2001 Prius Inverter Pump

I found out today that during the above mentioned 8.9V and below (no flow) condition, and the pump still making some noise it really is still flowing however reduced flow, that is just where the flow sensor switches from reduced pressure. Well it'll never run at 8.9V or lower in the vehicle anyway, so it's still a go for installing in the 03 of course after it soaks in antifreeze for a couple more weeks and survives that ok. All is working well so far. I need to change the inv. coolant in the 03 before this winter anyway so will kill two birds with one stone then. Next step is to build the led box - since the switch is closed with good flow, it will need a not gate -a single transistor, and led, a 9v battery, on/off switch, a few resistors, and a small plastic project box. I plan on Velcroing it under the dash..This way the led will be off when the flow is good and on when flow is not good or stopped. It certainly has pretty instantaneous switching with flow/no flow condition.

 

The flowswitch test setup has been running for over a week now in 50/50 antifreeze. I assume this is close enough to the Toyota long life red coolant, so if the flowswitch is still operational in a couple more weeks, will install it into the 03 prius with an led box under the dash.

 

With a Drimel tool and three cutoff disks, the pump came apart. There were earlier attempts involving hammers and screwdrivers that may have damaged the part beyond the original failure mechanism:

These are the major subassemblies. There are no seams, it has to be cut apart.


The cracks inside appear to be from an attempt to punch through the pump housing.


The impeller has a ceramic bearing insert that rides on the single shaft. The impeller assembly has four magnetic poles, two N and two S.


So it is a brushless design with the stator fields and control electronics in the base.

Bob Wilson

 

Interesting report of a new campaign to replace the inverter coolant pump for early model, NHW20s:
http://priuschat.com/forums/gen-ii-...4-2007-inverter-water-pump-3.html#post1218331

Bob Wilson

 

Thanks for the credit bob.

 

Indeed, thank you for the part.

There appears to be a seam, very well sealed, about 3-4 mm below the cut:

It is also evident on the connector interface. The only way it makes sense to assemble the part is the cover must be hot in a plastic state and pressed together to the base with the impeller. One puzzle remaining is where does the vibration come from.

I've looked at the shaft and there are no wear marks. There is enough of a gap that the impeller wobbles. Yet I don't see scruff marks on the impeller nor inside the assembly. There are four distinct 'nodes' as it rotates.

I put 12 V on the unit and it powers this unit's rotor in quadrants:

1. 1st forth is powered
2. 2nd forth is unpowered
3. 3d forth is powered
4. 4th is unpowered

What this suggests is there are two quadrants where the rotor in this unit could freeze and not move. In fact, if I slowly turn the impeller towards a power position, it vibrates and pushes back. A little more, ~5-10 degrees of forward force, and it passes into the forward, powered quadrant and rotates forward. The only way the impeller rotates is if it gets enough momentum to pass through the unpowered quadrant and overcomes the initial resistance and on into the next powered quadrant. But the slop in the shaft leads to vibration and high drag. This behavior suggests one pair of drive coils (or drivers) are dead but due to the potting, it is difficult to see enough of the logic board to ohm it out.

Given the pulsed nature of the OEM pump drive and small, single bearing, it is easy to understand why the ZVW30 pump was redesigned with two bearings. Again, thanks to Yakuzahi who provided the failed part.

Bob Wilson

 

Just started reading this thread and since I'm looking for an inverter coolant pump I found this on ebay: Toyota Prius New OEM Inverter Coolant Pump 2004-09 - eBay (item 270661494841 end time Jan-08-11 07:58:51 PST)

So, as I understand it, the purpose of these tests are to see if you can make a gen 2 pump work in a Gen 1 prius, because the new pumps are better designed. My feeling is that my pump has lasted 160k, and still works, so if a new one lasts another 100k, that oughta be good enough for me. But if I can get a gen 2 pump for $89 and it will work, hey I guess I should try eh?

 

Yes I got a new oem gen1 inv. coolant pump off ebay for $105 a couple months ago, and using it for experiments and/or my spare, our 03 is also still good. Well I think Bob is headed toward using a gen2 pump in gen1, however my test is to simply warn the driver if the pump ever fails, way before more expensive damage can occur. As it stands now I am ready to install the flowswitch inline with inv. coolant flow, as soon as I can find the time to do the coolant change as well. Just been too busy with other work to get anything done lately. I plan on just installing a rather loud buzzer along with the flowswitch under the hood that will give a sharp buzzing sound if flow ever stops -say going 65mph down the highway etc. The only problem I haven't figured out yet is the fact that when you shut the car off it will buzz as well, thus the need to open the hood everytime..I suppose the switch wires could be brought through the firewall but this will only be the first attempt at it just to see if all works.

