Diagnosing Inaccurate AR-15

[Scandiacus]

I recently assembled an AR15 upper from scratch, but the accuracy has been terrible, with 5 shots at 100 yards consistently grouping around 8" with M193 ball ammo off of bags from a cold bore, and even worse with Federal 55gr 223 Rem. I'm trying to diagnose and fix the issue.

Here's the relevant parts list:

- Aero Precision stripped upper
- 16" Faxon pencil barrel, 1:8 twist, 5.56 NATO (free floated) with pinned, low profile gas block
- 13" Faxon carbon fiber free float handguard
- Faxon 5.56 M16 Bolt Carrier Group
- DPMS mil-spec A2 flash hider
(It's basically a WWSD2017 upper, if you're familiar with that from watching InrangeTV)

Here's the troubleshooting I've tried so far, without any luck:

- Cleaned the barrel very well. Three times. Checked for damage to the crown and found none.
- Verified headspace before first test firing using Forster 5.56 Go/No Go gauges
- Ruled out my own marksmanship. With my Ruger AR556 I was able to group 5 shots into about 2" at 100 yards firing under the same conditions. With the AR556 and irons, I shot 3". I got about a 4" group of 5 shots at 100 yards using my iron-sighted 1894 Marlin with 357 Magnum handloads. I'm not an amazing marksman by any stretch of the imagination, but getting 8" from an optically sighted AR15 doesn't seem right compared to the other rifles.
- Tried it with three different sighting systems, to rule out an issue with them: A Vortex Spitfire 1x prism sight, a Vortex Crossfire red dot, and a set of Magpul backup iron sights.
- I had initially torqued the barrel nut to 75 ft. lbs. to get it timed up properly with the gas tube. It shot terribly. I later saw that Faxon recommends 40 ft. lbs. for best accuracy, so I reinstalled the barrel with a shim to get it to exactly 40 ft. lbs. Still no luck. Before each barrel installation, I did season the threads by torquing up to 30 ft. lbs. and then removing the nut 3 times before applying final torque.

Here's what I'm planning to try next:

- 45gr 223 Remington ammo, and 62gr M855 ball ammo, just in case this rifle is extraordinarily picky about ammo. Though I would not be happy with an AR that doesn't like 55gr ammo.
- Fire with the A2 flash hider removed. Maybe it's interfering somehow?
- Test the bolt from my Ruger AR in it (after confirming headspace) to rule out issues with the Faxon bolt.

If those don't fix it, I'm out of ideas. I don't expect jaw-dropping accuracy from a pencil barrel, but 8" is pretty bad. Does anyone know of other stuff I should check? Could my barrel be a lemon?

Thanks for any advice you can offer!

[Tacdriver83]

When you run a patch through do you feel anything abnormal? Is it a consistent push all the way through? I would send the barrel back after trying different ammo/bolt.

[Rootsy]

Make sure the gas tube isn’t contacting the inside of the float tube.

[Scandiacus]

When you run a patch through do you feel anything abnormal? Is it a consistent push all the way through? I would send the barrel back after trying different ammo/bolt.

Patches go through smoothly and consistently. Subjectively, it feels ever so slightly tighter than my AR556, but that might just be my perception, or the fact that it's far less worn in.

[Scandiacus]

Make sure the gas tube isn’t contacting the inside of the float tube.

Good idea. Just checked, though, and it has plenty of clearance.

[Scandiacus]

Could an out of spec upper receiver cause something like this? I'm trying to think of a way to rule out the Aero upper, short of buying another one, or disassembling my Ruger upper (which is NOT gonna happen, it shoots well and I don't want to mess it up). Coming up blank, though.

[Moleman-]

Can you scope it to see what groups you get with a 9x on it? I'd recheck the barrel snout contact. Before you take anything apart, put the buttplate on the ground and support the stock with your feet. Grab the end of the handguard with one hand and see if the barrel wiggles with the other. I've run into improperly made free float barrel nuts that bottomed out before properly holding the barrel extension ring to the upper, as well as loose barrel extensions. Some movement is normal as you're bending either the handguard or barrel, but you'll be able to feel if something is loose. How tight did the muzzle brake get tightened? Too tight is as bad as too loose. Did the pinning process of the gas block make a constriction in the bore?

[Tacdriver83]

No keyholing on the target or any weird looking impacts?

[Scandiacus]

Can you scope it to see what groups you get with a 9x on it? I'd recheck the barrel snout contact. Before you take anything apart, put the buttplate on the ground and support the stock with your feet. Grab the end of the handguard with one hand and see if the barrel wiggles with the other. I've run into improperly made free float barrel nuts that bottomed out before properly holding the barrel extension ring to the upper, as well as loose barrel extensions. Some movement is normal as you're bending either the handguard or barrel, but you'll be able to feel if something is loose. How tight did the muzzle brake get tightened? Too tight is as bad as too loose. Did the pinning process of the gas block make a constriction in the bore?

I don't have a 9x to test with, unfortunately. Barrel doesn't seem to be loose - no wiggling. I should've mentioned, this handguard uses a milspec barrel nut, and when I reinstalled the barrel to 40 ft. lbs. I also tried a brand new barrel nut made by CMMG. I didn't use a torque wrench on the muzzle device (standard A2 flash hider), but I used a shim with the crush washer so that, when hand tight, it was about 45 degrees out of time. Then just used a wrench to rotate it into time. From what I read, that sounded like the usual approach for A2s. I don't know how to check if there's a constriction around the pin hole, but it was drilled by Faxon at the factory, so if that's the problem it should hopefully be covered by their warranty.

Thanks for the suggestions!

[Scandiacus]

No keyholing on the target or any weird looking impacts?

Nope, bullet holes looked normal.