9mm CMMG Guard Full Auto Tuning

Goals for this platform are:

  1. Soft shooting
  2. Flat shooting – Minimize ‘dot’ bounce so gun stays on target. some equate this as being soft shooting but I think they are different but related.
  3. Slow enough cyclic rate to easily pull single round bursts.  Here is where the AR platform really shines over the MP5, MPX and many other platforms.  This was achieved by selecting from a large assortment of buffer tubes, buffers and springs on the market today.  These options overshadow every other platform in existence and this doesn’t appear to be slowing down.
  4. Good suppression platform.  Like the MP5, the CMMG Guard utilizes a delayed blowback mechanism which suppresses very well.  Better than any blowback 9mm Colt pattern AR.  As I side note I also have a post sample Sig MPX and IMHO the MP5 and CMMG Guard suppresses better.  Too much noise comes out of the piston with the MPX.

Documentation Breakdown – I’ve broken it down to 3 sections based on the amount tuning involved.

I.     No custom parts – just specific, component selection
II.    Custom tungsten carrier weight that is spring loaded
III.   Fixed Ejector and modified bolt lugs to further delay unlocking
IV.    .40SW Guard bolt used in 9mm upper

I.     No custom parts – just specific, component selection

The recommended configuration from CMMG is to use a standard carbine spring and standard carbine buffer.  Due to the radially delayed blowback this will not be hard on your lower like all the blowback 9mm AR/M16 configurations out there.  If that is all you are after, then go with that.

While it does work for full auto, I find it to be too fast….above 1000RPM.
I also find it to be harsher and more bouncy than an MP5.

My recommended configuration for those that don’t want to do any tinkering with custom parts is to go with:
1.  Tubb 556 Flat Spring
2.  A5 buffer tube or low cost equivalent
3.  9mm Kynshot/Blitzkrieg Hydraulic buffer – Think of this as being a ‘hydraulic A5 buffer’, don’t get hung up on it being a ’9mm’ buffer.  

Bear in mind that the above configuration works VERY well in many other configurations not just the 9mm Guard…. I’ve tried to document some of that here.

Back to discussing the Guard, below is what I get unsuppressed with this configuration.  622 RPM….NO carrier weightGuard-556Tubb-9mmKynshot-A5

686 RPM suppressed. Again, there is NO carrier weight.NoWeightMK9K-686RPM

Below is some of the trial and error I did to get to these component selections.

I was previously using the JP captured setups in several configurations (9mm straight blowback and 5.56).  I have the spring tuning kits, standard version in Gen1 and Gen2, Gen2 H2 version for full auto as well as the 9mm version (straight blowback).  Initially tried the H2 version and it was too heavy and was getting malfunctions.  Tried the standard weight and was getting light primer hits and cycled way too fast.   I decided to ditch the JP system since that system wouldn’t allow me to add weight to the carrier.

Did some preliminary testing on smoothness with the Tubbs flat wire comparing to the Sprinco Green and no-name AR rifle spring and think the Tubbs flat wire was the smoothest.  After that did all testing with the Tubb’s flat wire.

Testing below was with custom 5.1 oz tungsten carrier weight.
H2 RoF too fast and couldn’t get singles
H3 As seen on timer only got 7 rounds tried another time and got to around 8 rounds before malfunction again.  Was also bouncy
The Enidine Hydraulic was noticeably stiffer to compress than the 9mm Blitzkrieg/Kynshot
Blitzkrieg/Kynshot seemed to be the smoothest but did encounter some malfunctions which I think was due to just too much weight.
AAC tungsten was kind of bouncy and didnt’ lock the bolt back.
MGI was also bouncy and didn’t lock the bolt back.
The Griffin Armament AR SOB was surprisingly quite good and I’d say the best choice behind the hydraulics….Enidine and Blitzkrieg/Kynshot.
Did a little bit of testing with no weight and it was just too fast although quite smooth.
Finally switched to the 3.5 oz Colt Steel carrier weight.
Didn’t encounter anymore malfunctions with the Blitzkrieg/Kynshot, singles were quite easy to get.
Enidine was noticeably more bouncy than the Blitzkrieg/Kynshot.  Wonder how an carbine Kynshot would do compared to the Enidine since the Blitzkrieg/Kynshot is not as stiff.
** Update – ended up getting a Kynshot carbine buffer and the stiffness seems identical to the old Enidine.  Interesting that they both feel stiffer than the 9mm ‘Blitzkrieg’ version. **
Documented the different hydraulics here
Again, Griffin Armament AR SOB was pretty good and would probably use that if was not able to use the Blitzkrieg/Kynshot.
All pictured RoF results below were with a standard steel Colt 9mm bolt weight (3.5 oz) except for the AAC 7.4 oz buffer was done with 5.1 oz tungsten carrier weight.