 

My 06 was in that campaign,
I just brought my 06 in to Stapp Interstate Toyota get the "redesigned" inv. coolant pump done and sure enough I checked it a day later and there's no flow at all, the pump is running but no flow. This is the same dealer that likes to fill the engine oil 1/2 to 3/4 inch over full mark and now proves to not know how to bleed a coolant system. How much will you bet they didn't even bleed the system? They said the fluid would be changed, and on first look it does look like new fluid, but now no flow. I've been doing my own oil/filter changes on the 06 the last two times because of the above, and I did install an aftermarket eng. air filter, do you think they got mad and just didn't bleed the gen2/06 inv. coolant system?
I mean c'mon dealer service dept. can you do anything correctly?

BTW I sold the 03 here, it was a good car still but a buyer came along and gave what we wanted. So will not finish the above inv. coolant flow sensor test but I can say it still works good on the bench, maybe I'll put it in the 06 instead now.. lol

 

Sounds like if the pump was replaced and its not showing a disturbance in the resivoir you would need to:

Make sure the pump is getting power or plugged in.

and or bleed the system, it takes less than 30 min.

If you loosen the bleed screws on the front of the inverter, you can then squeese/pump the large coolant hoses to force the air out to the top. Thats how I did it and it worked well and quickly. Once you see turbulence in the resevoir you should be good from then on. Just check the fluid level a couple times for the next few days to make sure any additional bubbles dont cause your fluid level to drop. Then you should be ok.

 

Yes I did a manual bleed of the system, clear tubing from the bleed nipple right into the res. tank, and watched the bubbles go up the hose like a machine. After two times of this I got some flow back however it seems a much weaker flow than the original. The pump is always on as long as the ign is on, and there's no air in the sys now. The dealer said you should hardly see any turbulence in the res. tank (I saw rapid flow/turbulance before the changing pump) however there is some flow, just not as turbulant as before. I wonder if one of the "redesigned" things in this new pump are quiet (slower) flow? This is of course for the gen2.

 

2001 Prius.. Quit on me... P3125 flagged... Bought a used inverter and installed it today. Before installation I turned my ignition on, heard the coolant pump running and noted that the coolant was flowing through the reservoir. After installation, I followed the procedure of attaching hoses to the two bleed valves and running them back to the reservoir and filling it. But, I only got fluid out of one bleed valve, the other was totally dry. And, when I run the pump there is nothing flowing through the reservoir. The donor inverter came out of a 2003 model. So, any idea what I should try from here? I might add that the car has 279000 miles on it and everything is original except the foot pedal position sensor, the HV battery, the right rear wheel bearing and the brake pads.

Update: Took it to Toyota for diagnosis. They thought it needed a transaxle and possibly a new ECM. I bought a used transaxle and had it installed. Did not fix the problem but now the coolant is flowing good. Still flagging the P3125 code. So, my last resort is the ECM, but, I hate to spend even more money if that still isn't the problem. Does anyone know of anything else that could flag the P3125 code? I suppose it is possible that the used inverter I bought was bad even though they claimed is was pulled from a working vehicle. There are used computers available on Ebay but I don't know if it is possible to re-program a used computer to work in my car? If I have to purchase a new one and have it installed by Toyota, I will have spent about $3000 trying to fix the car which is pretty much the Kelly Blue Book Value. In retrospect I should have sold it for scrap but hindsight is 2020! :-(

 

It is very difficult to get air out of the Classic inverter coolant loop. I recall spending several hours on this when changing the inverter coolant on my 2001.

Nevertheless this must be done before you can put the car back in service.

If you have a handheld vacuum pump, maybe you could try applying vacuum on the valve which is dry to see what happens. You might also try taking a short drive around the block to see if that helps - but do not drive for an extended time without a functioning coolant system.

Worst case, you will need to tow the car to your local Toyota dealer for help. Good luck.

 

There are reports of a 'gel' forming in the cooling passages from what I suspect are aluminum salts. Hobbit found some and others have reported it, too. I suspect a pressure flush can get the gel out but I haven't had to deal directly with it before.

The two, tubes connected to the bleeding valves, you did see a flow a coolant through one back to the reservoir with the car turned on enough to run the pump but not start?

Thanks,
Bob Wilson

 

I did not try running the pump with the bleed valves open. I'll give that a shot and see what happens.

 

It's not that hard to get the air out of the system of a classic Prius. I have a few comments regarding air bleeding and pump replacement.I recenty replaced my inverter that had a heat death due to coolant loss, and I was encouraged to replace my pump while I had the system apart.

