key slot to the shaft and putting both rings and rotors together at the back of the pump. (In phase) I would replace the partition with a solid piece and put it all the way at the front (sprocket end) of the pump. Or, if you want to go hog wild (and you know who you are) you could shorten another ring and rotor, to make it a full length pump and leave out the partition.



Now "That would be an oil pump".



Please note that the late turbo pump is 50MM not 40MM and has a gallery to feed the second segment without pulling oil through the rear segment. This to me is very telling. Nobody in the car industry changes anything that is working well.

They tend not to change things that work poorly because the tooling is so costly.

But here is the 93 13B twin turbo with a bigger oil pump with a different design.

Very expensive business.



That tells me that they know the pumps cavitate under certain circumstances and they could not afford to have their new baby torching motors along the freeways of America. Very bad press. So, with a gun held to their heads, they changed the design.



Your mission, should you choose to accept it, will be to develope an adaptor plate to mount the new pump on the old iron without affecting the performance of the new pump.



You could shorten the sprocket and put it on the crank backwards with a bit of a spacer to keep the stack the same. That gives you the room. for the adaptor plate.



Other things to do with a die grinder while the engine is apart, that can make the stock pump work much better. First is to measure the end play of the pump. The stack of parts in the pump are shorter than the body of the pump.



Lapping the pump body to shorten it down to within .005" of the stack is good if you are not exposed to freezing temps. If you are about .0015" is minimal.



OK so you do this with silicone carbide paper and the kitchen table top, not a die grinder. Big deal.



The suction side of the pump is like the tiny signal coming into a radio. The pump is the amplifier, and the pressurized oil is the amplified signal coming out. So a small improvement in the suction makes a big difference in the output of the system.



The suction side starts with a sharp edged tube inside the bug screen in the pan.

What a piece of **** this is. On the flow bench, this is a flow disaster area. We should get federal funds for a pickup that looks like this.



I used to braze on a real thick (1/4" or more) washer flush with the end of the tube. Then radius the hole and the tube end into a look like a little tuba. A kind of a bell shape. Mother nature really likes that and will give you back a few gallons per minute for free. Some of the vortex covers are not welded all the way around and if this is your case, braze any potential leak points.



The running pump creates a low pressure area all around the end of the tube. The oil has some air in it, more if you have no windage tray, so the air in the oil expands a bit and makes a easy path to the surface so that the tornado you made in the bath tub as it drained follows you into you stunted car crazed adulthood and makes this problem in your oil pan. Anyway the big cover thing that holds the bug screen spreads out this low pressure area so the low pressure vortex cannot pull air from the surface of the oil. Makes you want to keep the oil topped up does it not?



The thick washer eliminates the "vena contracta" at the end of the tube. I used that before and nobody blinked, so I guess you all know about that already.



I continue.



The path to the suction side of the pump is formed by drilling holes in the cast iron. This hole lines up with the suction tube and intersects a kidney shaped hole in the casting on the oil pump mounting boss. The intersection has razor sharp edges and pisses mother nature off to no end. Stay on her good side and smooth off all sharp edges on the inlet side of the pump boss. If you have an old pump, saw off the last 1/4" of the housing and bolt it to the iron. drop in a ring and rotor on a shaft. Observe that there (may) be some of the iron extending into the opening where the pump parts cannot mask it from the flow. Remove any iron and smooth the radius from the kidney shaped hole into the vertical drilling.



On the outlet side the same situation exists. The lateral drilling intersection with the kidney shaped hole is razor sharp and needs to be rounded and smoothed out.

Next is the 90 degree turn into the "O" ring pocket that likes to leak so much. Use a little ball bit in a drill not a die grinder (unless you are really good with a die grinder) to round the inside of this junction. There is very little material above the center line of the lateral drilling so go real slow there.



The best part is to detail the vertical partition between the suction and the pressure side on the iron, but that will be a long one, and takes some pictures to prevent screwing up the iron. If anyone is interested, I will modify a piece of iron and take the pictures.



No sharp corners anywhere flow is involved.



Have you ever wondered why suff the on the turbo side of the engine gets crispy?

What would 1600 degrees look like?



Lynn E. Hanover

bill shurvinton

I use a Pace 3 stage dry sump setup on my car. Not because it needs the oiling per se, but because I needed to lower the engine further than the stock oil pan would allow. Its the pump shown on the homepage, with a 2 stage scavenge and a single stage pressure. Oil is de-aerated in the external tank before being returned to the high pressure pump. There are filters pre and post (the stock unit) tank to hopefully gather up any crud.



I did however miss a trick that I won't do next time. In order to ease the plumbing I have the oil cooler before the tank, meaning I am cooling aerated oil. not ideal, but that effectively the way it is done in the stock application, so in my naivity I thought that it would be suitable.



Sadly I can't report how well it works yet, as other parts of the car have been causing grief. However I can report 40PSI at hot idle which bodes well.



Bill

Jeff20B

I bet is is powdered metal like the corner seals.

Lynn E. Hanover

I use a Peterson rectangular pump that has split open the pressure section twice.

Rectangular is a bad shape for a pressure vessel.



I now have A Moroso as a back up.



I have 85 PSI at idle, but my idle is at 2,000 RPM, so I cannot compare that to yours. What do you have at 2,000 RPM?



If you have a big cooler, cooling the scavenge side is not a problem. If you are short on oil cooling, then the pressure side would be better.



I have two Setrabs model 50-634-4063 (13 row) in series on the scavenge side, and one in the pressure side.



I use two K&N oil filters, P/N 3001. They have a 20 micron rating and a 550 pound burst can.



Here is a pix of the coolers.



Lynn E. Hanover

83turbo

So the baffle plate is a useful item? I wasn't sure and didn't want to use one if there was any possible detriment.



Interesting info on the old style pump. Here's a quote regarding the infamous "Granny's 4 rotor":

Jeff20B

Sounds like cavitation is possible with Granny's setup.

bill shurvinton

Sadly I haven't managed to run the car up to 2000 yet to see what the pressure is. Kids, work and the house seem to be eating all my time at the moment.



I might consider putting a second cooler in the pressure line, if I can find somewhere to locate it. As you can see space is a little limited with me, but I should be able to go in front of the main radiator if i need it.



Luckily air temps never get that high over here and, by the time I have upgraded my engine to the point where it needs the additional cooling I should have worked out how to fit everything in.



Bill

bill shurvinton

The real reason I went for the dry sump. I can get the engine really low, which has a significant bearing on the polar moment in a sub 1300lb car.

Lynn E. Hanover

My first attempt at scavenging the rotary worked real well. For the front I used the front pressure relief gallery. I just cut a notch in the bottom facing aft. I pulled the oil out of the stock fitting on the drivers side of the cover, where pressurized oil used to come out. I plugged the gallery to the oil pressure "O" ring pad.



For the rear, I used a stock pressure relief valve, modified by gutting it. I cut open a section of the side of the body about 1 1/2" long. The open side Goes down and is indexed and held in place by the jamb nut. I pulled the oil out through the stock banjo fitting where oil used to go into the rear iron. The oil gallery up to the filter adapter was blocked off with a paper gasket under the home made aluminum block that gets oil into the rear main bearing gallery.



Then the pan can be dead flat, and the engine still looks kind of stock.

There were no bug screens, and I had no problems because of that, but you could put in some "In Line" filters from Moroso or Peterson.





This bottom view shows the extra oil line to the front main bearing and the suction lines to the front cover.





Lynn E. Hanover



Lynn E. Hanover