Now that we have mostly settled just what the pressure reduction resistor does and does not do, the obvious question (which some are already asking) is what can we do to reach converitible top pump pressure reduction nirvana.

Let start out with a set of desired features:
-Sharply limits the presssure to 950 PSI like the relief valve, in other words a true regulator
-Does not require opening up the hydraulics to install like the relief valve
-Totally plug and play in 3 minutes like the resistor
-Inexpensive
-Does not slow down the top like the resistor or effect it in any other way
-Reliable

As others have noted, there are two commonly used approaches to do this in industry: one is a mechanical pressure relief valve and the other is electronic closed loop feedback PWM. In a nutshell, closed loop feedback PWM drives the pump motor with pulses whose width is set by measuring the actual pressure. This sets the speed of the motor, and thereby can control the pressure of the pump. Simple PWM circuitry itself can be fairly inexpensive, but it requires an electronic pressure sensor. That would bring it to at least a cost level in the same ballpark as the mechanical pressure relief valve and requires openning up the hydraulics to connect.

So if these two 'in the box' solutions are too expensive and difficult to install, we have to look a bit outside the box. If you look at the pressure and voltage plots I took with the .2 ohm resistor, you will notice that the voltage reduces across the motor somewhat in proportion to the pressure increasing. Since the voltage source is a close to a constant, this means the voltage across the resistor in increasing. By ohms law, when you have a constant resistance and an increasing voltage, the current is increasing. So the conclusion is that the current increases roughly in proportion to the pressure. This allows us to monitor the motor current (cheap and easy) and determine with a fair bit of accuracy what the pressure is. This means we should be able to build an accurate enough PWM controller and without an expensive pressure sensor!

Another observation I made (but did not post yet) was that the motor, although it is driven both forward and reverse, only generates the high pressure spikes when driven forward. This information lets us make another simplifying assumption: that the existing relay in the car can be used to drive the pump in reverse, and the PWM circuit need only drive the pump forward. This assumption significantly reduces the complexity and cost of the PWM drive circuit. It also means that all of the signals needed can be accessed by unplugging one relay in the trunk fusebox and replacing it with this new device which I will call the 'pressure limiting relay'.

Right now this is just in the concept phase, but I have decided to put together a proof of concept device. I'll keep the forum posted.

 

Cool, with something like this you could even speed up the total operation by adding an add on module to increase the voltage


Don't know enough about electrics, but it sounds plausible what you suggest, nice plan.

 

Thanks for the idea Andre! That puts an interesting 'performance enhancement option' (no blue pill jokes please ) spin on the problem. I might play with that a bit and see what is feasible.

 

I'm pretty much on the same page with a few exceptions.

- We know what a 0.2 ohm resistor will do, not what series resistance in general can do. It probably be taken a bit further with higher resiistance, but I'm also inclined to go in a different direction.

- PWM is ideal when you'd like flexibility in controlling voltage, power efficiency, and can afford it. I think here it might be cheaper and equally effective to use a fixed voltage reducer, zener-based perhaps, to reduce pump voltage .... but only during latch operations, triggered by the latch solenoid on the pump. (The zener and latch solenoid ideas were discussed in earlier posts.)

- I don't see much payoff in treating the two pump directions differently. (The recent raise-top graphs do not show it of course, but there IS a high-pressure spike during the open latch operation ... at the beginning of the lower-top cycle. There the latch is under mechanical load, as it is gripping the latch.)

- while it is definitely safe to decrease pump voltage, it might not be safe to increase it. The risk/reward here doesn't work for me.

So we might end up with two potential solutions to let people chew on. Always a good thing.

 

I think I figured out why the latch wouldn't close with the 0.3 ohm resistor during my experiments. The latch solenoid needs at least X volts to operate. With the 0.3 ohm resistor installed the combination of the resistance and the motor load dropped the voltage to the solenoid below that which it requires to operate. Therefore the sequence would stop at the point the solenoid wad supposed to activate. I didn't post the first video with the 0.3 ohm resistor, I was using WhiteXKR's timer to close the top and I thought it had just timed out. Then I got my wife to hold the button for me and the same thing happened all over again; the sequence just came to a dead stop at the point the solenoid activated. I think if you're going to design this system, you need to verify the working voltage of the solenoid.

 

Dennis..thanks for your comments.