Enidine Hydraulic buffer – 4.3 ozEnidine-A5-Tubbs

H3 buffer 5.3 oz

Slowest ROF AAC 7.4oz buffer w/ 5.1 Tungsten carrier weight – bouncy and didn’t lock bolt back.


Griffin Armament AR-SOB buffer – 4.8 oz

Blitzkrieg / Kynshot 9mm Hydraulic buffer – 5.9 oz.  3.5oz Colt weight was installed in the carrier****** – note that I found reliability problems with the weight installed and this buffer/buffer tube and spring arrangement and removed it.

A5 Buffer Length – HeavyBuffers.com

Blitzkrieg/Kynshot 9mm buffer (RB5007) – compressed

Blitzkrieg/Kynshot 9mm buffer (RB5007) – extended

Blitzkrieg/Kynshot 9mm buffer (RB5007) – weight

** Update, after more testing, I removed the weight from the carrier which has increased my RoF.  Same buffer system:
556 Tubb flat spring
9mm Blitzkrieg/Kynshot buffer
A5 length buffer tube


II.   Custom tungsten carrier weight that is spring loaded

I continued to tinker because I felt the MP5 still had the edge in terms of softness compared to the Guard.

As of Nov 2018 pictured below is where I was at on the tuning of the CMMG Guard:

572 RPM, softer, flatter and slower than shooting the configuration I was using the Section I.

  1. A5 or equivalent buffer tube
  2. 556 Tubb Flatwire buffer spring
  3. 9mm Blitzkrieg/Kynshot hydraulic buffer
  4. Insertion of the static bolt weight pin.
  5. 18-8 SS washer placed after the bolt weight pin to prevent spring distortion
  6. Spring is then inserted (pictured below is a conical compression spring, others tested as well)
  7. Custom 2.2 oz Tungsten weight is inserted in the carrier.  Depending on the spring, I made sure it was either flush with the end of the carrier or slightly protruding so it isn’t beating on the buffer.


I have tested with a 1.8 and 2.2 oz Tungsten weight.
Conical compression spring used for this confined space: https://www.mcmaster.com/1692K919
Previous spring testing was deforming the spring from the Tungsten weight compressing the spring onto the bolt pin.  Thick 18-8 SS washer is now used to prevent spring deformation: https://www.mcmaster.com/90377A144

I had tested with the CMMG Tuning kit which comes with 3 weights: 1oz, 2oz and the 3.5 oz (same as the stock Colt weight).
In my experience adding a weight that is pinned to the carrier also allows that weight to pound into your shoulder and makes it more bouncy.  Ideally, you only want to affect the initial recoil impulse.  Think about an MP5, the rollers don’t do anything when the carrier is to the rear.
By using a spring loaded weight in the carrier, the weight can further delay the blowback but on the rearward stroke has less of an effect as a static weight that is pinned to the carrier since it is free to move independent of the carrier.  Additionally, by using the hydraulic buffer, the buffer may also be absorbing energy from the little Tungsten weight as well as from the BCG.

