2630 John Deere water coolant pump rebuild

Whoa! Page 2 here. Thanks!

Back from the dealer to return the shortie tube.

Things are looking up! No problem with the pump housing gasket since they had the right one in stock and it even was a hamburger's price cheaper than other one I had from them. They even had the seal and driver in stock--John Deere and their inventory has them as separate parts, but the kit--both together with a kit part number-- they don't have that and that fooled me last time there. Ha.

I now have everything I need to reassemble the pump, the top tube off eBay has been shipped and from the length measurement they gave me, looks to be the correct length.
 
Last edited:
Whoa! Page 2 here. Thanks!

Back from the dealer to return the shortie tube.

Things are looking up! No problem with the pump housing gasket since they had the right one in stock and it even was a hamburger's price cheaper than other one I had from them. They even had the seal and driver in stock--John Deere and their inventory has them as separate parts, but the kit--both together with a kit part number-- they don't have that and that fooled me last time there. Ha.

I now have everything I need to reassemble the pump, the top tube off eBay has been shipped and from the length measurement they gave me, looks to be the correct length.
How about posting the part numbers for the individual seal and driver parts you got and the kit number they are in, it might save someone hunting time, in the future.
 
Here is my parts list, all from John Deere. Thanks for helping me compile it.

Included is the short piece of coolant hose which goes into a metal tube headed toward the oil cooler. I found mine to be hard like a rock so it should be replaced, it is expensive. The long metal tube attaching to that was corroded underneath the hose ends, but it was salvageable after cleaning. Also in that circuit is the nipple pressed into the water pump housing--mine had corrosion where the hose inserted over it, but deemed viable after removal of the rust and scale with sandpaper. I sprayed some primer on it.

I don't know if that press-in nipple is available as a separate part, but I think it would be the devil to remove it. That's one of those things which should be evaluated before DIY rebuilding, which might tilt one toward a rebuilt, or new, pump.

I also included the gasket for the lower thermostat housing where that bolts onto the block--note that this gasket has a different part number than the gasket of similar shape at the thermostat itself. I reused the impeller.

Part numbers for 2630 John Deere early water/coolant pump rebuild:

Gasket, back plate-to-block: T20243
Gasket, pump housing-to-back plate: R55405
Bearing assembly, KOYO: JD9257
Seal, spring-loaded half: AR97882
Seal pressing tool, required for above: R518805
Seal, ceramic washer: R32174
Seal, rubber retainer for ceramic washer above: R40243
Tube/bypass insert, early engine serial, 2.4" long: R54809
O-rings (two needed) for tube bores in pump and thermostat housing: R56461
Hose, small, short, angled, water pump nipple-to-tube toward oil cooler: R48766 (sub. for U42685).

Thermostat: AR48675
Gasket, between upper and lower thermostat housing: T20215


*If one has the later engine serial, then two short tube/bypass were used instead of the one long tube for the early engine. The two "shortie" tubes need a section of hose between them fastened with hose clamps. If not rebuilding the pump, buying an aftermarket or rebuilt pump instead, the tube already included in the pump could be the short tube, so one additional short tube is needed, the one inserting into the lower thermostat housing. However, if using the pump for the early engine serial, I would guess that a short tube insert would need to be removed for the long tube's use, instead. Here are the parts needed for that, as follows:

Tube/bypass insert, two needed, later engine, 1.34" long, stepped diameters of 1.110"/1.125" : R59050
Hose between above inserts: R59052
Pipe clamps, two needed: ___________

The fiber-rubber gaskets are branded "Interface Performance Materials N-8092, made in USA". They are 1.0 mm thick. The spring-loaded seal is indicated on the package as also being a domestic part and the ink jetted numbers on the seal are "261513" and "11:25" . The packaging for the plastic seal tool indicates a domestic part. The packaging for the thermostat indicates it is a domestic part, as well. The bearing assembly is KOYO, made in Japan.
 
Last edited:
Hi again folks.

The bypass tube R54809 arrived today. It is a plain steel pipe with no varying diameter. If ordered from John Deere, the cost is 70$. The measurements on it are 2.380" long, 1.125" diameter, 0.065" wall thickness. Looking around on the internet, I see tubing with an OD of 1.125" (1-1/8") is a standard tubing size.

After removing the short coolant hose on the nipple coming off the right side of the pump, I found a fair amount of rust on the nipple. The nipple is a press fit into the housing and I wish I had checked it and removed it before painting the housing.

The nipple has not perforated, but it does seem from experience, that pit rust which has penetrated steel is quite hard to stop. Even after cleaning the rust-off and treating it, thinking it would be ok to reuse, doubt crept in, and I decided to order that nipple from John Deere. It is a press-in tube and has a stop step machined into it, part number is:

R34985 Water Pump Tube.