If, for example, you're doing a full inverter swap and also changing the pump, you can skip removing the bumper cover and and headlight to swap out the pump bracket. Look at your new bracket, you'll see 2 metal tab/arm that can bend down to let you slide the pump and it's rubber shock mount off the bracket. If you do this, you can skip the fairly involved process of removing the underside covers, the bumper cover, and the headlight. I don't know if this is possible if you're not removing the inverter, and I'd certainly suggest removing the bumper rather than the inverter.

With regard to bleeding the air, one thing that wasn't immediately obvious to me from the descriptions I saw was that you should be bleeding air only when the pump isn't running. The reason for this is that the bleed valves can suck air into the system if they are open with the pump running. So, with an empty system, open the bleeding valves and fill the reservior until it's to the full line (and close the bleeding valves if you get the pink coolant and no air bubbles coming out of them). Then, close the bleed valves and turn the ignition on (but not all the way to ready, you don't want to run the engine, just the pump). Now, the pump should run and will likely be making a bit of noise. It may also not be actually pumping (it's probably full of air, which it can't pump). If you open the fill cap of the reservior and look down into it, a circulating system will have a whirlpool of moving liquid. If you're not getting any actual pumping, it's useful to "massage the hoses". Locate the hoses heading between the inverter and the pump, and squeeze them by hand while the pump runs. If you encourage any coolant into the pump (and any air out), you'll hear a change in the noise. Do that for a minute or as long as you get any progress. Then turn off the ignition (and pump) and open the bleeding valves again, Any air you got out of the pump should come out of the bleeding valves. Add more to the reservior if needed. Then close the bleeding values and turn on the pump and massage again. Once you get the pump actually pumping strongly, a lot of the air will make it's way to come out at the reservior (filling) tank, but you'll still want to use the bleed values to remove any air trapped in the hoses heading to them. Once you get the air out of the system, the pump will be quite quiet.

 

That does match the procedure in the shop manual: close bleeders, run pump, stop pump, open bleeders, close bleeders, run pump, ..., oy.

A little further to that, I noticed a while back that only the bleeder nearer the midline of the car is on the suction side of the pump and likely to draw air in. The bleeder nearer the fender is on the discharge side of the loop, and coolant flows out this one if it is left open while the pump runs. That one can be left open to a hose that just dumps back into the reservoir.

Some time after I posted that, I went back out with a manometer and measured the pressures involved. With the pump running, pressure at the high side bleeder was about 18 inches of water, and at the low side about minus 10. It would be possible to keep both bleeders open with the pump running, provided some way to keep the low side at < -10"H2O pressure. The bleeders are more than 10 inches above the ground, so connecting the low bleeder to a coolant-filled siphon hose leading to a jug on the ground would do the trick. Of course this would be a small flow out of the system instead of back into the reservoir, so it would be important to watch the reservoir and keep it topped up.

Then you could just run the pump, massage the hoses, and watch the bubbles come out.

-Chap

 

I thought the protocol was to run two, clear tubes from the bleed nipples to the reservoir. Turn on the car and when the air stopped coming out, tighten the nipples and remove the tubing. Then carry some spare coolant for the next couple of days and top off when the reservoir dipped down . . . or that is what I remembered.

Bob Wilson

 

I think it has been described that way in some places, but not in the shop manual, which clearly instructs to open the valves only when the pump is off, close them before starting the pump, and to do this iteratively. The idea of just cracking both valves and letting the pump run is a first reasonable attempt at coming up with a procedure less Sisyphean than the one in the shop manual, and it's pretty much impossible to go through the procedure in the manual without getting passionate about finding a better way.

But the reason the first attempt doesn't lead to an effective procedure is that with both tubes run between the reservoir and bleeders, only one tube will be carrying flow to the reservoir. The other tube will suck from the reservoir.

Clearly that tube will have to be secured deep enough into the reservoir to suck coolant instead of air, otherwise the game is lost. At least if it is well-secured and kept immersed, the pump could be left running and there would be a harmless coolant circulation from the reservoir, in one bleeder, and back to the reservoir out the other bleeder (and some bubbles will be carried out through that tube). The question that remains is, even supposing that's done right so it doesn't suck any air in, and allowed to circulate for any chosen length of time, does it eventually carry all the trapped air out via the "exit" bleeder? It doesn't, because both bleeders are placed at local high points of the loop, and air at each high spot will remain trapped unless/until it can flow out through that bleeder. The one bleeder whose natural flow direction is in can't function as a bleeder while the pump runs, without some way of changing pressure differentials so it flows out.

-Chap