Using a spring loaded weight in the carrier is nothing new.  I can think of two carriers in production today that use this:
The Master of Arms Nyx BCG: https://www.masterofarms.com/index.php?main_page=product_info&products_id=69

and the Surefire SOB carrier: http://www.defensereview.com/surefire-optimized-bolt-carrier-long-stroke-obc-ls-drop-in-bcg-and-h7s-buffer-system-for-better-weapon-control-reliability-and-durability-best-tactical-ar-15m4m4a1-carbineshort-barreled-rifle/

In both cases these carriers are for 556 and not compatible with the CMMG Guard carrier.  From my observations, the carrier is machined differently to accommodate the different bolt design as well as different cam pin path.

Regardless, while the attributes of having a spring on the backside of the weight makes sense in that it will slow the cyclic rate and give the magazine more time to feed, I also think that it can increase the amount of force going into your shoulder on the rearward stroke since the weight is captive to the carrier.  I have considered trying a weight on the backside of the weight in the Guard and leaving the weight NOT captive to the carrier but unless a new hole is drilled into the carrier, there just isn’t much room to mess with this.

Pictured below is the distance behind the bolt pin.  .887″ is not much room to work with but better than nothing. DistanceBehindPin

I had some left over tungsten from another project and cut off a piece.  The groove in the middle is where a hole was before.

I went to a surplus hardware store and sifted through various springs.  I selected two different strengths that had the right diameter but needed to be cut down.  One was pretty weak while the other was considerably stronger.

A note about RoF testing, it can vary quite a bit depending on ammo, springs (buffer and hammer), magazines…and how dirty/clean the gun is.  In this kind of testing, I try to do a baseline RoF test the same day immediately before testing anything as the gun is the same in regards to how clean it is.  Below is my baseline config I have been using w/ no carrier weight added yet.
A5 or equivalent buffer tube
556 Tubb flat wire spring
9mm Blitzkrieg/Kynshot buffer

622 RPMNoWeightBaseline622RPM

Next was to try the 1.8 oz Tungsten weight with the ‘weak’ cut down spring which was 630 RPM.  At 630 RPM vs 622 RPM, that was really about the same…like it made no difference.  I was kinda disappointed as I could feel no difference either.  It was like the weight/spring did nothing.

I moved on to the stronger cut down spring and this resulted in 602 RPM and I could feel the difference in smoothness and the dot stayed on target much better than before.

It is hard to quantify how controllable something is.  So while the RoF information is good to know, it has no bearing on the control ability of the gun.  As an example the static 3.5oz Colt weight results in 525 RPM but it is really bouncy and hard to keep the dot on target.Guard-556Tubb-9mmKynshot-A5-3_5ozColtWeight

Initial tests were positive but I knew the surplus cut down spring were just a proof of concept and I would need to get something better.  I found a crappy washer in my tool box that happened to have a .625″ diameter and figured I would try that to see how it would hold up.  Picture below shows how the cheap washer is being bent over the bolt pin.  This was only 30 rounds of full auto with mixed burst fire.  Looking at this, you know that the weight is clearly moving with considerable force and further delaying the blowback.

The washers I’m using now mentioned at the beginning of this article are .060″ – .090″ in thickness and made from 18-8 SS.
I have had no more spring deformation issues with this washer.

McMaster carries 3 different strength multiwave compression spring that have .30″ length and .625″ diameter.  6, 12 and 20 lb.

The 6lb was like the cut down weak standard compression spring and acted like it wasn’t even there.
I can’t find a picture of the RoF results I did for that one right now.

The 12lb worked pretty decent in regards to smoothness.  RoF not bad either.  659 RPM.  I know that is faster than the stock 622 mentioned earlier but this was also done a few weeks earlier and the gun was probably more lubed at that time.

The 20lb wave spring at 612 RPM was getting results like the strong cut down conventional spring and was very nice.  This was done the same day as the 12lb test above.20lbWave

However, not leaving well enough alone, I felt that having to resort to this thick washer didn’t allow for much movement and the wave spring doesn’t have much deflection.  So I also ordered the conical compression spring mentioned earlier.

Conical compression spring with the 1.8 oz Tungsten weight 634 RPM and very nice.