I am aware the old nipple it might be hard to remove, but after watching a few Youtube videos of folks showing tricks for removing hose nipples from water pumps of various old cars, etc., I think I can twist it out without catastropie, the key strategy commonly proposed being to insert something close to the nipple I.D. and twisting out the nipple with vice grips or pipe wrench, the insert being used to keep the tube from collapsing under the twist. Heat is also suggested by some.

Looks like I put this old pump on in 1989 as a rebuilt from John Deere, so it is 37 years old. I don't remember doing it, it has been that long ago.

One minor correction to my previous post (which I am no longer allowed to edit), is that the two-hole fiber rubber gaskets for the thermostat housings are manufacture's gasket type N-8091, not N-8092, just in case someone wants to cut these from sheet.

I'll also mention that I bought an aftermarket pulley hub to replace the one I broke. Looks about the same except it is not quite the same length--shorter. Glad I checked this! According the the manual, my JD pulley is to be pressed on until flush with the end of the shaft, but due to the shorter length of this one, I'll have to press it on a little past the end of the shaft to compensate.
 
Last edited:
The short hose barb nipple off the pump's right side, R34985, arrived today and I am going to post the dimensions in case someone needs to make one:

The inserted end is 0.725" in OD diameter, tapering up to 0.750" OD at the hose barb. It seems most likely that the tube was all one diameter but that the process which makes the hose barb ring expanded the metal OD to either side of the rise.

The diameter around the barb itself is 0.825".

Wall thickness is 0.065", which is noticably thicker than the nipple I removed. The overall length is about the same as the old nipple, 1.715". Insertion depth into the housing bore, taken from marks on the old nipple, is around 0.60". I should be able to simply tap the new nipple in, I think. Looks like plain steel, magnetic.

Because the insertion depth into the housing is not much, the old tube came out easier than expected, twisting with the help of a mini-pipe wrench, adding some heat to the inside of the tube from an oxy-acet torch, and inserting a close-fit plug into the tube to keep the tube from collapsing during the twisting.

Once I could twist the tube, I added some penetrant and twisted on it more for about a minute, then I re-heated, switched to some vice-grips and resumed twisting back and forth but now with a pulling force added to get the tube out.

On the bypass tube I mentioned in the previous post, R54809, I just noticed a burr in one end's chamfer. I'll smooth that off so it does not cut the o-ring during insertion.
 
Last edited:
Alas!

A problem with the assembled water pump: the impeller is taking excessive force to rotate!

Not sure what has happened here. This is not the first water pump I have rebuilt. I have an official JD shop manual and I have been careful. It seems there is excessive binding at the seal itself.

I can barely rotate the impeller by hand.

Figuring I would need to press it all apart again, I decided first to countersink the bearing in the bore a little, about at the edge of the housing bore chamfer (the manual says however to press the bearing flush with the housing). After doing this, I took the pump out of the press and for whatever reason, the impeller spins like it should now!

The spring-loaded seal did not unseat from the bore as I did this adjustment, and the impeller has been pressed on the shaft already, so the only moving object of any significance would be the inner part of the seal, a metal ring, which the shaft goes through and which contacts the white ceramic disc mounted in the impeller.

Now, the seal pressing tool, manufactured undoubtely of the finest plastic with utmost precision, is supposed to seat the seal in the bore, AND push the inner steel ring down over the shaft, in magical proportion.

Counter-sinking the bearing the little that I did, would move the inner steel ring, already secured against the shaft, away from the rest of the seated seal. Now it is properly rotating, so the freeing-up of the whole apparatus must have something to do with the inner steel ring needing more clearance from the part seated in the bore!

Not sure what is going on here, but my guess is that the "tool" pressed the inner ring in too far down on the shaft. I conjure that the the tool is out of spec. In other words, I am speculating that the tool pressed the inner ring down in the shaft too far in relation to the part seated in the bearing bore. When I pressed the bearing in a hair from the other side, that caused more clearance between the ring and the seated part of the seal and resulting in a normal force of rotation.

I am inclined to keep the bearing where it is, just below the housing surface. I find this interesting because in my very limited experience in the mechanical world, a bearing or seal would be inserted below flush anyway, to where the chamfer ends. In other words, if I had no manual and were trying to figure this out cold, the bearing is seated right now where I would have seated it.

I am likely to leave the bearing where it is now, since the impeller now spins like it should. Of course, now I will need to press the impeller in a little bit more so it will clear the back plate.

Regarding the impeller, that confuses me too. The shop manual says to press the impeller about 0.010 thousands below the surface of the pump casing, measured off a fin. But these fins are not perfectly flat in relation to the the center of the impeller, all of them tapering down a hair from the center. So, getting ten thousands clearance at a fin, leaves the center of the impeller just a little proud of the case! Not sure what is going on here either, but I'll have to take clearance at the center of the impeller, the hub, rather than the fins, otherrwise I'll have interference at the center!

