Let’s agree to disagree then, I see you will not let it go.

I am no expert here but have already some experience with different air pumps, and understand their specifics and know their best usage. That’s similar with dc motors, you use one that is designed for a specific purpose, and that includes not just voltage, but also current! Pending on the pump type that is used, peak pressure is about torque (or power, so voltage * current).

 

Sorry to hear I've left you with a bad impresson on this.

Was trying to avoid the misperception that the kind of DC motor we're talking about would be designed differently depending on whether it is to bel be "controlled" or not.

"Control" of these motors is nothing more than varying the voltage supply. The fancy-sounding ECU is really just a variable voltage reducer, often using a technique called PWM. A series resistor is a fixed voltage reducer.

For any applied voltage, The motor's current draw, torque, and RPM are then determined by the mechanical load on the motor. A google of "DC motor torque curve" captures much of this.

 

Hey Gus, how about I drive up there on Tuesday, you hook my car to the fancy pressure equipment, and I make a video of the test? Because I honestly don't know how my pump can be making 1600 PSI when about 1 time in 20 I have to start the engine to make the latch close. To me that indicates the latch is closing at the minimum possible pressure.

We can test your car then my car back to back. It'll be fun. I'll even buy lunch.

 

Best to replace the hose kit between the latch and pump. It's a 6 hour job from my experience with my XKR. I have a new HJB8256AB kit for sale since my 2nd XKR (daughters) leak was from the latch itself and that the hose kit had been replaced by the previous owner (who knew). Paid $228 but will consider offer. Contact zeecat@***.net if interested

 

I think this is a great idea.

BTW, I don't think your pump is making close to 1600 PSI when you are running off the battery with the resistor.....Gus's test results had the engine running, which added about 2 volts. This is not an apples to apples comparison.

 

This is an option and I could be up to this because I have nothing to hide what I am having a problem with is why do you think I am hiding something or misleading anyone?

 

And here I was just going to write to Dennis to just call you and have an open talk about the tests/results, as I was sure you would be open for that, so pretty cool that you offer time and equipment to do further tests!

 

I wasn't implying that you were hiding anything or misleading anyone. I was just saying that I don't see how the pump can possibly be making 1600 PSI with the resistor installed because every once in a while the pump on my car doesn't generate enough pressure to close the latch. I then have to start the engine to finish the closure. If the latch closes at around 950 PSI, and my latch doesn't close, I assume it's because the hydraulic pump was generating less than 950 PSI for some reason.

 

Dammit... My wife just reminded me that I have jury duty this week. We'll have to put this off until next week.

 

I look forward to the video.

IIRC, the tests were always done with the engine running. (#335). So might the 9V resistor target be exceeded with alternator power?

If nothing else, this thread has forced me to revisit Ohm's Law and applications thereof., together with the schematics of the convertible roof.

One thing I did notice, could not wear and tear in the latch components and hinges etc, have some effect on the pressure required to complete the 'close' operation?

A sticky latch, and or guide pins requiring more pressure to complete.

All good that you guys are going to get together, must be of benefit to all.

 

Sounds like a good idea to me.

It will be interesting to see the results and comments from proponents of each solution to the pressure problem after observing tests at the same time and being aware of all the conditions and variables.

 

I for one would welcome such a test with any "disinterested" 3rd party present and using Gus' setup to precisely measure pressure versus time/top position and plot the results.

To suggest that anyone here has any particular agenda for favoring one solution over the other is ridiculous. The relief-valve makers I am sure are not betting the success of their already established manufacturing company on the sale of a few dozen Jaguar valves and I am sure that Rev or whoever else is making/supplying the resistor solution finds it more of a burden to be providing these for a few bucks each.

So why not let's stop this nonsense about anyone having a particular secret motive or "faking" test results. And how about putting the brakes on spewing megabytes of bandwidth and storage on endless theoretical discussion on electro/mechanical/hydraulic theory.

Just provide pressure measurements for both systems under the same operating conditions using the same measurement equipment and provide the results in a graph similar to what Gus is presenting. Thus far the retort to Gus' information has been akin to "who are you going to believe...me or your lying eyes".

Forget about all the BS theoretical dissertation and merely provide the justifying data in an easy to understand graph--story over.

Doug

 

I want to go on record say that I believe that the data published by both Dennis and Gus were in good faith and honestly presented.

With this in mind, I spent some time this evening trying to reconcile the disparate results. This is my theory:

Dennnis used a simple mechanical pressure gauge for his testing. He discusses in his write up that he knows there are inaccuracies in this techniques when measuring pressure spikes (they will not read out as high as they should due to the response time [inertia] or the mechanism). With the engine running, Dennis measured 1340 PSI, and with a .25 ohm resistor, he measured 1120 PSI. This is a difference of 220 PSI, attributed to the resistor.

WIth the lab grade equipment, with 100 samples per second, Gus measured 1635 PSI. I think it is fair to assume that Dennis' 1340 PSI readng of this same parameter is due to the limits of his mechanical gauge. If we extrapolate Dennis' resistor pressure reduction of 220 PSI onto Gus' 1635 PSI reading we would expect Gus to see the pressure drop to 1415 PSI with the resistor....but instead Gus' recorded 1578 PSI with the resistor...a measly 57 PSI drop vs. the 220 PSI that Dennis saw. This is a 163 PSI discrepancy.