A note about the concical compression spring.  See picture below of new unused conical spring.  Circled area shows how the last coil at the bottom is directly below the coil above it.  You need to bend that out of the way or cut it or it will get pinched off.conical-top

Picture below shows that the weight was coming forward so hard that it was able to pinch off the coil below it.

I had some more Tungsten laying around so had another piece cut to .681″ which is 2.2 oz.

That 2.2oz Tungsten weight, conical spring and SS washer is the current setup and I’ll be running it hard for a while to see how well it holds up.

I am also using the new Mean Arms Endomags.  I had tested using unmodified Colt mags using a modified Hahn mag block and modified mag catch as well as the CMMG Guard Glock lower.  I’ve always hated Colt mags and have had them fail on me too many times as well as ‘volcano’ing’ if slightly bounced.  I have also found that the Colt mags require more force to strip rounds off than the Glock or new Endomags.  More force required to strip rounds means a smaller operating window of reliability.  Again, I have specific goals in mind.  If cyclic rate is not a concern then going with a different buffer/spring arrangement would be fine w/ Colt mags but not for my goals.

So for me, it came down to the Glock or Endomags.  I really like double feed mags more than single feed for ease of loading.  So I am currently using the Endomags.  I really hope they come out with a 40 round version at some point.  The Endomags have been pretty reliable although I occasionally have an issue where a round is not fed, like it couldn’t keep up.  There is no jam, just no round in the chamber.  I have 8 Endomags for the Guard and just haven’t spent the time to figure out if it is specific mags or what.

As mentioned in the goals I want, the Sig MPX is not desireable to me due to its lack of tuneability and the suppression being superior on the MP5 and Guard so I would regularly fire the MP5 for comparison.  Below is what the Guard was up against.

MP5K PDW below that is running at 907RPM.

I would also compare to a full size MP5 running at 789 RPM:

Also tested against a Post Sample Sig MPXhttps://www.youtube.com/watch?v=XKVgczZKObI
In my opinion the MPX does not suppress as well as the MP5 or CMMG Guard.  The RoF is also high and like the MP5, both are NOT tunable for cyclic rate.  Sig MPX was eliminated from my comparison testing since there is nothing I can do about its suppression performance relative to the MP5 and Guard.

***Note that all this testing was unsuppressed, running with a 9mm suppressor turned this configuration into a single shot***
As a workaround, I tested a SWR 40SW can and a .45 Gemtech Blackside can.
Interestingly, these both work fine whereas any other 9mm can I have tested makes my setup w/ the spring loaded weight a single shot. (I have tested, 9mm Liberty Mystic, 9mm SWR Trident, Octane 9 and Gemtech MK9K).

Pictured below is the RoF with the .45 Gemtech Blackside. The Pact Timer counted 581 RPM while the Competitive Edge timer picked up 577 RPM….both counted 10 rounds.
I had previously gotten 564 RPM unsuppressed with the same exact setup.

III.     Fixed Ejector and modified bolt lugs to further delay unlocking

As mentioned previously, section II testing was exclusively unsuppressed.

In my experience, adding a suppressor on a SMG only adds about 30 RPM (you typically see about 200 RPM increase for rifle caliber like on a 556 M16) or so more to the cyclic rate and if anything increases reliability but is dirtier and then it will get more fouled.

So I was thinking that putting the suppressor on would be no big deal. To my surprise, adding a suppressor with my spring loaded tungsten weight gave me a lot of problems. With the 300 BLK spring it turned it into a single shot with my MK9K.
Swapping to the 556 Tubb spring was better but still jamming like every 3 or 4 rounds. I had to remove the spring loaded tungsten weight and it is fine but after shooting more with it, I can definitely feel how much smoother it was with it and am disappointed it was’nt working with a suppressor.

With the Guard, it has at least 60RPM faster cyclic rate when suppressed which I think is higher than other SMG’s.  This is with various 9mm suppressors.  >>> Note that I did do some testing with a .40SW and .45ACP suppressors on the Guard and it was reliable with the configuration described in Section II.