This means there will be more than ten thousands clearance for the fins. My impeller looks ok, and maybe it has had wear but it doesn't look that way. Wish I had a new one to compare with.

I could also remove the impeller (which would require complete disassembly of what I just put together) and sand it down level with the fins. This might have been a good thing to do with the impeller before assembly, if I had noticed the fins and center not being level with each other.
 
Last edited:
I have a new impeller (T30897) in a kit I have. That is the same impeller part number I find in the 2630 parts catalog. I just checked the impeller, and it is machined flat across the fins and center. Sounds like you have a worn impeller. It would not make sense that the end of the fins is .010" below the flange of the casting and have the center proud of the casting flange, even with a gasket between the flange and back plate, I would expect an interference with the back plate. Grinding it down to flat is going reduce its effectiveness. Your tractor, your call.

Impeller.jpg
 
Thanks so much Jim. ME for checking yours and posting that. I will replace the impeller and agree that machining the old one flat is a bad idea. Come too far to do not to do it right. I think I can disassemble everything else and reuse.

Quite uneven:

IMG_0946[1].JPG
 
Last edited:
Undoing it all to start again.

I took the bearing out and pressed-off what we now know is a worn-out impeller, then used split bearing plates to press the spring-loaded seal off the shaft.

I noticed that the spring-loaded seal has two black washers beneath the inner steel cup/ring/collar. Holding the outside of the seal, you can twist the inner steel cup with its washer, against the washer under it. These must do the main sealing. The top washer could be considered part of the rotating shaft, and the bottom washer, part of the fixed housing. A spring holds pressure between the two washers at all times.

The seal is 0.860" in length, uncompressed, and 0.760" compressed, therefore the travel is 0.100".

The seal rotation feels gritty and I noted tiny black particles coming out between the two black washers which rotate against each other.

So, I am going to replace the seal, fearing it was damaged during installation; maybe pressing the shaft out of it wasn't good for it, either.

I think the seal overly compressed due to the plastic insertion tool being off.

Placing the tool over the seal uninstalled, the outer lip of the tool does not contact the metal ring which inserts into the bearing bore: there is about a 0.075" air gap (interestingly about the same distance I counter-sunk the bearing from the other side, as mentioned in my previous post, to relieve the seal binding).

So, when this tool, as it is, stepped, first starts pressing, all the pressing force is on the inner cup/ring/sleeve! When the outer ring of the seal meets the bore, the spring compresses past resting position, as the outer diameter of the tool contacts the outer ring of the seal, until the outer ring is seated. How much it gets extra compressed is the 0.075" gap on the tool.

Since the travel of the seal inner ring and spring is 0.100", and this tool compresses, and leaves, the spring down 0.075", that leaves the spring just 0.025" from bottoming out at the seal's final resting place.

My theory is that not enough force is shared with the outer ring during the final pressing. Even after the tool is removed, the spring will be extra compressed and the compressive force of the washers against each other will be excessive, causing the shaft to bind.

When the seal is finally seated, the spring cannot rebound, since the inner cup/ring/collar is pressed onto the shaft and not loose on it. If the spring is compressed more than at rest, the two washers will bear the force of that extra compression.

Seems to me that the tool should be pressing the inner ring and the outer ring evenly from the start. Then, when the outer ring is fully seated in the bore, the relationship of the inner ring to it would be about the same to it as at rest (uninstalled). The two black washers are still in compressive contact with each other, even at rest, and now then full length of the seal is available to contact the other seal imbedded in the impeller.

I'll think about this some more and look at the situation again tomorrow, will try to get a picture of the tool over the seal.

At any rate, the tenative plan is to use a machinery bushing under the outer diameter of the tool when I press in the new seal, maybe around 75 thousands thick, whatever will maintain the relationship of the outer and inner seal like the are at rest. The outer ring will seat fully in the bore without leaving compression on the spring over it's resting preload.

If there is any extra compression on the spring, it should be coming from the seal on the impeller as the impeller is pressed on to its proper relationship with the housing, I would think.
 
Last edited:
Found a new JD impeller at a dealer, in stock, so today, Friday, I drove to get it. Looks much better than my old one and everything is level across the top!

They also had the seal tool, but it looked exactly like mine, and with the seal in it also had the same gap between the outer diameter and the seal outer ring! However, today I watched a video where it looked like there was no bottom gap on a plastic driver being used. Also, it looks like there is a metal version of the this driver available.

I also bought another new seal. The gap between it and the bottom with my plastic driver tool is 0.10", as shown in the picture:

IMG_0948[1].JPG



What 0.10" gap means, along with the 0.10" of spring travel, is that when the seal is fully seated, the spring is fully compressed and there is maximum force on the black sealing washers.

What I'll do is press the bearing back in, then press the old seal back, using shims filling the gap, as an experiment.

