CMMG Guard Ejection Issues

I have noticed that if you try to slow down the cyclic rate on the 9mm CMMG Guard, the faster it kills ejector spring.***

*** UPDATE ***
I had initially thought it was just people like me trying to slow the cyclic rate down running into these issues but I’m hearing more reports of others having ejector spring failures not running my preferred buffer/tube/spring arrangement.  Most people seem to be having failures around 1K rounds as well.

First of all, I have 2 CMMG Guard 8″ Barrel and BCG combos.

The first set I got came with a dual feed M4 barrel extension and 1/2-36″ threads.

The second set I got came with a single feed ramp barrel extension and 1/2-28″ threads.

I had done all my tinkering trying to slow the full auto cyclic rate down and started running into failures to eject…about to give up, I tried the BCG from my second order and it ran everything fine.  I later determined that the ejector spring was different as well as exctractor design was also different on the second BCG.

CMMG told me that the .45 Guard uses a proprietary ejector spring that is shorter but stronger than a standard 556 ejector spring which the 9mm Guards initially shipped with and they later changed the 9mm Guards to use the same spring as the 45 Guard.

Pictured below is a Sprinco / 556 Ejector spring which is of typical length as others in the industry.EjectorSpring-Sprinco

Pictured below is the typical wire thickness of a 556 ejector spring.WC-thickness

Pictured below is the Guard 45, and LATER 9mm ejector spring. (I also believe the .40SW is using the same spring as well).GuardEjectorSpring-New9mm-45_10-2018

Below is the thickness of the Guard ejector spring.Guard-thickness

Comparison of some ejector springs.comparison

Putting the bolt in a vise and using a punch to compress it, I then used a trigger pull gauge to figure out approximately how much force was required for each ejector spring.
Below is a 556 Wilson Combat ejector spring.NewWC

Below is a new Sharps 556 BCG

New Guard Ejector springNewGuard

This is a worn out 556 ejector spring where I started to get malfunctions.Worn556

Worn out Guard ejector spring where I started having malfunctions as well

Note that in both cases where I started having malfunctions was at a slow cyclic rate.  These springs would ‘probably’ still work if it was cycling faster.

Pictured below is a 5.45×39 bolt modified for the Guard with my own angles cut on the bolt to further delay the recoil which killed the ejector springs in less than 100 rounds.
Bolt was modified further to accept a 308 ejector spring.

308 Ejector on the left, 556 in the middle and makeshift ejector for the Guard on the right.

View of hogged out ejector hole

View of makeshift ejector installed in bolt

This worked but only for like 30 rounds before the 308 ejector spring was completely dead.
I gave up on the ejector in the bolt at this point and went to a fixed ejector done by another C2 manufacturer that has a prototype 9mm piston upper that uses a fixed ejector.
This has turned out wonderfully.

Note that I now see that CMMG is releasing a .40SW Guard and they have updated the ejector to be a bit different from what we normally see:

What may also be different is that lugs may have a less aggressive angle as the 9mm Guard like what I was doing on my 5.45 bolts to slow the unlocking.

I have UZI bolts that have the bolt face opened up to work with .40SW and 9mm.  I’m hoping this new .40SW bolt may actually work for 9mm as well for those that want to slow the RoF and not have ejection issues doing so.

I just got my .40SW BCG and Barrel and the lugs clearly have less of an angle on the .40SW vs 9mm. 40SW-vs-9mm

Other than now having a black finish, the ejector is now rounded like seen in the pictures from the pre-release version above.
As a side note, I used to have a .45 ACP gas operated upper by RMW.  He cut the face of the ejector to have an angle to help kick the brass out at the correct angle.  I think CMMG is basically doing the same thing here but by making it rounded, they don’t have to deal with ‘indexing’ of the pin.  The following link also discusses a similar ejector mod for cases larger than 556:


The length appears to be the same as a 556 ejector.Ejector

The spring appears to be same as what is used in the 9mm and 45 ACP Guard…shorter but stiffer than a standard 556 ejector spring.40SW-ejector-spring


Update 6/22/19

Put some more rounds through the .40SW bolt and measured the current strength of the ejector spring.  Almost 6 lbs.  In my experience, once you get under 5 lbs is where you will start having issues with the 9mm when using my preferred buffer / tube / spring configuration that runs in the 600′s.

However, I don’t know what effect the rounded ejector will have….the spring is already holding up better than anything else I have tested.  My modified 5.45 bolt with angles similar to the .40 SW killed the ejector spring in less than 100 rounds.  While it was a standard 5.56 ejector spring, I would still anticipate a factory Guard spring to die relatively quickly.

The rounded ejector may also continue to work properly once the spring gets under 5 lbs as well.  I won’t know till it has gotten to that point. Ejector-6-22-19

While not on the CMMG website (at the time of this writing), Optics Planet is selling a spare parts kit:

They list two kits at that link.  One specifically states it is for .45ACP while caliber is not specified on the other one but by looking at the part number it appears to be for 9mm.

Looking closely at the picture, you can see it includes the new rounded ejector.
mkg45-repair-kitI contacted CMMG about this and they confirmed that the rounded ejector is only being included for .40SW and .45ACP (which were previously standard 556 ejectors).  I also inquired if the rounded ejectors have shown to increase the ejector spring life and was just told that it ‘worked better’ for those two calibers.  Kinda vague and decided not to keep asking the same question over email.