I didn’t like the way the Guard felt with either of my uppers suppressed.
I have an early Guard barrel that came with the dual feed ramp extension and that one was testing with the MK9K. While I was using an SWR Trident on my other Guard barrel that has the single feed ramp.
They are both bouncy to me suppressed.

I felt like my setup in Section II unsuppressed had rivaled the MP5′s I had tested against.
However suppressed, the MP5′s kicked the Guard’s ass. Not from a sound perspective, Guard sounds fine like the MP5. It is the flatness, smoothness suppressed compared to the MP5 that the Guard was way behind.

Back around 2016, I had worked on a prototype piston operated AR/M16 9mm upper with another manufacturer.
We had used 5.45×39 bolts for testing as the bolt face was perfect for 9mm.  So I had a couple 5.45 bolts modified so the tail end matches the profile of the Guard bolt so it could fit in the proprietary Guard carrier and accept the bolt spring that the Guard uses.  Bolt on the right is a standard 5.56 for comparison.


I then cut my own angles on the 5.45 bolt to further delay the unlocking for suppressed usage.  This yielded 504 RPM on my first bolt and was incredibly soft and smooth.545-mod-504RPM-MK9k

However, I found that it was killing the life of the ejector springs very quickly.  I would say in less then 100 rounds a new ejector spring would die.  This was obviously not going to be acceptable.

As a side note, there have been many discussions about the CMMG Guard ejector springs.  CMMG uses a proprietary ejector spring in the .45ACP Guard which is shorter but stronger than a standard 556 ejector spring.  They are using thicker coils.
Later, CMMG standardized using this proprietary ejector spring in their 9mm Guard as well.

Ironically, when I had worked on the prototype 9mm piston upper with the other C2 we had stopped messing with it due to erratic ejection relying on the ejector in the bolt.  We had concluded that the system needed a fixed ejector to be reliable and we didn’t want to mess with it at the time.  Fortunately, he continued to work on it and has a proprietary ejector, bolt and carrier made up for his piston system.  (not yet released).

He retrofitted my Guard upper with his fixed ejector setup and I am happy to report this has brought the Guard to a point that I think it exceeds the MP5 unsuppressed AND suppressed.

However, I am not at liberty to discuss all the details of his fixed ejector setup at this time.

Fixed Ejector RoF: 546 and 552 depending on which Timer you look at.FixedEjectorUnSuppressed

610 RPM suppressed with the MK9KFixedEjectorSuppressed

Below is a Youtube video comparing this ‘hybrid’ Guard configuration with a fixed ejector to an MP5K PDW.
I will try to get around to doing one suppressed as well.


Did a side by side with my ‘hybrid’ Guard using the fixed ejector and the MP5K PDW.
(2) mags loaded to 10 rounds each.
1st mag fired at center mass then 2nd mag fired at head/neck area.

I put the image below at the end of the video but it got cropped off so pasted it below.
Got 18 hits on both. I suspect the two missing rounds were when I was going for the head/neck area. I think I had to compensate for the high cyclic rate more on the MP5K PDW why more rounds seem to be lower on the MP5 target.

You can see in the video it is about 50 feet away.

BTW, along w/ using the fixed ejector, my Guard was using the 9mm Kynshot, 300BLK Tubb flat spring, A5 buffer tube and no weight in the carrier.
I only did this test once and would like to do it again suppressed. The Guard is considerably lighter than the MP5 and curious how things would play out with more mass at the end.

The 556 Tubb flat spring may also be a little flatter but I think about 30RPM faster. Nice thing is that the Guard can be tweaked while the MP5 is what it is.

Note the obvious that the MP5 has no BHO so manipulation of the charging handle is required


IV.     40SW Guard Bolt for use in 9mm

I’ve been watching the release of the .40SW Guard. I have a good bit of .40SW ammo I want to burn up so decided to get the 8″ BCG/barrel combo. Also considering that I’ve switched over to Endomags, I have a Guard Glock lower doing nothing.
I’ve posted some details of the new .40SW bolt and ejector details here.