High wattage zeners are expensive and inefficient, requiring a large heat sink. The classic solution to this is to use a small zener with a pass transistor, but this is still inefficient and requires a large heat sink at the power levels we are dealing with. Ths technology is dated and has largely been supplanted by switching power sources using low cost, super effiicient high power MOSFET transistors. These tiny power transistors can switch upwards of 30 AMPS without a heat sink! Switching power supplies also operate on PWM principles. The bottom line is this technology is cheap, space efficient, runs cool and is widely available.

Using the the pump solenoid as a trigger is definitely a simplifying option, but I can build a current sensor for the same cost or less as the connector to the solenoid and simplifiy the installation in the process.

Since MOSFETs are unidirectional devices (as are zener regulators for that matter), only dealing with one polarity is a simplying and cost saving approach.

Increasing the pump voltage is not being proposed as a general part of this solution. It would only be considered as an enhancement option, after adequate testing. This is an interesting option to explore, but is not the main thrust of this investigation at this time.

 

My phone screwed up and posted my last post before I got done typing it. I edited it and finished it. You might want to go back and read that comment.

 

Steve,

Your circuit design skills are surely beyond mine, and mine are pretty stale in any case. I'll read up on what you described. Thanks.

Also, please look at the edit to my post below ... there is a high-pressure spike during open latch ... but you won't see it during raise-top. It happens at the beginning of lower-top.

Sam,

We're on the same page, I think. The series resistor system sort of falls off a cliff if there's too much resistance. Two things are happening at once ... the motor is trying to pull more curreent as it encounters more mechanical load ... but that very thing makes the voltage drop across the resistor go up ... and so the pump voltage goes down. A double-whammy, and we get failure-to-latch. Agree?

 

I am confused. The solenoid is driven directly from the security and locking module...I do not see how the voltage across the motor can affect it.

 

Well, if that's the case then I guess I'm wrong about that. I'll post my first video with the 0.3 ohm resistor here in a second. If you watch the two 0.3 ohm videos you'll see that the cycle always stops right at the moment the latch is supposed to close. There isn't even a pressure spike because the instant the latch is supposed to activate the whole sequence just comes to a halt. Something is going on there, but I'm not sure what it is.

 

Your data closely matches the data that I took back when we first analyzed this problem and then recently when the resistor data was measured. If you look at the roof down cycle you will see a similar problem when the latch opens to release the top. In both instances the pump is held on for a second or so to guarantee that the operation is complete. The other pressures in the system are necessary to do the work and while not benign, are unavoidable if you want a functioning roof system.

I do note that your battery (under 12V) is not in very good shape. The battery voltage directly impacts the ramp rate of the pressure (versus the constant width of the pulse) and accounts for the different peak with resistor. I am quite sure that is your battery was something like 12.5V you would get results similar to what Gus and I measured as the ramp rate would be steeper.

When this was originally evaluated various solutions were discussed. My first thought was an electronic pressure control system using PWM control of the pump voltage. We also make a product for Jerr-Dan wreckers that regulates pressure based on engine RPM's that could be easily adapted. This is natural since my company manufactures products like this and I have controls on the shelf that could easily be adapted to do what you suggest. Given the variability of the cars (battery, mechanical efficiency, measurement tolerance, climate variation, wear and tear, friction, phase of the moon, etc.) I decided that it would be very difficult to make something that would be reliable under all circumstances and still be suitable for DIY. After considerable thought I went ahead with the relief valve because it is very easy to install and solves the problem regardless of the variables. The valve snaps off the pressure and the right place, does not affect normal roof operation, and is completely reliable.

I researched various sources for relief valves (including Chinese) and the valve we have been selling is the best I could come up with. The low flow characteristics of the roof system are not what stock hydraulic components are designed to to do so finding a proper cost to function match did not happen. The valve we are selling is therefore much larger that necessary for the job but the correct size doesn't (to my knowledge) exist.

As far as comments about difficulty of installation, those comments are coming from people who haven't done the job. The fact is this is easily done from the instructions in less that half an hour and without a drop of spilled oil. There is a lot of variability in the DIY skill and quality of tools so many take it to their favorite mechanic. We sell quite a few kits to dealers and specialty mechanics.

 

Here's the first attempt with the 0.3 ohm resistor:

Closing the Top, 0.3 ohm resistor, Attempt 1 - YouTube

And here's the second attempt:

Closing the top, 0.3 Ohm Resistor, attempt 2 - YouTube

As you can see in both videos, something happens right when the latch is supposed to close. The system comes to a complete halt and there is no pressure spike at all. The system doesn't even attempt to put pressure on the latch.