Update 8/3/19 
40SW ejector spring is starting to fail now. Note that previously didn’t have any FTE’s in 40SW and now it is occurring in this caliber as well.  Indicating ejector spring is failing now.Ejector-8-3-19_2


Update 8/24/19 

In the ‘Guard’ thread someone mentioned trying an HK 416 ejector spring.  Never thought of that.


A lot longer than a CMMG or standard 556 ejector spring


I did do some searching and found this thread:

In that thread, someone stated the HK416 ejector spring is weaker than a standard AR ejector spring and he was right.HK416-EjectorSpringStrength40SWBolt


Although it is weaker, at 6lbs it should still function fine for .40SW.  I am interested in longevity.  If it stays at 6lbs for at least 2K rounds then it is an improvement.

Tested it today and ran (4) 30 round mags in full auto suppressed in 40SW and had zero malfunctions.

However, when I tried it putting the BCG in my 9mm upper I couldn’t get past one round without a failure to eject.

A couple things still to try.
1.  Maybe use a standard flat faced 556 ejector in the .40SW bolt since the rounded ejector is really for the larger cases.
2.  Some others have mentioned to put a rod inside the ejector spring to reduce the spring from binding.   I have some drill rod laying around and plan on trying that.

However it was a major pain to install that HK spring and ejector in the bolt and don’t want to mess with it till I get a bolt ejector tool to make it easier to swap the ejector and spring.


Update 8/25/19 

I was thinking about having an ejector made like the DPMS GII ejector seen below. You can see in my pictures how the springs get deformed and almost look like they are binding.
BTW, I’ve been emailing CMMG support and also pointed them to the thread on, this  website, this idea and that several people are having ejector failures around the 1k mark.


I was thinking it would be a pain to make an ejector like this and that this would be a failure point being that the 556 ejector is smaller than .308 as well.

Talking to a few guys over email, I had the idea to use a nail as an insert to the spring. One of the guys got these nails from Home Depot. These are stainless steel and seem pretty tough.
So my next attempt will be like pictured below. I had to sand down a little right after the tip as nails are kinda deformed there. After that it slides right into the spring nicely.
#18 is supposed to be 1.207mm. A 1.25mm drill bit fits nicely inside the ejector spring so this is about perfect.
The head of the nail was a little big but no big deal to stick it in a drill and sand down so it moves nicely into the ejector hole now.
I need to measure the length, I may not have to even cut the nail or I could even cut the nail and put another on the other side.
This should work out better than a custom longer ejector with a spring insert since it can’t snap off.



Update 8/30/19 

Measured the strength of the HK spring and it lost some strength after the 120 rounds.

Took it apart and it shrunk quite a bit.HK416-EjectorSpringStrengthPost120rndsComp

I tried to put the nail in the HK spring but that spring has a smaller ID than the CMMG spring so going to shelve the HK springs for a while and test a new CMMG spring with the nail.

Approximate nail length.  I think it can be a little longer.

Measured force w/ nail in the .40SW bolt with zero rounds fired.  Also confirmed that the ejector compresses fine. It actually still goes a little beyond the bolt face so I could have made the nail slightly longer to limit the distance the ejector can travel.


Update 8/31/19 

Ran 461 rounds of .40SW
(2) 30 round mags of 180 Gr suppressed (6.0Gr Autocomp) suppressed – 2 FTF (Failure to feed)*
(2) 30 round mags of 155 Gr unsuppressed (5.6 Gr WST) unsuppressed – no malfunctions
(2) 30 round mags of 180 Gr suppressed (6.0Gr Autocomp) suppressed – 1 FTF (Failure to feed)*
(2) 30 round mags of 155 Gr unsuppressed (5.6 Gr WST) unsuppressed – no malfunctions
(1) 30 round mags of 180 Gr unsuppressed (6.0Gr Autocomp) suppressed – no malfunctions
11 rounds of remaining 155 Gr (5.6 Gr WST) unsuppressed – no malfunctions
(2) 30 round mags of 180 Gr unsuppressed (4.9Gr WST) suppressed – no malfunctions
(2) 30 round mags of 180 Gr unsuppressed (4.9Gr WST) unsuppressed – no malfunctions

*If you look at the .40SW page, you’ll see that FTF’s are the malfunctions I have been running into with the .40SW.  I don’t think it has anything to do with the ejector.  It is interesting that I had zero malfunctions with the loads using WST powder.

Switched to 9mm using new Curtis Tactical custom suppressor and had a few FTE (Failure to Eject) malfunctions.  Switched to Octane 9 which worked before and had two FTE’s again after the first few rounds and stopped trying to shoot 9mm with the 40SW bolt.

Ran two more mags of .40SW to make sure it still worked fine.
(2) 30 round mags of 180 Gr suppressed (4.9Gr WST) suppressed – no malfunctions

Strange that I couldn’t get 9mm to run with the .40SW like I did before when the bolt was new.  Maybe it is because I didn’t run as much .40SW with the new spring the first time and this time I ran 401 rounds before trying 9mm.

The spring lost over a 1 lb of strength to run 461 rounds of .40SW and the few rounds of 9mm I tried.

The ejector spring shrunk a bit.

The nail still looks straight.  The spring still looks like it got deformed even with the nail.

I am going to try putting a standard 5.56 ejector in the .40SW bolt to run 9mm and see if that helps at all.  Beyond that, I’m about to give up on the spring loaded ejector.

Update 9/7/19 

Another (4) 30 round mags of .40SW.  Round count at 581 rounds.

1st 30 round mag had a single failure to feed.  The other (3) 30 round mags were malfunction free.

This was using a standard 5.56 ejector.  Also tried to run 9mm a couple times and immediately had FTE’s.