Details of testing the .40SW Guard (which I did before the following testing) here.

What interests me are the angles cut on the .new 40SW bolt vs 9mm and the new ejector pin design (more on that later).  I had suspected that CMMG would have to make the angles less aggressive than what you see on the 9mm to handle the higher pressures of 40SW.


The angle I cut on my 5.45 bolt and 9mm fixed ejector bolt are between what we see for the CMMG 9mm and .40SW.

Below is an old CMMG post (1/9/2018) talking about changing the angles of the lugs.cmmg-post

I personally think that is the optimal way to go since we are not increasing the reciprocating mass…..but also more expensive for the consumer to buy more bolts vs just swapping weights in the carrier.

Having previous experience with full auto open bolt UZI’s where the bolts were opened up to handle .40SW and still work with 9mm, I was thinking the same should work in the CMMG Guard.

My thought process would be analogous to an MP5 where you change the locking piece from a 100 degree to 80 degree piece when running 147 Gr and suppressed.

Running without a suppressor, mag loaded to 10 rounds – it made it to 8 rounds then a jam.  Using the configuration seen in the picture.556Tubb-A5-9mmKynshot-40SWBolt-NoCan

I’ve been doing a lot of shooting with the 300BLK Tubb spring for the 9mm Guard which is weaker than the 556 Tubb so tried that next – Jammed again after 4 rounds.300BLK-A5-9mmKynshot-40SWBolt-NoCan

Same configuration as above just adding a suppressor (Octane 9) and now it works. 300BLK-A5-9mmKynshot-40SWBolt

Now swapping to the stronger 556 Tubb flat spring and still works with the 40SW bolt.556Tubb-A5-9mmKynshot-40SWBolt

I also loaded a live round then inserted a fully loaded 30 round Endomag into both of the setups above, firing bursts with no malfunctions 2x each.  This is to apply the maximum amount of resistance to the bottom of the carrier.

You may be thinking,  ‘who cares since the RoF achieved with the first configuration  suppressed was 686 RPM vs 663 RPM.  However, what is hard to quantify is smoothness.  As discussed earlier, it is no big deal to get down to even the 400′s but then it is too bouncy.  Cyclic rate data is good but you really need to shoot it side by side to get an appreciation of the progress.

The fact that it doesn’t even run without a suppressor then works with the suppressor mounted should tell you that this is a pretty optimal setup for suppression.  I can tell this is smoother than configuration 1 but even for me, it takes some side by side testing.

The issue now is whether or not the ejector spring is going to hold up.

The .40SW uses a new rounded ejector.   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’m guessing 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 spring appears to be identical to the proprietary spring used in the 9mm and 45ACP Guard.   So I would think that using my preferred buffer/spring/tube combination would result in a short life but that is yet to be determined.  I first ran the 40SW BCG in .40SW to get that dialed in.   I then did the testing above.  So the ejector spring probably has about 300 rounds of mixed .40SW and 9mm run through it.  It seems weaker than out of the box but not weak enough to be replaced yet.

40SW is higher pressure than 9mm so I would think it should die faster in 40SW than 9mm.

It will be amazing if for some reason going to a rounded ejector extends the ejector spring life.  Maybe less surface area contacting the 9mm case….?

At this time, I am still very happy with my fixed ejector setup.  For some reason, my bolt spring is almost dead and barely pushes the bolt out of the carrier and the extractor spring is about dead also with very little tension….but it still runs 100% so I’m going to keep running it till it stops working.

So I plan on running the .40SW with my .40SW upper till it fails or I run out of ammo…but will keep my other 9mm Guard upper as a backup to be used with the factory 9mm Guard BCG or the 40SW BCG (suppressed only of course).

However, I now know I if I want to modify another bolt for my fixed ejector setup that I should be able to clone the 40SW lug angles for a suppressor only bolt.

I would hope that CMMG would make bolts optimized for suppression.  However, I can see how it is much easier to just sell weights to put in the carrier since that is less expensive for the general consumer and a reversible configuration.