It's well known that the convertible top pump produces a peak pressure of about 1600 PSI. I've always wondered why 1600, why not some other number? What makes the pump act this way? It turns out that ...

The pump creates a peak pressure of ~ 1600 PSI because it has within it a relief valve set to 1600 psi. It follows that we can have lower peak pressure just by replacing this valve with one set to lower pressure.

See the photos below. We're looking inside the pump reservoir. The latch hose ports are on the left, the rams on the right. The three tubes with hex cross-sections are relief valves, and the one on the left is of most interest to us here. It regulates the pressure in the latch circuit, and is stamped with its relief pressure of 110 Bars (1 Bar ~ 14.5 PSI; 110 Bar ~ 1600 PSI). The presence of this component explains why the pump makes a peak pressure of about 1600 PSI. (The ram circuit is regulated to a lower pressure by the valve stamped 85 Bar, about 1230 PSI.)

It's known that only about 1100, not 1600, PSI is needed to operate the convertible top and latch. Since higher pressure probably contributes to hose failures, some guys have modified their pumps to produce only 1100 PSI by adding an external relief valve, along with associated hoses and fittings. Having now identified the internal 1600 PSI valve, the same result can be achieved just by replacing it with one set to 1100 PSI. No additional components, or fittings in the pump body are required.


Securing a Replacement Valve: I've located one supplier of a suitable replacement valve. It can be factory pre-set to 1100 PSI. But it has NTP threads and, not surprisingly, our threads are metric. It will be cleaner to use a metric-thread valve if I can locate one. I would buy some valves in quantity if there is interest.

Cost: About $50.00 for a suitable valve, best guess as of now. Nothing else is needed.

Testing: Testing should be done to confirm the pressure reduction. For this, the in-place 1600 PSI valve could be swapped with the 1230 PSI valve (on the ram circuit) temporarily ... just to confirm the principle. This would leave the ram circuit susceptible to higher pressure than before, so we shouldn't leave things this way permanently. It would be OK temporarily since we know that in a system working normally, the ram circuit sees less than 1000 PSI pressure.

Questions:
- Any comments or criticisms from the hydraulics practitioners out there? Anyone know of a possible supplier for a metric valve? Is SteveInFrance still around?
- Would anyone with suitable pressure-logging equipment be willing to test this, or to loan me the gear so I can do testing here?


Notes:
- The replacement valve can be installed without disturbing any hydraulic fittings, or even draining the reservoir. The technigue is to take the pump down from its mounting and position it vertically, motor side up. The reservoir can then be removed with very little fluid loss.

- The pump shown here is from a MY 1998 car, or thereabouts. Later versions of the pump have only two valves, not three. I don't know why this changed.

- The ports and valving within the pump are complicated, at least to me. I'm not sure whether the 110 Bar valve is regulating both the open-latch and close-latch pressures, or only close-latch. Because of how close to 1600 PSI both the measured open-latch and close-latch pressures are in a stock pump, it seems likely the valve is regulating both. Testing with a lower-pressure valve will tell for sure.

- An aside: the small hole shown in the upper right of the photo is a passage through the pump body, all the way over to where the motor is mounted. An overflow? It's vertical position is close to the upper of the two hash marks on the reservoir which serve as targets when adding fluid. This may relate to the specified fill points, top-up and top-down, but that's just conjecture as of now.

 

I'm in.

Maximum caution about removing the electrical connections to the solenoids.

 

Ah, yes, worth mentioning. The solenoid connectors on the pump do not need to be disturbed (instead, disconnect at the yellow clip attached to the tower that the pump is mounted on). But it's important to be careful with those connectors while wiggling the pump around.

 

Looking at the valve, it seems to have an adjustment. Can the relief valve be adjusted to a lower value without replacing (and if so how much reduction can be achieved)??

 

I thought about that too, but don't think it can be adjusted. The thingy on the end is an orifice or nozzle. (I'm told it is better practice to include this rather than to leave the output side of the valve wide open. I can see why this would be better for the motor, anyway.) There's probably some sort of ball and spring in there but ...

