@Andre the Giant,
Turns out you are right... wife says while driving it today the noise is less. So I'll leave things alone, at least for a while. Now she says the "whining" noise has reduced to a slight "rattle." Sounds like things are improving.

 

 

Still too loud.

 

 

At this point my guess is the PS pump got some crap in a bearing somehow during that process. I would get a good used one (should be a dime a dozen) and swap it out.

Dale

 

The bearing in the pump is a sealed ball bearing, not in contact with any fluid

 

If the noise keeps up, I'm still planning on removing that high-pressure hose connector at the very top of the pump, which is higher than the fluid level in the reservoir and higher than any other point in the system (except maybe the high-pressure hose itself). I'll loosen the drive belt so I can turn the pump shaft by hand, slowly. If I see air, I'll put in fluid. If full to the top, I'll just give up and reconnect it. (Can't see how I could possibly add much more air. I have a little Dexron that I had to pull out of the reservoir since the Mazda dealer overfilled it.)

 

My uneducated guess, from looking at Section N of the 1994 workshop manual, is that the noise comes from the vanes of the pump scraping along the inner circumference of what Mazda calls the "cam ring." If there is air in there, I would expect noise.

 

Usually the noise is caused by "aerated" fluid, with tiny air bubbles in suspension. Once the air bubbles have time to escape from the fluid, the noise will lessen and go away. Yours hasn't gone away which tends to indicate another problem.
The pump is a positive displacement vane pump. There is no spot in the mechanism where an air bubble can just be trapped indefinitely, causing noise while it is in there.

The noise level has reached a plateau ? Not getting any quieter?

 

Andre the Giant,

Right, noise is still there, still worse when turning the steering wheel while driving. But I keep thinking that air could be collecting at the high-pressure hose connector at the top of the pump, above the reservoir level, above the input to the pump (level near the top of the reservoir), and is not staying in the reservoir, bubbling to the top as it should. There is no longer any foam appearing in the reservoir at all. That was gone even before we took the car to the Mazda dealer. And there was none when the car was returned to us. Except for overfilling the reservoir, I can't see anything the dealer did that would make things worse... it's pretty much as it was when we turned the car over to them. I guess "plateau" is the right word.

 

Air won't get "trapped" anywhere in the circuit, especially after putting some miles on it. The fluid is constantly flowing though the system even at idle. Not at garden hose flow speed, but flowing fast enough that any air is pushed along with the fluid and will end up back in the reservoir and separating out.

When you did the original hose change, and bleed procedure from the manual, did you have the wheels on the ground or in the air? If the wheels are on the ground, it puts way more strain on the pump when turning the wheels to bleed. It's possible the pump just didn't tolerate being run partially dry and having a load put on it combined with the wrong formulation of fluid.

I don't want to keep being the negative Nancy here but I'd also hate for you to keep chasing your tail trying to find a phantom cause for your problem.

I'd also love it if your problem just magically went away on its own
Some car issues are fickle like that so you never know.

 

Actually I deviated from the manual procedure and did NOT put the wheels on the ground as they specified for the engine-on portion. For both engine off and engine on, the front wheels were at first in the air. Also I did not pause the wheel at either right or left wheel lock position, since the manual warned not to leave the steering wheel in that position for more than 5 seconds (in one manual paragraph) or 15 seconds (in another paragraph). But after driving the car and having the noises still present, I did put it back in the garage and run the second part of the procedure with engine running while the front wheels were on the garage floor. It did nothing for the noise. If it damaged the pump, it was during this second try.

We will keep driving the car, at least for a while, but since no bubbles or foam are showing up in the reservoir, I am having trouble thinking that whatever air is left is still being circulated.

 

Sounds like you did everything you could to ensure that the process went smoothly....
Fingers crossed that it gets better for you.

 

Still noisy. There is a cutout through the pump body threads for the high-pressure connector bolt at the top of the pump, just below the connector, and way above the fluid level in the reservoir. I can see no fluid in there, which must be at the top of the vanes in the pump. I had already removed that bolt and the control valve below it, and the fluid level was about even with the fluid level in the reservoir.

My next idea is to use a piece of small-diameter flexible tubing, filled with ATF, and gravity-fill that cavity to the level of the bottom of the hole. We will run the car some more for a while, but if I can ever get my new computer working (another "project") I'll get back to that.

 

After another 100 miles running since the onset of noise in the power steering pump, the level of the fluid in the P/S reservoir has dropped from MAX to MIN, and the noise level has decreased slightly. Seems like this extremely slow process of ridding the system of air should have been stated in the manual. It's still a problem, but not as crazy annoying. Hopefully the less air in the system the faster it will recover completely. But it's not there yet.

 

Wish I could offer some advice. I had previously removed my entire PS system and went with a Maval depowered rack. Hated it and re-installed all the PS stuff. Basically went from a dry system and never had any noises from the pump. The bleeding process described in the manual worked for me immediately. Makes me wonder if there isn't something else wrong. How is your steering feel? Is it providing consistent assist at all times even when the noise is present?

 

Action of the power steering seems normal. Even when it is noisiest when wheel is turned, assist is there, then the "extra noise" quiets (mostly) when car is driven straight. But one can still hear the whining from the P/S pump as engine revs up. (I suppose I could tell people "it's the turbos" and get some respect!) But I have hopes that eventually it will quiet down.

 

We waited long enough for the noisy power steering system to clear the air out, and decided to do something about it. I removed the high-pressure connector (17 mm hex bolt) at the top of the pump, then unscrewed the 15/16" component below that. Looking at the hole that was at the top of the vanes in the side of the housing's internal threads, there was no ATF fluid apparent. So I used a piece of small-diameter vinyl tubing and manually filled that hole. When it was full, the level in the fluid reservoir was way above
MAX." Nevertheless I judged that to get rid of all the air, the level had to be that high. But if a lot of air was pumped through the system and arrived in the fluid reservoir, foam would probably overflow the reservoir.

So I found a large rubber hose to use as a snorkel and extended the height of the reservoir fill tube. Then we fired up the engine with the front wheels off the floor and wife rotated the steering wheel both ways as described in the manual. The reaction was awesome; foam came shooting out of the "snorkel" (see attached pics of setup) as the steering wheel was turned back & forth. In all the time the car was driven with air in the system, no foam appeared in the reservoir. Now suddenly the air was moving out of the pipes, hoses, and pump into the reservoir.

We haven't driven the car yet, but in the garage the pump seems much quieter and the noise when wheels are turned is much less. Hopefully the remainder of air will eventually be driven out of the system... the top is back on the reservoir, but the level is still above MAX; I'll leave it there for a while.

 

The above did NOT fix the noise problem, but a couple of days ago I removed the pump reservoir and checked it for leaks. There were none, but the "O-ring" at the pump input was really skinny and did not fill its notch properly. I replaced the O-ring with a used one from another RX-7 power steering pump, which fit much better. Also I traded the new hose I had installed with a used hose from the same "spare" pump. The noise problem went away completely. Not sure whether it was the O-ring letting in air to the pump or possibly the new hose letting air in somehow. But there are now no leaks and there is no more noise. I guess I'll never know for sure...