Diaphragm 1st stages blowing out at depth?

[Eric Sedletzky]

So what kind of pressure is behind the diaphragm, is it the intermediate pressure of 130-145? And the spring counters this, plus pressure onto the outside of the diaphragm from depth makes it depth compensating as the diaphragm flexes away doing it’s thing. What is “overcompensating” and does that have anything to do with it?
So I’m guessing a huge blowout would have to be caused by a huge IP spike from a bad HP seat or seal.
Probably unfair to blame Zeagles for this. It sounds like it could happen to any diaphragm style reg.

 

[tbone1004]

So what kind of pressure is behind the diaphragm, is it the intermediate pressure of 130-145? And the spring counters this, plus pressure onto the outside of the diaphragm from depth makes it depth compensating as the diaphragm flexes away doing it’s thing. What is “overcompensating” and does that have anything to do with it?
So I’m guessing a huge blowout would have to be caused by a huge IP spike from a bad HP seat or seal.
Probably unfair to blame Zeagles for this. It sounds like it could happen to any diaphragm style reg.

so this is really really simplified
tank pressure pushes against the HP seat and causes the regulator to "seal". The spring presses against the tank pressure and spring pressure essentially equals the IP which is how it locks up.
When you breathe, the IP drops and causes the diaphragm to drop and stays down until the tank pressure pushes it back up and it locks again.
In an unsealed diaphragm, the water pressure also presses against the diaphragm and "adds" spring pressure so the regulator stays balanced.

When you seal it, the external diaphragm does the exact same thing but when a regulator is "over compensating" or "overbalanced" as they usually call it, the surface area of the sealing diaphragm is larger than the transmitter and causes a non-linear increase in ambient pressure. Think a thumb tack as an exaggerated example of how this works.

Now. When my failure happened, the diaphragm itself released before the second stage released pressure. Unsure why this happened, but it did. Mine had a 2-stage failure where the diaphragm failed, then the sealing diaphragm failed. If it was unsealed, it would have just blown and leaked gas out. Since it was sealed, it blew, leaked into the environmental chamber, then when that pressure exceeded the pressure the silicone could withstand, the silicone environmental seal blew out.

All diaphragms are subject to it, but it is not something that most people should be worried about. Had I checked the IP on that reg, I would have seen the creep and wouldn't have used it. This was spring of 2011, regulator got rebuilt immediately after with an Apeks parts kit *was a HOG D1 first stage*, and has not been rebuilt since. One of my students has it and it has at least 500 dives since then and it's still rock solid. Fluke incident

 

[grf88]

My son's D1 HOG had the same issue. It had less than 40 dives since new and failed in high current with a virtual overhead (shipping lane). Tank was almost empty by the time I shut it down as I could not see anything through the mass of bubbles. Diaphragm had extruded and envirnmental seal was shredded. I noticed in the manual that the original plastic washer was to be discarded and depending on the service kit should be replaced with the one provided or omitted if the kit did not contain one due to a thicker diaphragm being provided. This was an air share where I was thankful that I use a long primary.

 

[rsingler]

Torque is so critical for diaphragm reassembly.

Two examples of over torque damage.
The cut is self explanatory.
In the other, too much pressure had extruded the anti friction washer, so that diaphragm was flapping at the edges.

And undertorque is similarly a self evident problem.

Some diaphragm designs are more prone to issues based upon the edge clamp design.

Poseidon's XStream being the exception: a big floppy fabric-lined diaphragm with attachment under a keeper ring. Minimal force. Beautiful.

This was one CRUDDY reg that came in. But functioning flawlessly!

 

[Luis H]

The first stage diaphragm is exposed only to the intermediate pressure (IP). In a properly operating first stage, when the IP reaches its set (adjusted) pressure the diaphragm deflects against the spring and the external ambient pressure, to allow the valve to close.

A damaged first stage valve would allow the IP to increase, but the second stage (or stages) should vent any excess pressure.

It is important to keep in mind that the LP (low pressure) hoses are always exposed to the same IP pressure as the diaphragm. They are both exposed to the same volume of gas.

I can’t remember the published burst pressure of most LP hose, but I could be around 500 psi (maybe higher).

I have a hard time conceiving that the IP can runaway high enough to rupture the diaphragm or LP hose without first having a huge free-flow from the second stages.



IMO, the most logical explanation to this incident is from a defective or damaged diaphragm. It may be possible that it could have been a bad installation, but I am not familiar enough with Zeagle regulators to know if that is even a realistic cause.

I am very familiar with Conshelf first stages (and all its derivative regulators) and I don’t consider it to be possible for poor installation to be able to cause this kind of event, unless it involved an accidental cutting of the diaphragm during handling. It would be a very unlikely event.

The IP diaphragm for a Conshelf (or any of its derivatives) is a heavy fiver reinforced diaphragm with a large solid spring pad on the outside. I have a very hard time imagining it bursting due to increase pressure, without the LP hose also rupturing, but I guess it is possible; unless the diaphragm was already compromised with a cut or some type of defect.