It looks as though the orifice was threaded in and then a pin or similar driven into the body of the valve from the side, locking it in place (see photo). The orifice will not thread out, even trying harder than I probably should have. I don't want to do anything destructive to it until sure of a working replacement. Maybe then, drilling out the lock, we'll find it can be adjusted. How would we calibrate it?

 

Hey Dennis--this is like deja vu all over again. After the endless debates about the benefits of one or the other possible solutions (pressure valve versus voltage) don't tell me you've come to the pressure manipulation dark side--yes?

Doug

 

I'm but wouldn't it be easier to find a used or broken pump and just swap out the ram valve for the latch valve. Or could the threads on awe valve be retaped.

 

I recall reading here about a company that rebuilds the rams, and possibly the pumps. Maybe they could chime in with a source for those pressure relief valves, or some technical expertise on why it will or will not work.

 

Sure, a pump being parted out could be a good source for a valve. Maybe we'll find or think of several sources. Requirements:
- settable to ~ 1100 PSI
- can live in an oil bath (Pentosin)
- M6.0x1.0 male thread, non-taper (I think, pending a check)
- fits within the confines of the reservoir.

 

I'll follow along on this. Sign me up if you find the right valve, the 1000psi relieve valve in my possession will take a lot of $ to hook up with the correct adapters and hoses.

 

That would seem to be the most expedient thing to do. Would the extra 130 lbs matter ?
Would still be 500 lbs lower than stock ?

 

Those pumps are going to be very expensive compared to a pressure valve. Not a likely solution path.

 

I'm in too if you can find a suitable valve. I'm based in the UK so I'm trawling google to see if I can find a metric valve from a European supplier.

 

That would be great, thanks. I will re-post the requirements when I'm sure of the thread size, but one added spec' in any case: thread length 11 mm or less.

What Sklimii said about modifying the valve that's already there would really be a great way to go, if possible. Zero cost.

If I just had one spare working valve, I would take a drill to one of them and see what's in there. If it's a simple ball and spring, the relief pressure should be proportional to length of spring compression. Might be possible to reset the pressure pretty accurately by changing the compressed length. It's not like we have to hit any particular number on the nose.

 

I would be fairly sure they are all a poppet and spring design along the lines of the Kepner valve in your original post. The orifice size will control the flow rate when it opens but the spring strength and compression will determine the opening pressure. We would need to get both the flow and relief pressure in the right ball park for it to work.

 

The 85-Bar valve installed on the ram circuit has the same size orifice, so it seems that component will probably be OK down to our 1100 PSI (75 Bar) target. Agree?

Having to do only a spring adjustment for opening pressure would be mighty nice ...

 

Agree. The orifice just needs to be big enough to relieve the full flow of the pump when fully open. Once you have the connector size it might be worth approaching Kepner's tech sales folks. They may know a supplier if they don't have a suitable part.

 

Yup, I have talked to Kepner. Nice people. They don't have a metric valve or know a source but I'm almost certain we could use their valve with an NPT / Metric adapter, and that should not be too hard to find.

I'm holding out for an all-metric solution at the moment.

 

!@#$%!! Every other fastener on this pump is Torx or metric. Even the snap rings (I think). But not the relief valves. These are 1/4" x 28 thread, which looks an awful lot like M6x1.0 when you're thinking metric, BTW. Very close. Sure fooled me.

Sorry for the wild goose chase, anyone looking for a metric-thread valve. On the plus side, it should now be much easier to find a few options for a replacement valve, at least here in the Colonies.


So I think the following represents an accurate set of requirements:
(Please note that these values supercede incorrect draft values given in posts #9 and #14 below.)

- relief pressure settable to ~ 1100 PSI
- able to live in an oil bath (Pentosin)
- 1/4" x 28 male thread, non-tapered
- max thread length = 7/16"
- fits within the confines of the reservoir.

The seal is not made on the threads, BTW. There is ... sorry, I don't know the correct name for this ... a sort of compression washer surrounded by a metal collar at the base of the threads to make the seal. See photo in post #1.

Everything sound right?

 

Everything sounds great, Dennis; you've done all of us a great service. Now it's just a matter of someone willing to be the guinea pig!