PDA

View Full Version : MSM Fal Para Return Spring Review


gunplumber
July 18, 2017, 09:41
INTRO
Original Para springs are rare
DSA reproduction para springs are a little weak (-10%+)
Falcon Para springs are grossly defective garbage (-30%+)

So I was pleased when some members decided to make a better reproduction. I was pleased that I could assist in some manner with sourcing tech specs.


BACKGROUND
I received a set for evaluation some months ago, but I was unable to put a review together earlier because I had no para FAL among the 30+ rifles in for builds. They looked good, but that wasn't enough.

Then Daniel/R1shooter, DSA's paid shill, posted:
Having now seen this thread and for a while wondering about it I went to the office and actually tested 3 original para spring sets, well results are not what you may initially think, question is what are the true numbers of the sets you sell?

I called bullshit and challenged Daniel to reveal what type of strain gauge equipment he used, how he measured, what results he obtained, and what standards he compared them to.

As I have come to expect from R1shooter, he just obfuscated, changed his story, and carried on like a whiny little bitch. . .unable to produce any evidence that any such test ever occurred. Because he is a lying sack of shit. It is The Chicago Way.™

Now everyone knows that I don't bullshit anyone. If I say something it is because I believe it is true. And I'm not particularly stupid or foolish (although I have been fooled), so when I say something is, one can be pretty sure it is. Or at least I have strong reason to believe it is.

At the same time, while "DSA para springs feel at least 10% lighter than factory springs" while absolutely true in my experienced opinion, is still subjective. So it's time to get some objective data.

EQUIPMENT
I did not believe my various manual and digital strain gauges were of the appropriate range and precision to make for a fair test & comparison. Strain gauges perform best in the middle of their range. So a Mk10 Series 2-50 from Cole-Palmer (which is a pretty nice gauge for under $500), performs best at 25 pounds. The further from the 25# median, to the 0 minimum or the 50 maximum, the greater the potential error. Error in strain gauges is a percent; the higher the range, the higher the error - because a 0.5% error is more on 500# than 20#. The higher the range, the lower the resolution (significant figures.) In this case 0.5% full scale. A Mk 10 Series 2-20 performs best at 10 pounds, half way between 0 and the 20# max.
Mk10 M2-20 (.02#)
Mk10 M2-50 (.05#)
Mk10 M2-100 (.1#)
Mk10 M2-500 (.5#)

I ended up purchasing a Nidec - Shimpo FG 3006 push & pull digital force gauge at the 24 pound range (~$250). I chose it because the accuracy was superior at that range (0.3% vs 0.5%) to the Mk10, and I'm a tool slut, always looking for an excuse to buy another neat tool. It seemed the best value for my purposes. Max error on this unit from the calibration certificate was -0.10% @ 97.9 Newtons (20 pounds), or -0.1 Newton error (.02#).

http://www.arizonaresponsesystems.com/wp/ars-shop-shimpo-dfg.jpg

SETUP

I decided to use a modified version of the REME Arsenal Inspection and Repair Standards for the L1A1 return spring. That is, an assembled rifle with bolt/carrier retained by the BHO. Measured is the force required to move the carrier enough to disengage the BHO. This test is performed with the receiver cover removed. As the para receiver cover incorporates the spring guide, I have to work off the charging handle instead of the carrier.

I used a new PAC Gear Logo IMBEL, a DSA Para carrier, and a DSA folding charging handle. First step was to cut the receiver for the para receiver cover nose.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-01.jpg

Next a lightly filed and stoned the receiver rails to remove any high points.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-02.jpg

I did the same to the bolt carrier and the charging handle. Purpose was to remove any added friction from the measurements -as that friction may not remain consistent from one test to another.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-03.jpg

Testing bolt carrier and charging handle independently. Must fall under gravity with receiver at a 45 degree angle.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-04.jpg

After testing independently, I repeated as an assembly. Carrier and charging handle together fall under their own weight.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-05.jpg

At this point, I attempted to install an IMBEL bolt into the DSA carrier. It would not fit. I tried another IMBEL bolt. It also would not fit. Incomplete machining of the firing pin clearance prevented assembly. Geez! And here I thought I had a decent carrier because it only took a little filing to get it to fit in the IMBEL receiver. So I have to pause the test and review until I can machine the correct firing pin clearance on this out-of-spec DSA para carrier.

(an hour later)

Ok, I corrected the problems with the DSA para carrier so the bolt would fit and I mounted the whole assembly with a DSA spring assembly to a stand, C-clamped to the bench so as to eliminate any movement.

I discovered an additional problem with my original testing plan - I don't know if I got Kg and Lb mixed up, or just under-estimated the greater force requirement in a para, but I'm not getting BHO release before 26 lbf, at which point my 22 lbf strain gauge goes into error mode. I waiting to hear back from the Itin Scale Company to see if they'll allow an exchange for a 50 lbf unit.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-06.jpg

Now all is not lost - this is the maximum force at the maximum compression of the three spring assembly installed with the bolt and carrier. It does not prevent me from testing the springs independently.