 

As I have suggested before, an excellent modification to the system would be to make the latch/unlatch a totally manual part of the operation. Eliminate all hoses to and from the latch and install some handle in the space provided for the allen key. The main problem would be the way all of the microswitches speak to each other to tell the system the position of the latch, the position of the top, etc. so trying to convert the latch to manual at this point would probably be complex. Still, it may be easier than trying to come with pump modifications and intricate circuitry to modulate pump pressure.


Doug

 

That's odd. How do you explain my results, which are nearly identical to WhiteXKR's? My battery is less than a year old, and it was fully charged during the tests. I started the tests less than 30 minutes after turning off the engine after taking a 90 minute drive.

As I've said time and time again, the relief valve is the ideal solution. You won't get any arguments from me on that point. It's just a lot more expensive than a resistor.

You're correct, I've never installed your relief valve kit, but I've read the comments from other members who have. Several of them have had difficulty installing it. At least one person had to take his pump to a machine shop to get the plugs removed ($) and another damaged a hose during installation, requiring replacement of the hose (more $). When I was hooking up the pressure gauge I had a lot of difficulty removing the plug into which the gauge was inserted. I came close to stripping out the Torx head on the plug. I wasn't too concerned, however, because I have the necessary tools to remove it if it had stripped. Others might not have the tools for that, and they would have to take it to a professional. That's not a criticism of the relief valve, it's just a fact that there have been a few reported instances of problems with the installation.

If both the relief valve and the resistor were free, I'd install the relief valve. But at $300, the relief valve has no value to me. If my hose bursts, I'll just spend $300 on the high pressure hoses and install them myself. I'll end up with hoses that I am certain won't burst.

Having said that, I completely understand why some people would choose to install the relief valve. If they have a burst hose they would have to pay someone to replace it. That could cost them $1800. They chose to install a relief valve for $300 and save $1500 in the long run.

 

Walt...I appreciate your comments. I believe your company's valve is an excellent product that works as intended, is not terribly difficult to install and is provides a fair value for the dollar. I also think your engineering decision to go in that direction was a prudent one.

I also think that the variablilty that you mentioned, if anything might be the reason my idea turns out not to be feasible...but I am not fully convinced of that yet. It is an interesting challenge, so I plan to explore it.

I would hope that neither one of us loses sleep over competition in this very small market.

And yes, I believe that you are correct that my battery is not in the best of condition and that could have skewed my results slightly, especially with the engine off. The only way to be sure,though, is for you and Gus to monitor voltage during your tests.

 

Some observations for the record:

I installed the pressure relief valve on my 2000 XK8 and found the installation to be very easy and would expect that most people with basic mechanical skills would find that to be the case.

About a season later my top latch hose leaked in the typical fashion - split at the terminating junction. (I had the old style lines with 11 years on them but fresh fluid). Don't know what to think about that, really; I have not heard of old hoses failing after the pressure has been reduced. I presume that at some point, any pressure at all will cause a badly deteriorated hose to fail.

I replaced the lines with Colliflower lines (get the link from Gus' website). Very high quality! In the process I checked the other junctions for leaks, etc.

I'm very confident in the set up now. It is somewhat expensive to get there ($275 +/- for the valve and $300 +/- for the hoses) but it seems like a solid and logical solution.

I've added the Key-Fob option so that once I install it I will have a reliable remote system.

 

If you need anything from me just ask.

 

Thanks guys for your hard work--both electrical and hydraulic methods. We really mean that.

But it's almost enough to make me put my top up in October and drop it in April. Then, keep it in the garage the rest of the year when not driving. Minimal cycles = less risk of the green curtain.

 

I hear you on that one. It took a while for me to get comfortable using the top again after the dealer replaced the hoses (leaking at the latch). Finally when I was happily using the top again, whammo, the pump dies! The dealer replaced the pump (it was not all that easy finding a replacement) but this second incident with the top has further spooked me about operating the top at all!

Doug

 

It'd be nice if some really knowledgable people could work together on this issue for ALL OUR sakes, for Jag owners and forum members alike. It would be amazing if this could happen.


As it stands there are now several threads dedicated to this issue, and back n forth it goes. All mods will be made aware to check these threads constantly, any bickering, name calling of any kind etc to be removed, infractions given, threads will be closed and persistant offenders removed from the site.

From today, edits will be made, posts removed as deemed necessary, infractions given and more, with regard to these threads, without any further warnings, so please try to get along!


Personally, I'd much rather go with option 1 if humanly possible