Jetstream 3960 high pressure leak

[kombiguy]

Not just a leak, but a full-blown out of control IP. Turn the tank on, the ip rockets to 200+ psi and hits the stop on my gauge. Cleaned and rebuilt twice, and I have no idea what the problem could be. I have the manual, the assembly tools, and have rebuilt 3960s before. Apparently, what I lack is some brainpower. Any hints?

 

[Caveeagle]

Take a close look at the little steel cone that seats into the plastic HP seat, That is often the issue and can be either polished or replaced.

How old was the service kit you used? Look for a little number cast Into the plastic seat, there were a few that were recalled a coupe years back. If I can find my notes, I will post back with the the bad #

 

[rsingler]

#4

 

[rsingler]

This can be polished out with Micromesh.
But it seldom causes a "rocket" to 200 psi. It's usually gradual.

Here's the only time I ever had a newly serviced 3960 rocket. It was a manufacturing flaw in the HP seat oring. It split along the mold line and then blew out.

 

[rsingler]

Get a $25 USB microscope that attaches to your phone or laptop from eBay or Amazon. You can get a 30x view of those parts and likely find the culprit.

 

[MichaelMc]

Get a $25 USB microscope that attaches to your phone or laptop from eBay or Amazon. You can get a 30x view of those parts and likely find the culprit.

Are there any USB microscopes you particularly like?

A New York based electronics distributer/manufacturer, adafruit, whose stuff I find to be high quality, has this 20-220x one for $75. USB Microscope - 5MP interpolated 220x magnification / 8 LEDs. I have no relation to them, beyond buying many of their small boards and other fun bits.

 

[rsingler]

It's always a cr*p shoot, given our Chinese friends.
Here's one for $500

and one that looks similar for $18:

That fat style seems to have become more popular, because it moves the LED lights out to the side a bit more.
But I usually use it with the lights off, and light the part separately from the side.
This was my favorite:

But given where the scopes are made, there's no guarantee you'll get a great image. The only good thing is that it didn't cost much if you got screwed. I've got several; some on long cords for looking in tanks. I've had three good, one mediocre.

If your friend has quality stuff, that's worth something, I'd say.

 

[rsingler]

There are only three places that can cause IP creep in a 3960 (green arrows):

1) the HP seat oring (that split in the post above)
2) the interface between the piston valve and the plastic seat (the cone picture above)
3) The balance chamber O-ring.

It HAS To be one of those three spots. The seat can be replaced. The cone can be polished. The o-rings can be replaced. But if the land where the balance chamber o-ring sits has been scratched, or the land where the high pressure seat oring sits has been scratched, the regulator is toast.

 

[rsingler]

It's why the Poseidon XStream is the best diaphragm regulator in the world, IMO.

It reduced the number of failure points from 3 to 2, since it's now the HP seat that moves with falling tank pressure! There's no balance chamber needed.

And with the stainless steel ball now replaced by a synthetic ruby, it never scratches from grit in the HP tank airflow, like the sandblasted cone in the pic above.

 

[kombiguy]

Well, I replaced the seat and its O-ring. That reduced the large ip increase. It now is just a slow leak. I've got a new piston coming, and I'll start checking the lands as suggested by rsingler above.