I have been servicing this type of regulators for almost 50 years and I have seen a few (very few) damaged diaphragms leak, but never burst.

I have seen thinner HP diaphragms that belong to first stages made by other manufacturers, but I can’t recall which manufacturer. If my memory is correct at least one was a molded diaphragm. I remember reading the assembly procedure for that type of regulator, and it was critical to assemble it in a particular order. It seems that the procedure was critical to avoid over stressing the diaphragm.



In the case of a Conshelf type of first stage I have assembled in either direct: the HP valve side first or the diaphragm with retainer side first, and it doesn’t make any difference. I just have to keep the IP adjustment screw backed all the way out so that I am not fighting the heavy spring while assembling the valve side.

I never serviced a Zeagle regulator, therefore I have no idea what the diaphragm looks like and if it is a traditional thick heavy diaphragm or one of the newer thinner types.

Are pistons vulnerable to anything like this? Other than possible squeezing out an O-ring or something? I have several pistons and they are my go to regs.
Now I’m starting to look at my Conshelfs with suspicion, or am I being silly?

I don’t want to turn this thread into another diaphragm versus piston debate, but…

Couv and I were having some debate last April in Bonaire. He was only diving one this single hose things (just kidding) because he was testing some Freedom plate or something… He didn’t even bring one of his double hose regulators…

In any case, he said that years ago he either witness or experience more first stage failures in diaphragm regulators that in piston regulators. We didn’t get into the type of failure (catastrophic or minor leak), but in any case it is interesting information.

Personally I have seen very few undeserved (not involving accident or gross neglect) regulator failures and AFAI remember, they have all been minor leaks or free-flows due to contaminants or particles due to corrosion.

 

[Luis H]

I wrote my previous post earlier today before I saw @rsingler post.

Those first two regulators seem to have fairly delicate diaphragms and it makes sense that they require a fairly controlled torque-ing value and process.

Wow, I haven't kept-up with what the inside of some of the new regulators look like.


I have serviced some very old Conshelfs that someone in the past used a pick to remove the diaphragm and they ended up scratching the diaphragm sealing surface. The regulators were basically ruined, but instead of just trashing them, I just tighten the diaphragm until I got a seal (do not do this at home!). I have had to tighten the heck out of that diaphragm to get it to seal, but they always worked just fine.

Those regulators normally end up in my shop as utility regulators with air blow gun or filling tires, etc. They never get serviced and they continue to work without any issues.

I should point out that Aqua Lung does have two different thickness diaphragms, but both are fiber reinforced and even the thinner one is fairly substantial. I have used both and they are both available. VDH normally carrier a reproduction of the Aqua Lung heavy diaphragm. That is the one I use the most.


@rsingler, Can I ask what brand name are those two regulators?

 

[rsingler]

First one was a Poseidon Cyklon 300.
I'll look thru my pics to see if I can identify the second.

 

[Luis H]

First one was a Poseidon Cyklon 300.
I'll look thru my pics to see if I can identify the second.

Wow! It has been over 40 years since I service one of those regulators, but that was one of my favorite all times regulators. And I sold and service a number of them. It is almost embarrassing that I didn't recognize it.

I still own a number of those regulators, but I have not service or use the first stages.

Actually, the second stage is still one of my favorites. I didn't particularly cared for the non-balanced first stages so I paired some of them up with two Conshelf. I adjusted the IP to about 165 PSI.

Thanks

 

[Eric Sedletzky]

I don’t know if this is true, but
It would make sense to me that modern diaphragm designs use a larger diameter and thinner diaphragm to make the pin move more thus causing more depth of flow for larger air volumes. I have read that pistons typically flow more air than diaphragms simply because there isn’t reslly a limit to how far the piston can slide to open up the orifice. The limiting factor would be spring movement on a piston and not the stiffness of the diaphragm on the other regs that would limit the movement of the HP seat.
I noticed older diaphragm designs like the Conshelf have smaller but thicker diaphragms which I’m assuming would restrict air flow more under huge demands, but hold up better and are more bullet proof.
I have several Connies and love them for beach diving.
I used to have a Hog but sold it after maybe a dozen dives. I didn’t really care for it. I bought a service kit with it (which I never used), and remember noticing how thin the diaphragm was compared to my Conshelfs.
I figured they did this to make it more high performance and to compete with the flow of a piston.

 

[halocline]

In an unsealed diaphragm, the water pressure also presses against the diaphragm and "adds" spring pressure so the regulator stays balanced.

No, the water pressure on the outside presses on the diaphragm to compensate for depth changes. It's got nothing to do with balanced vs unbalanced. Diaphragm regulators are balanced by placing the poppet (HP seat) stem in a chamber sealed by an o-ring so that the upstream tank pressure is a net zero (or close to) on the HP seat, resulting in stable IP throughout the entire range of tank pressure.

tank pressure pushes against the HP seat and causes the regulator to "seal".

Not in a balanced diaphragm. There is a second, smaller spring that pushes the seat against the orifice.