To me this 163 PSI discrepancy is explained by two factors:
1. If you carefully look at Gus's graph, you will see that the spike using the resistor is significatly narrower than the spike with no resistor. Apparently the resistor narrows the pressure spike (I do not know the physics behind this, but the data is clear). A narrower spike will exascerbate the performance limits of the mechanical gauge Dennis used, and it will show a lower reading than expected.
2. Dennis took his data with a .25 ohm resistor (.2 ohms was later recommended). Gus took his data with a .2 ohm resistor. This again will make Dennis's pressure reading lower than Gus'

In combination, I think that these two factors can easily combine to create the 163 PSI discrepancy.


With the above being said, I still would like to see a set of data on Gus' lab equipment showing pressures on battery power (without the engine running) as well as pump voltage data plotted along with the pressure data before declaring a final conclusion.

 

Entire post exceptionally well stated and clearly non-emotional and almost coldly scientific (a good thing). Sure beats certain of the folks here and one disgruntled poster on another forum (who seems to have a bug up his a** about this forum) whose primary response to those favoring the relief-valve is that the relief valve folks all have some hidden agenda and just want people to (unnecessarily) spend more.

Let's put all this hidden agenda BS to rest for good and just post results of each system using the same equipment under similar conditions under the watchful eyes of someone who does not have an "agenda".

Doug

 

Except... It doesn't explain why sometimes people have to start their engines to complete the latch closure with a resistor installed. If the latch takes 1000 psi to close, and the resistor is only reducing the pressure to 1578, then why isn't the latch closing?

 

I think that there are three factors in combination that can explain this:

1. The pressure with the resistor and the engine NOT running is almost certainly somewhat less than 1578...how much we do not know. This is why I suggested we need test data from Gus without the engine running and with voltage monitoring at the pump.

2. I believe there have been some users of the hydraulic pressure regulator that have had to adjust the trip pressure upward to achieve reliable operation.

3. To me there is little doubt that the factory system was designed with 'margin' to allow for aging and loss of lubrication in the system. If we reduce the peak pressures (using any means) on an older vehicle it stands to reason we may run into issues.

 

Thanks, WhiteXKR, for the effort that went into your analysis.

I'm sorry, but until Friday I won't be able to contribute much to discussing this. For now I will just say that my own results are not consistent with the pump making anything like 1600 PSI with 0.2 ohms of resistance in series. The discrepency is beyond experimental error.

An independent measurement is the best way to resolve this. Part of that, BTW, should include a measurement of voltage at the pump to assure that the resistance intended is really there and working properly. Something like 8 volts across the pump should be seen at latch closure time.

An aside: It's a normal part of technical work to assure that one's results can be duplicated by a 3rd party. This is not to accuse anyone of anything dishonest, but rather to be sure we haven't made some mistake. It's good (and standard) engineering practice to welcome this.

 

I believe that this article may have some relevant points to this discussion:

http://expert4pvc.com/Documents/Can%...at%20Gauge.pdf

 

WhiteXKR has made some accurate comments about the data.

The pulse magnitude with resistor is narrower because the pump rate (ramp time) is reduced by the slower turn rate of the pump. The maximum pressure is reduced because the pump leakage is a higher percentage of the total capacity of the pump as the rotational speed is reduced. If you were to dissect the pump, as I have done, you will see that the total pressure is limited only by the ability of the pump to continue turning and the relative percentage of leakage that escapes out the sides of the rotor.

Mechanical gauges will not capture dynamic values. The response time is slow and the gauge acts as a low pass filter on the reading. These gauges are good for low resolution readings on static pressure. The same is true with voltage and current measurements using a VOM. The meter filters the value and you only see what the filter lets through.

If you look at the spike that occurs at the beginning of the curve when the latch is opening you can see the effect of ramp time very clearly. The spike is terminated by the controller. The resistor curve maximum is therefore truncated due to a lower ramp slope. More resistance will increase the ramp time so the spike will diminish. The valve snaps it off at 950 psi regardless.

The new data only shows roof up. Similar things happen on roof down when the latch is releasing. The pump is held on in dead head condition for an extended period I assume to make sure that it fully released. Only Jaguar and Power Packer know for sure why they used the timing we see.

The ultimate flaw of the resistor method is that it affects the operation 100% of the time. This means that at some point (i.e. with more increasing voltage reduction) there isn't enough pressure to do the job at hand. This is why we chose to go with the relief valve as it allows the roof to work without impediment. I wanted a car that works as it was designed not some bailing wired contraption.

Regarding the comment that some have had to adjust the valve, I am not aware of that although it was suggested as a possible remedy. Some systems have had other problems including low oil and friction in the latch mechanism. There are a lot of variables in this mechanism so plenty of opportunity for other problems. If you look at Gus' new data on his web site, his roof closes about 1 sec faster than my roof in stock condition. This obviously relates to different amounts of friction in the roof mechanics. It was also noted that his roof does not retract as far when down. This is probably due to a different relationship between the cylinder switches and the rest of the car.

The comments that the pressure transducer is lying because it has an agenda are the stuff that makes me laugh. People try to help out and are accused of being liars and cheats. If anyone wants to take over this huge business of providing relief valves to cheap *** Jag owners be my guest. I did this to help out people who shared the same problem I had. I put a lot of time and money into sorting this out and my thanks are being called a bunch of names. I should have done this for my car and the rest be damned. We will continue to offer the kit at least until the current lot is off the shelf but after that it will probably not be available. This crap is a bit much.

 

Always refreshing to see a reasoned and detailed post here by Walt--welcome back (if even temporarily)!

Doug