I've enjoyed reading Dennis' work, I've not much to add but wanted to mention this, from the pics, the spring does not look symmetrical but progressive. Note the right end appears larger diameter winding than the left. If this is the case and not just a optical delusion the tension will not be linear but logarithmic. So the # of turns are not straight forward. ie each turn does not produce the same change.

Good Luck

 

Thanks for the interest. I think what you see there may be just an artifact of the camera. In person, looks like a plain-vanilla, constant-everything spring.

Stiff little guy it is too. Only ~ 2 mm of travel from 0 to full compression. Too stiff to even budge it between your fingers.

 

I was just thinking about the car's that still have the old oil (the oil that thicken as time goes on). Could this valve be the culpit to the blown hoses. The oil gets think, the valve get clogged and pow! hose lets go. Hummmm!

Chuck
05/XK* Conv. 1 of 200

 

Dennis did I miss a post with a picture of the spring taken out of the valve. Sounds like Jamdmyers (above) has a picture.
Just checking.

Chuck
05/XK8 vert

 

Here you go Chuck......

 

Sorry for the delay, gents. I hope to be able to put up, by Sunday, some facts and figures that can be used to reset the valves.

 

When you say spring, that's one hell of a spring.

Course 1600 lbs of pressure is a lot of pressure. Seeing this spring, leads me to believe that very small turns of the screw will produce big results. However, it may take larger turns to change things. Some of the engineer types will know more.

Now what?

Chuck
05/XK8 vert CF 1 of 200

 

Using the adjusting screw on internal relief valve to change relief pressure

I've been trying to work out a rule for how the relief pressure varies as we rotate the adjusting screw. The best method found was to use the stock valves marked 110 Bar and 85 Bar as "standards", that is to assume they are accurate and then to infer a rule based on the differences between their relief pressures and screw positions.

But practical considerations really get in the way here. As Chuck S. pointed out, this is a very stiff spring and so very small changes in screw position will create large pressure differences. Also, we are operating near the lower limit of pressure settings possible with these valves. Overall, there is pretty large uncertainty (see below).

Also, because only very small adjusting screw rotations will be used it seems best to talk in turns of "flats" rather than "turns". A flat is the rotation that takes you from a spot on one of the six hex faces of the valve body to the corresponding spot on the next face (i.e. 1/6 turn). Flats are much easier to "eyeball" than fractional turns when making a screw adjustment.
So, with all that said, my best estimate is ...

Relief pressure of our internal relief valves varies by ~ 200 PSI per flat of screw rotation, +/- about 15%.

For example, to adjust a 1600 PSI valve to 1200 PSI we would rotate the screw counter-clockwise by 2 flats.


Testing Problems

My intent was to take a 110 Bar valve from my "backup" pump, set it to about 80 Bar, and then insert it into the latch circuit of my "working" pump, the one in my car. But when I opened the reservoir on the working pump I found ... see below. The valves look entirely different. Long story short, these valves mount to the pump with the same threads as the older style valves taken from the backup pump, are marked as 110 and 85 Bar as before, but something else is going on ...

When I first installed the reset "backup" valve, the top would not operate. Pump running, but insufficient pressure. Tried raising the setting on the "backup" valve in increments, but the top would not raise even when the valve is returned to its stock position, 110 Bar. The top does operate just fine when I re-install the original valve.

So at the moment, I've got at least these possibilities:
- in working with the backup valve it has somehow been changed so that it no longer functions as it did when stock, even with the screw set to the stock position. (A change in the size of the outlet orifice?) Or ...
- a 110 Bar valve of the type found in the working pump is somehow not the same as a 110 Bar valve of the type found in my backup pump.

Not understanding what's going on, I'm reluctant to modify the original 110 Bar valve from the working pump.

I will try to get some info from Power-Packer, but based on past experience I'm not optimistic. Any thoughts as to what might be happening would be much appreciated.

For the moment anyway, I'm completely stuck.

 

That's a pain Dennis - why is nothing every easy!