The way the FN print specs the springs is force over distance (Hooke's Law). Robert Hooke in 1678 stated ut tensio, sic vis, or "the extension is proportional to the force". F = kX where k is a spring constant (stiffness). It applies in a linear fashion except at the extremes. New springs are measured in OAL as free length. After taking initial set from maximum compression, the new OAL is "solid".

(Diagram) https://upload.wikimedia.org/wikipedia/commons/f/f0/HookesLawForSpring-English.png

FN prints for springs (like my Russian AK spring prints) has a right triangle representing compression, with the hypotenuse indicating force, following Hooke's law.

FN Specifications

Outer Spring
L(f)= 297mm (11.69")
L(s)= 287mm (11.30")
wd=1.05 +/-0.07mm
od=8.76 +/-0.14mm
id=(not listed)
coils min= 88.75 (not including 1/2 end coils)
coils max= 90.5 max
end coil max d= 9.05mm
end coil bend= 4.3+/-0.3mm (this is to trap the spring on an FN guide)

Step 1: 232.8 mm +/- 10.6mm (9.17" +/- 0.42") @ 1.75 kgf (3.86 lbf)
that is 1.75 kgf must be reached @ 243.4mm-222.2mm (9.58-8.75") compression
Step 2: 130mm +/- 18.3mm) (5.12 +/- 0.72") compression, 4.75 kgf (10.47 lbf)
that is 4.75 kgf must be reached @ 148.3-121.7mm (5.84-4.79") compression

note: Original specification was for end coils to be squared by grinding, but this feature is crossed out on my print.

Inner Spring (2)
L(f)=
L(s)=
wd=
od=
id=(not listed)
coils min= (not including 1/2 end coils)
coils max= max
end coil max d=
end coil bend=

Step 1 (assembled):
Step 2 (assembled):

These ratings are well within the range of my strain gauge, so I just need to throw together a guide rod for pushing on the spring.

Measuring Springs for Comparison.

Counting coils is tedious, but clicking a razor blade across the top of each coil makes it easier to keep count. Closed end springs measured as free coils + closed half coils.

Overall length measured on the mill using the DRO and a pointer. Free length (Lf) as new. Solid length (Ls) measured after 10 manual cycles to maximum compressing - "going solid".

note: It came to my attention only after completing the MSM test that the set MilSurpMonkey sent to me had already been cycled 1000 times. This both accounts for the smaller change between Lf and Ls, but also is a surprise there was any change. I speculate the springs "relaxed" in the month they were out of the gun.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-07.jpg

Measuring Spring Strain in Isolation

Pretty simple set-up. Just a polished rod as spring guide with the two specification points, and the allowable range, indicated on the DRO. Springs are tested after ten cycles to maximum compression (solid); using a sleeve over the spring guide, compressed by the strain gauge.

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-08.jpg

Mark target in green and permitted range in red. FAIL if kg rating is achieved outside red zone. PASS if designated kg rating is achieved anywhere in the red zone, with the green line being ideal.

Step 1: 1.75 kg = 243.4mm / 232.8mm / 222.2mm
Step 2: 4.75 kg = 148.3mm / 130mm / 111.7mm

Following Hooke's law, if the positions were plotted on a graph, the rise/run should be 1:1 for the majority of the range

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-09.jpg

Inner spring specs are from an "assembly" print, in combination with inner spring guide. Currently missing print for individual inner spring. FAIL if kg rating is achieved outside red zone. PASS if designated kg rating is achieved anywhere in the red zone, with the green line being ideal.

Step 1: 2.0 kg = 245mm / 234mm / 223mm
Step 2: 7.0 kg = 148mm / 128.5mm / 109mm

None of the test results come anywhere near 7 kgf for essentially full compression of the paired inner springs, so I suspect it is meant for outer and inner combined - will run again and see what happens.. Yep, after running Outer + Inner + Inner, the readings came into line with what I expected. In reexamining the spec, I realized I was an idiot - it clearly shows the spring sections on the outside being larger diameter than inside spring section - so yeah, "l'assemblage" test is the whole thing - inner x2 + guide + Outer. Hereafter referred to as O+I+I

http://www.arizonaresponsesystems.com/wp/fal/fal-para-returnspring-test-10.jpg


New DSA Outer

L(f)= 300.4mm
L(s)= 297.6mm
wd= 1.05mm
od= 8.8mm
id= 6.7mm
coil= 86 + 2/2 RH closed end, round

New DSA O step 1 = (target 1.75) 1.69 / 1.93 / 2.20 | AVG 1.94 kgf
step 1 target = 0.18 kgf > spec | mean = 0.19 kgf > spec
step 1 target reached within range = PASS

New DSA O step 2 = (target 4.75) 4.70 / 5.07 / 5.67 | AVG 5.14 kgf
step 2 target = 0.32 kgf > spec | mean = 0.39 kgf > spec
step 2 target reached within range = PASS

