The common denominator with the working lower was the MGI Buffer. New springs and yet another round of cleaning the carrier and fresh lube did nothing for the bursting.

Put the upper on my RRA with the MGI, no burst. Insert the MGI in the new Stag lower, no burst.

300 rounds later, no more problems.

I was running 'H' buffers in the Cav. Mk. II. and the Stag lower. I guess it just needed more weight to slow the cycle rate.

Also notice that the gas deflector is already chewed up from brass. This never happened with my last 9mm at all. Pic of the deflector.








Yes, I've read the FAQ and other threads with this problem but I think in this case, it's an upper problem.

9mm Olympic upper. Fresh and clean.

Lowers used were:

Cav. Arms Mk. II
RRA with 6 position and an MGI buffer
Brand new Stag complete lower

With the Cav, after about 75 rounds it would start to burst. Since I bought this lower used, I thought the LPK may be worn. So I stopped and put the upper on my RRA lower.

RRA lower and the 9mm upper run fine. Also using reduced power trigger springs in this one.

Brand new Stag lower...
First trigger pull, bang, second pull created a nice brass rainbow. Ammo being used was WWB.

Had the range officer there in about :30, and local PD just happened to be there running drills.

First thought is slam-fire from a stuck firing pin. Pulled the bolt, firing pin is not stuck, and firing pin return spring is strong. Carrier moves freely.

All lowers pass the disconnect test and don't allow the hammer to fall while holding the trigger back and hand cycling the bolt. All lowers work fine with my 5.56 uppers.

I'm stumped, and open to suggestions before I take it to a smith.

tvone

"Also, the one test you didn't try and would have shown the problem is to lock the bolt back, hold the trigger down, and it the bolt release to let it slam home. Chances are, it will fail this test, and retiming the disco for greater engagement will solve the problem."

Thanks for this test. Tried it on all three lowers and all passed. Last night I put a set of reduced power FCG springs in the Stag and will test it today. I also pulled the MGI buffer from the RRA to see if a heavier buffer will solve the problem.

Maybe not going back far enough to catch the disconnector but far enough to eject,load, and slam fire? Never had a 9MM and do not now how far it needs to operate.

Retime the disco until the hammer rear hook just touches/every so slightly rubs the disco hook with the trigger at reast (read .002 gap).  

Best guess is that the hammer is being danced off the disco when hot, and this is causing the problem.

Also, the one test you didn't try and would have shown the problem is to lock the bolt back, hold the trigger down, and it the bolt release to let it slam home.  Chances are, it will fail this test, and retiming the disco for greater engagement will solve the problem.

I agree. I may have fixed the problem, but not really the source.  As stated earlier though, each lower works fine with 5.56 uppers. All lowers passed the various tests.  I installed a different  hammer, disco and new springs in the new Stag lower, but still had the problem with the 'H' buffer.

My next test will be to take the 'H' buffer, and install it in the RRA lower that I know to work with the MGI buffer. If I have the same problem, then at least I know what I need to fix it.

I'm leaning towards high carrier speed, not allowing enough time to stop the hammer. Would that also explain the damage to the gas deflector? The violent ejection of the brass is chewing up the plastic.

Ok time to stop tap dancing around the problem.  

First off, 223 uses a locked bolt that is gas action opened.  
The 9mm is direct blow back action.  

This means that the only thing holding the bolt closed on the 9mm is the angle of the hammer against the non ramped bolt, and tension of the recoil spring, and the mass of the buffer and bolt.  If you need to adjust/slow the action down, then additional mass is needed.  If you try to use heavier recoil springs, then you slow the reward cycle, and speed up the return stroke (read if the hammer wasn't being danced off the disco before, it sure in hell will now).


Now the point that I and Tweak are pointing out is that the only thing holding the hammer back during re-cock is the disco.  Tweak points out to make sure that the tail of the hammer is not making contact with the back of the disco and causing the disco to be bounced open (with a slight dwell time) as the hammer moves forward for engaugement (this would cause more of a hammer follow threw).  My thoughts are to re-time the disco to the max retension to insure that even during the worst case of the bolt slaming home, the disco is going to retain the hammer.

Again, do not think 223 and 9mm as being even close to the same kind of actions; with what is working for the 223 surely should work with the 9mm uppers.  In regards to installing the MGI buffer, you inceased the mass, which means that the impact of the bolt on closer lighter, and the disco is less likey to allow the hammer to be dance out due to the impact. My guess, drop on a can with even the MGI buffer, and your going to be back at burst city.

Sorry not to nit pick, but your ejection port door spring looks to be installed backwards.