I'll leave the impeller off until I am satisfied with how the shaft turns, the put the new seal in and install the new impeller.

I may even order the metal version on the driver as a comparison, but probably not since it is 50$ shipped.

......thinking about this a little more, another reason that the spring should not be compressed at all past the pre-load is this: the ceramic sealing ring inset in the impeller has to be in contact with the surface of the inner ring/collar to be of any use. I suppose you could call this a secondary seal. I would think that there should be full travel of spring available, so the impeller seal will seat against it when the impeller is driven on to proper clearance with the case edge. I think the travel in the housing seal is to accomodate varying position of the impeller seal as the impeller is pressed on the shaft. This would also mean, I think, that the insert in the impeller can re-position the inner collar of the other seal to where it needs to be.

Another reason I think I need to install the seal with my tool modified.
 
Last edited:
Wow. More surprise! Now it is the impeller.

Seems that my old impeller was not worn-out! The problem it has is that it wasn't milled flat on the top after casting!

Looking at the picture below, the impeller on the right, it looks like the milling machine caught one vane entirely (at 11:00 o'clock) and a piece of the one next to it (at 9:00 o'clock), but didn't touch the rest. The way it was cast, all the vanes tilt downward from the hub (like an umbrella) so it should have been fully milled. Obviously it has worked this way for many years, but my main concern with the new impeller is proper positioning of the seals, efficiency after that.

The backside of my old impeller is also unfinished, but that's no consequence.

Obviously, if I were to re-use my old impeller as is, it would have had to be done with clearance to the backplate measured at the hub, leaving the clearance at the vanes way over the spec given in the JD shop manual.

The heights of the vanes measure in the perimeters are the same, as are the outer diameter of the impellers.

Here is the comparison:

IMG_0949[1].JPG


There is a very small difference between the distance between the top of the hub to where the seal face is going to rest, that is 1.075" on the new impeller, and 1.050" on the old impeller.

Another difference is the cavity where the seal and rubber cup seat: the cavity for the seal is more shallow in the new impeller so that the seal seats a bit proud. I don't see that as being a problem, but I can't help but think that the depth of the seal cavity in the new impeller should have matched the depth in the old cavity.

I can only hope that when I install the new impeller to the spec of 0.010" from the backplate, that the seal in the impeller will mate with the housing seal correctly!

A comparison of the depth of the impeller seal cavities, obvious difference, about 30 thousands:

IMG_0951[1].JPG
 
Last edited:
Ready to press in the new seal. I'm not going to bother with experimental pressing with the old seal, feeling sufficiently confident.

I settled on a 0.060 thickness machinery bushing to go between the outer gap of the seal press tool and the seal lip. Actually, the gap is not even all the way around, and seems more like 0.075" on the average.

This bushing is a loose slip fit in the gap and but still allows me to compress the spring a hair's width, which I can see at the area where the inner collar bellmouths to secure against the outer ring, viewing the bottom of the seal.

If the seal were driven so that the driving force favored the outer seal, the hazard is that the inner collar will tear away at its bellmouth and ruin the seal.

So, I would rather err on putting a hair's worth more driving force on the inner sleeve than the outer ring. We'll see.
 
Disaster again.

I seated the housing seal with my modified pressing tool and everything looked good! I could twist the shaft with my fingers and it was smooth!

However, when I pressed the new impeller on, by the book, at 0.010 clearance from the fins to the housing edge, the housing seal was compressed too much from the impeller and that cracked the housing seal. When I took the pump out of the press, the impeller was impossible to turn by hand!

I removed the housing seal and, indeed, the sealing washers had cracked!

Now, I think that if the new impeller had a recess for its cup and seal identical to my old impeller, I would have had success. The seal bore in the new impeller should have been about 0.185" deep instead of 0.110" deep, a 0.075" difference. Having that cup and seal sticking halfway out of the impeller just does not seem correct, regardless. Maybe the impeller being different at that spec at that point is the very problem.

I have the option to give up trying and buy an aftermarket pump.

The other options are to try again with a new seal and

1) reuse my old impeller as is. Do my clearance measurement off the center of the impeller instead of the fins.

2) buy another new impeller and have a machine shop deepen the bore to 0.185" for the impeller seal and cup. Buys me an extra 0.075". We know the spring seal is being compressed 0.10", then the washers are getting crunched. Backing off 0.075" might just do it.

3) reuse my old impeller. Have a machine shop level the top or try it with my belt sander, which would buy extra room between the housing seal and the impeller seal. This option especially if option 1 cracks the seal.

Needless to say, this is discouraging. Looks like trial and error to get it right.
 
Last edited:
Yesterday's Tractor Forums

We sell tractor parts! We have the parts you need to repair your tractor - the right parts. Our low prices and years of research make us your best choice when you need parts. Shop Online Today.

Back
Top