Question How much of an IP drop is acceptable?

[Learner Diver]

Sacrificing the turret for manufacturing simplicity and lower cost, it was designed in part by a former motorcycle engine designer; hence the odd (abortive) external design that never really took off.

I have read the design was influenced by Ducati.

Unless this is a new phenomenon in a reg that has shown better performance before, I'd pick the the spring.

Read your in depth analysis. Would IP creep be the same for a piston regulator like M25?

 

[rsingler]

Would IP creep be the same for a piston regulator like M25?

True piston IP creep (rising IP that, slow or fast, does not stop) can come from  only one place: the seat/knife edge interface. Not an o-ring or a scratched land. Only the seat or the knife edge.
Other leaks have multiple causes, including orings. What was termed IP creep in the OP (a  drop in IP) was perhaps due to a piston shaft leak which depleted supply pressure, and is not what I think of when using the term.
Diaphragms can have a multitude of sources of true IP creep due to their different architecture. That's another discussion. As were my comments above - I was not attempting to diagnose piston IP creep, which is straightforward, but rather excessive dynamic IP swing, which is not. Hope that helps.

 

[happy-diver]

 

[Jens Bölte]

The Mk21 was designed by Scubapro as their "value" Mk25.
View attachment 844861
Sacrificing the turret for manufacturing simplicity and lower cost, it was designed in part by a former motorcycle engine designer; hence the odd (abortive) external design that never really took off.
But in almost every internal part, it's a Mk25. The seat, springs and piston shaft seal are all Mk25. The piston shaft ID is identical, though it's a shorter part.

The posters above have hit all the key points. The IP drop MUST be related to one of four things: 1) knife edge/seat flow restriction, 2) friction, 3) input pressure/flow restriction or 4) measurement error.

I couldn't quite discern from the description of the test setup, but tapping the IP from a bcd hose is reasonable. The first thing I'd do is remove the (new) sintered filter and throw the reg set on a tank with the largest visible gas outlet in the valve, making sure it's full and wide open. That will eliminate 3). Eyeballing an analog IP gauge on a bcd hose is good enough for 4).
The only potential weakness in the Mk21 is its shorter piston. Much like the transition from the Mk5 to the Mk10, ensuring proper piston alignment was part of improving flow. The Mk21, like the Mk10 & 25, has its piston head land inside the reg body (instead of the screw-on turret). But it is so much shorter that (theoretically), there could be slightly poorer axial tracking, and internal friction could play a role.

If the knife edge is perfect and the seat is new, then gas flow should be nominal at the interface if IP is at specification  unless for some reason your spring constant is much higher, and valve opening is restricted at a given IP. So to eliminate 1), I'd replace the mainspring.

Finally, internal friction should be revealed by using a cheap USB microscope to inspect the piston shaft, the OD of the piston head and the piston head land in the body. Any scuffing would implicate 2).

Unless this is a new phenomenon in a reg that has shown better performance before, I'd pick the the spring.

If it truly is a new change and there's no scuffing inside, I'd pick the tank valve flow or the sintered filter. It's just logical that if the IP flow can't keep up with the wide open second stage valve, there must be an obstruction. A clogged or poorly manufactured replaceable filter would seem to be the statistically likely culprit.

A fascinating case!

Hello

Thanks for all the suggestions. I think I can rule out measurement error and input flow restriction. The test bench is used for all regulators being serviced in this shop. The valve was fully open and the test bench reported reasonable values for all other regulators that day. So I doubt that there is a mismeasurement. I suspect the spring is the problem, as it got a bit better when I added a plastic shim for testing purposes, which increased the spring force. Unfortunately I do not have a new spring to test. I will have to order a replacement.

Best wishes Jens

 

[Nemrod]

Hello

Thanks for all the suggestions. I think I can rule out measurement error and input flow restriction. The test bench is used for all regulators being serviced in this shop. The valve was fully open and the test bench reported reasonable values for all other regulators that day. So I doubt that there is a mismeasurement. I suspect the spring is the problem, as it got a bit better when I added a plastic shim for testing purposes, which increased the spring force. Unfortunately I do not have a new spring to test. I will have to order a replacement.