The fact that it works when you put the original valve back points to the problem being the modified valve. So it's either opening at too low a pressure or else fluid is leaking past the valve into the reservoir. Even though you returned the modified valve back to it's 110 bar state, it could be that disassembly of the valve has thrown it out of calibration. I don't know enough about the mechanics of springs but perhaps relaxing the spring means that you cannot rely on just screwing in the same number of threads to get you back to the 110 relief pressure - assuming that's what you did.

Could there be any way the modified valve connection onto the pump is not the same thread as the original? That could be a potential leak path. OK, I'm clutching at straws now.

Regards

Peter

 

Dennis, Thanks again for your efforts and I wish I was closer, love to be involved in this, even though it would all be way over my head.


Wayne

 

Can you turn in the screw one flat at a time and try it each time? It's tedious but without a gauge, what else can you do at this point? Eventually it should work.

 

Not being a fluid dynamics guy I'm trying to figure out how the valve actualy works. My guess is this valve is open to the pressure line on what I'll call a T tap. When the pressure on the line exceeds the set pressure of the valve then the pressure is relieved by the valve venting the pressure into the resavore. Does this sound correct?

With this being said it would mean the pump is actualy able to produce more than 1600 psi. The danger here is without a gauge to calibrate the valve and test it's operation you run the risk of exceeding 1600 psi and blowing your latch lines while testing.

This is not something any of us on the forum want you to do on your own car for our sake.

It would also be interesting to see the stream coming out of the valve when it activates, I'd bet at 1600 psi that's one heck of a jet of fluid coming out the hole.

I wounder if this could be tested with compressed air from a dive tank instead of using the top pump to test. I pump my tanks to 2000 psi and a simple bench test fixture could be constructed.

As for the valve in your car looking different I'd venture to guess it's just the type of stock they were machined from, one being round and one being Octogon.

Be careful my friend ....

 

Thank you, gents, for the insights and suggestions.

Just to recap a bit ... This pump is what's called a "positive displacement" type. Such pumps must deliver a fixed volume of fluid per rotation. If there is no path for the fluid (dead-head condition), the motor & pump must stall, or pressure goes up without limit until something breaks. So the designers put internal relief valves on both the latch (1600 PSI) and ram (1230 PSI) circuits.

The valves are tapped right into the ports on the pump that connect to the hoses.

Since the 1600 PSI valve from my backup pump will not work when installed in my working pump (the top won't raise so we know it is now limiting pressure to some value less than 1000 PSI), I have to figure out whether it is because I have somehow changed it unintentionally (I agree with Jandreu that this is probably it) or because it was of different design in the first place. The way it threads into the pump body is identical to the new style valve; very unlikely this is a source of leakage, IMO.

I don't think we can test using air pressure, since the behavior of the valve might be very different than with a viscous liquid like our Pentosin, but if that's wrong somebody please correct me.

Also, as suggested, I'm trying like hell to assure that I never end up creating a situation where more that 1600 PSI can occur.


So some new questions please for anyone who might know ...

- How would you go about measuring a spring constant (how much the spring compresses per unit of force)? Or where would you take it to get such a measurement?

- When a relief valve is specified at, say, 1600 PSI is there a convention that tells whether this is the pressure where it begins to open, or where it is fully open, or something in between?

Any help much appreciated.

 

Dennis,

With the minute length/pressure relationship and the fact that the disassembly of the reservoir/pump is required for adjustments leads me to the conclusion that only a bench calibration of the relief valve is going to be a reasonable solution.

While I am sure there are those that will take the pump apart two or more times to adjust to minimum reliable operation I would prefer to pay a reasonable fee for such bench calibration. For someone with only a working spare pump, an appropriate high pressure gauge and some fabricated fluid supply/return lines it would seem pretty easy to adjust to something around 1100 psi.

tp

 

Thanks, Tom

I think the valves I took apart are not acting right because, well, I took them apart. Drilling a new hole up top -- increasing the effective orifice size -- may be the reason. Nobody else will have to take theirs apart of course.

I'm still hopeful we can do adjustments using a rule-of-thumb for screw position rather than having to measure each one. At least for those not concerned with max accuracy.

But you may well be right ... a little test rig might be the best way to set them. It would certainly give the most accuracy. I wouldn't mind doing this as a sort of community service. I've got a suitable gauge. Need to look into cobbling together such a setup ...