Best wishes Jens

This reminds me of something. I once had a Mark V that had a large IP drop inconsistently. I fussed over it for days. It turned out that the spring, when compressed, had a bow in it and was binding or dragging on the housing. It is also reasonable to think that such a spring could provide an asymmetric force to the piston causing it to want to cock to one side or other and thus cause a bind. Perhaps the Mark 21 shaft being somewhat on the shorter side could exacerbate this potential.

 

[halocline]

Hello

I have a MK21 which, when set to an IP of 9.5 bar (137 PSI), shows quite a large drop in IP on inhalation. The IP drops by about 2.5 bar (36 PSI) which I think is quite a lot. I replaced the o-rings on the piston as it was also showing IP creep from 9.5 bar to 8.5 bar (123 PSI) when pressurised over a period of time. The IP remained constant after the o-rings were changed. The drop in IP was slightly less than before, 2 bar (29 PSI) instead of 2.5 bar (36 PSI). Does anyone have any suggestions on what to do about this rather, IMHO, large drop in IP when inhaling? What is an acceptable drop in IP during inhalation?

Thanks a lot Jens

The problem is likely in large part how you’re measuring IP drop during inhalation or purge, and there’s nothing you can do about that without using much more accurate measuring equipment than a IP gauge at the end of your inflator hose.

I went through this same issue years ago trying to figure out why my MK10 and MK15, both of which ar e high flow regulators, dropped more under inhalation than did my MK2. I even had a lengthy conversation with Peter Wolfinger about it. My wife is calling me now to walk the dog, so I’ll post this and then when we get back I’ll explain.

 

[halocline]

Ok, back from the walk/sniff fest. Here’s the basic issue. An IP gauge on the inflator hose will not read accurately while air is moving through the regulator because that movement creates a Venturi effect which effectively lowers pressure in the inflator hose. The faster the air is moving past the port to which the inflator hose is connected, the greater the effect. Even if you were to put an IP gauge on a tee right where the hose meets the 2nd stage, it would do the same thing, because we’re still talking about air moving linearly across the opening to the gauge.

There’s another thing; and this one is a guess. IP gauges are not designed to accurately register dynamic changes in pressure; they have springs and mechanical action which moves the needle in response to pressure. The needle itself has some mass, meaning as it’s moving it’s also acting on the spring. I don’t know how much this would really effect things, but I do know you’re trying to use the gauge to measure something it was not designed to do.

What you’d need is a flowmeter that is specifically designed to measure flow amount and pressure within a chamber that has moving air.

It’s possible that you have another issue with regards to the static IP ‘reverse creep’ over time while sitting there pressurized. It’s not uncommon for a pressurized regulator to have the IP measurement drop slowly over a long time as it sits unused. But a full bar, 15PSI, is a lot. My guess is that there is a very slight leak somewhere, I mean very slight, which is allowing a small amount of pressure to leak out of the IP chamber, and eventually it would drop enough to open the 1st stage valve and then re-pressurize and shut. There could even be a leak in the internal o-ring in the piston itself, and the fix for that is a new piston.

There’s the possibility that the gauge is leaking a small amount or even just is not accurate enough mechanically to stay exactly where it is over a period of time. Or, maybe the seat is breaking in and the seating force drops a little as the piston/seat mating surfaces fit more precisely.

These devices, both the regulator and the gauge, are not that precise. The round tipped MK25 piston generally locks up very well, but there are of course variances in the machined surface, in the seat surface, in the spring tension, and basically they’re designed to work for sport diving, not so much for lab-grade accuracy. So I would suggest taking it for a deep dive, breathing heavily for a few breaths, if you’re really obsessed you can attach a couple of alternate 2nd stages, bring along a couple of equally-minded regulator nerds (I’ve been there, done that!) and have them also take a few hard breaths at the same time, so you are really stressing out the 1st stage. If it works well, it works well.

 

[Nemrod]

Thing is this is said to be shop equipment (Post #3) and he has run other regulators back to back along with the discrepant one and it is the only one with the large drop thus ruling out equipment error. His readings may not be accurate but the IP on the problem child should not be drastically different in error vs the other units he has run? I do not like measuring dynamic IP from the end of an inflator hose, but it should be reasonably consistent in error .

 

[tursiops]

Ok, back from the walk/sniff fest. Here’s the basic issue. An IP gauge on the inflator hose will not read accurately while air is moving through the regulator because that movement creates a Venturi effect which effectively lowers pressure in the inflator hose. The faster the air is moving past the port to which the inflator hose is connected, the greater the effect. Even if you were to put an IP gauge on a tee right where the hose meets the 2nd stage, it would do the same thing, because we’re still talking about air moving linearly across the opening to the gauge.

There’s another thing; and this one is a guess. IP gauges are not designed to accurately register dynamic changes in pressure; they have springs and mechanical action which moves the needle in response to pressure. The needle itself has some mass, meaning as it’s moving it’s also acting on the spring. I don’t know how much this would really effect things, but I do know you’re trying to use the gauge to measure something it was not designed to do.

What you’d need is a flowmeter that is specifically designed to measure flow amount and pressure within a chamber that has moving air.

It’s possible that you have another issue with regards to the static IP ‘reverse creep’ over time while sitting there pressurized. It’s not uncommon for a pressurized regulator to have the IP measurement drop slowly over a long time as it sits unused. But a full bar, 15PSI, is a lot. My guess is that there is a very slight leak somewhere, I mean very slight, which is allowing a small amount of pressure to leak out of the IP chamber, and eventually it would drop enough to open the 1st stage valve and then re-pressurize and shut. There could even be a leak in the internal o-ring in the piston itself, and the fix for that is a new piston.

There’s the possibility that the gauge is leaking a small amount or even just is not accurate enough mechanically to stay exactly where it is over a period of time. Or, maybe the seat is breaking in and the seating force drops a little as the piston/seat mating surfaces fit more precisely.

These devices, both the regulator and the gauge, are not that precise. The round tipped MK25 piston generally locks up very well, but there are of course variances in the machined surface, in the seat surface, in the spring tension, and basically they’re designed to work for sport diving, not so much for lab-grade accuracy. So I would suggest taking it for a deep dive, breathing heavily for a few breaths, if you’re really obsessed you can attach a couple of alternate 2nd stages, bring along a couple of equally-minded regulator nerds (I’ve been there, done that!) and have them also take a few hard breaths at the same time, so you are really stressing out the 1st stage. If it works well, it works well.

The point of measuring an IP drop is some worry about breathing from a regulator that sees that drop. But is not the regulator ALSO attached to a port, so experiences the same venturi effect? I don't buy the venturi argument as being either large enough , or imporatnt enough to consider.

 

[halocline]

The point of measuring an IP drop is some worry about breathing from a regulator that sees that drop. But is not the regulator ALSO attached to a port, so experiences the same venturi effect? I don't buy the venturi argument as being either large enough , or imporatnt enough to consider.

I'm not sure you're understanding the mechanics of what I'm describing, or maybe I'm not doing a good enough job describing things. When you inhale on a 2nd stage, you lower the pressure in the 2nd stage, which collapses the diaphragm, which pushes the lever, opens the 2nd stage valve, which then allows air at IP in the hose to expand into the 2nd stage, which lowers pressure in all the LP hoses and IP chamber of the 1st stage. Forgive me if this is all obvious to you, I'm just trying to describe it fully.

Ok, so then the lower pressure in the IP chamber allows the 1st stage valve to open, letting HP air from the tank into the IP chamber where it flows through the hose to the 2nd stage and then into your lungs. It's this flow that causes a localized pressure drop in other hoses or anything that presents an environment where air is moving quickly and linearly across an opening. Again, sorry if this is obvious to you. But when you say: "is not the regulator ALSO attached to a port, so experiences the same venturi effect?" it sounds like you don't understand my argument. Air is flowing THROUGH the hose to the 2nd stage, not ACROSS the port to that hose. Venturi force is caused by air moving quickly across (or perpendicular to, sort of) an opening. So of course the air moving through the hose to the 2nd stage is not subject to this specific venturi force, it's that air which is causing it in other hoses.

There's more at work here than meets the eye, and I don't claim to have an engineer's understanding of the fluid and aero dynamics of the whole system. But I'm very confident that what I'm describing is accurate, and as I mentioned I did have a pretty lengthy conversation with Wolfinger and he confirmed it.

The information is Vance Harlow's book is really great, but in this case, it's not complete.