 

oNCE YOU HAVE A WORKING TEST RIG AND SUCCESSFULLY CALIBRATE YOUR VALVE SO THE LATCH WORKS, WE COULD SEND YOU OUR VALVES FOR RECALIBRATION. (Sorry, keyboard set on Autocad.)

Just a thougnt. We could pay you for your time and mailing. I would use a hand pump like a grease gun or similar to apply pressure to your rig. If I had some time, and a spare valve, I would try to rig an adapter. I have the gauge already off Ebay.

Also, have you contacted any local hydraulics companies that may be able to provide the service of resetting the valve?

 

I have been fascinated and impressed by this post and the persistence, expertise and knowledge of those involved - great stuff. As a TOTAL ignoramus when it comes to fluid dynamics and hydraulics, this may be total bull**** but ............ is it possible that the length of the "collar" on the shaft between the spring and the ball bearing could have an effect on the pressure required to open the valve. In other words - the spring is always the same length but this would very slightly change the distance between the spring and the ball.

 

Dennis, as our friend from MAUI said, please be careful. Even 10 lbs of pressure can hurt you.

Back to the problem.
Once you have the valve removed from the pump, look down in the pump, through the hole the valve was in, and see if maybe there is something else your not seeing. Also, look down in the valve itself, maybe the seat the ball sets on was upset and not as it should be. Your pictures show the empty valve, the screw, the spring, the plunger and the ball. They don't show a ball seat. Was the seat machined inside the valve? Unlikely as crude as the valve looks, so my assumption was that they machined a seat and inserted it into the valve. If this is true, it could have been upset somewhat and the ball didn't seat properly. My simple way of thinking is that if you put it all back the way it was, nothing basically has changed then and the thing should work. The size of that spring isn't going to let it change in it's ability to compress, so something is amiss.

I'll keep my thinking going, so don't let a small setback change anything. Remember this: It's been 50 years since John Glenn was the first man to orbit the earth. They didn't let little set backs stop them. ;-))

Chuck
05/XK8 Vert C.F. 1 of 200

 

There is (generally) nothing mysterious about a pressure relief valve. The relief valve when opened is bypassing fluid back to the reservoir. A bench test would involve plugging the return port, fixing a pressure gauge to the pressure port, and operating the pump to observe pressure. With the pressure port dead-ended at a gauge, the fluid will bypass at whatever pressure has been set by the relief valve. Test, adjust, test until you arrive at the desired value.

 

Thanks to all for the ideas. As the saying goes, I'm still confused ... but at a higher level. Comments please on the following:

I've been over everything a dozen times. the 110 Bar valve from the backup pump ... that wouldn't work in the working pump even when reset to its original position ... every detail of every component is where it should be, with the exception that drilling out the adjusting screw lock has probably left a bigger effective orifice there. But that can't matter until the valve opens, which should be at same pressure as originally. It should work. What else could be going on?

Imagine installing the valve's components just as shown in the photo, then turning the adjusting screw until the spring is fully compressed. Now back the screw off 2 turns + about 1/2 flat. That's the setting for 110 Bar. Back off another two flats and you're at the setting for 85 bar. Just another 1/2 flat or so and ... the spring is totally decompressed! You can hear things rattling around if you shake the valve, confirmed by calculation using the thread pitch and number of turns. So at the 110 Bar setting, more so at 85, the spring is almost fully decompressed. It can't exert more than a few pounds of force (confirmed by rough measurement), nowhere near the ~ 16 pounds needed to hold the valve completely closed up to 1600 PSI (110 Bar) or ~ 12 pounds for 85 Bar. (Ratio of spring force to PSI = size of the hole under the ball, ~ .01 square inches)

So it seems that the 110 Bar stamp on the valve does not announce that this is a simple 110 relief valve, but a valve that opens gradually, starting perhaps at a few hundred PSI, and when fully open, and with the installed orifice size, limits the pump pressure to 110 Bar.

Which fits with ... this kind of behavior, rather than the valve opening abruptly at 1600 PSI, is the only way I can explain why the valve I took apart did not work properly when put back to its initial setting. Seems to make sense only if the valve is at least partially open well below 1600 PSI.

Before I go any further with the implications of this ... what say you ... Possible? Probable? Lunacy?