New DSA Inner

L(f1)=145.9mm
L(f2)=147.0mm
L(s1)=144.7mm
L(s2)=145.6mm
wd=0.8mm
od=6.4mm
id=4.8mm
coil=59 + 2/2 LH closed end, round

New DSA I+I step 1 = 1.11 /1.39 / 1.64 AVG 1.38kg
New DSA I+I step 2 = 3.25 / 3.55 / 3.82 AVG 3.54kg

New DSA O+I+I step 1=(target 2.0) 2.79 / 3.34 / 3.75 AVG 3.29 kgf
step 1 target= 1.34 kgf > spec | mean=1.29 kgf > spec
step 2 target reached before range = FAIL

New DSA O+I+I step 2=(target 7.0) 6.75 / 8.35 / 8.95 AVG 8.02 kgf
step 2 target= 1.35 kgf > spec | mean=1.02 kgf > spec
step 2 target reached within range = PASS

MSM Outer
L(f)= 287.8mm
L(s)= 287.2mm
wd= 1.02mm
od= 8.8mm
id= 6.6mm
coil=88+2/2 LH closed end, flat

MSM O step 1 = (target 1.75)1.29 / 1.71 / 2.25, AVG 1.75 kgf
step 1 target= 0.04 kgf < spec | mean= spec
step 1 target reached within range = PASS

MSM O step 2 = (target 4.75)4.49 / 4.90 / 5.47, AVG 4.95 kgf
step 2 target= 0.15 kgf > spec | mean= 0.20 kgf > target
step 2 target reached within range = PASS

MSM Inner
L(f1)= 142.0mm
L(f2)= 142.2mm
L(s1)= 141.5mm
L(s2)= 141.4mm
wd= 0.75mm
od= 6.4mm
id= 4.6mm
coil (s1)= 59+2/2 LH closed end, flat
coil (s2)= 58+2/2 RH closed end, flat

MSM I+I step 1 = 0.72 / 0.98 / 1.22, AVG 0.97 kgf
MSM I+I step 2 =2.71 / 2.89 / 3.05, AVG 2.88 kgf

MSM O+I+I step 1= (target 2.00) 2.01 / 2.62 / 3, AVG 2.54 kgf
step 1 target= 0.04 kgf 0.62 kgf > spec | mean= 0.54 kgf > spec
step 1 target reached before range = FAIL (failure is by .01 kg. Precision of scale is +/ .01 kg)

MSM O+I+I step 2= (target 7.00) 6.1 / 6.59 / 7.5, AVG 6.73 kgf
step 2 target= 0.41 kgf < spec | mean= 0.27 kgf < spec
step 2 target reached within range = PASS

MSM Outer
L(f)= 287.8mm
L(s)= 287.2mm
wd= 1.02mm
od= 8.8mm
id= 6.6mm
coil=88+2/2 LH closed end, flat


MSM Inner
L(f1)= 142.0mm
L(f2)= 142.2mm
L(s1)= 141.5mm
L(s2)= 141.4mm
wd= 0.75mm
od= 6.4mm
id= 4.6mm
coil (s1)= 59+2/2 LH closed end, flat
coil (s2)= 58+2/2 RH closed end, flat

MSM O step 1 = (target 1.75)1.29 / 1.71 / 2.25, AVG 1.75 kgf
MSM O step 2 = (target 4.75)4.49 / 4.90 / 5.47, AVG 4.95 kgf
MSM I+I step 1 = 0.72 / 0.98 / 1.22, AVG 0.97 kgf
MSM I+I step 2 =2.71 / 2.89 / 3.05, AVG 2.88 kgf
MSM O+I+I step 1= (target 2.00) 2.01 / 2.62 / 3, AVG 2.54 kgf
MSM O+I+I step 2= (target 7.00) 6.1 / 6.59 / 7.5, AVG 6.73 kgf

DSA Old Outer
L(f)= mm
L(s)= mm
wd= mm
od= mm
id= mm
coil= LH closed end, round


DSA Old Inner
L(f1)= mm
L(f2)= mm
L(s1)= mm
L(s2)= mm
wd= mm
od= mm
id= mm
coil (s1)= LH closed end, round
coil (s2)= RH closed end, round

DSA Old O step 1 = (target 1.75) / / , AVG kgf
DSA Old O step 2 = (target 4.75) / / , AVG kgf
DSA Old I+I step 1 = / / , AVG kgf
DSA Old I+I step 2 = / / , AVG kgf
DSA Old O+I+I step 1= (target 2.00) / / , AVG kgf
DSA Old O+I+I step 2= (target 7.00) / / , AVG kgf


Falcon Outer
L(f)= mm
L(s)= mm
wd= mm
od= mm
id= mm
coil= LH closed end, round


Falcon Inner
L(f1)= mm
L(f2)= mm
L(s1)= mm
L(s2)= mm
wd= mm
od= mm
id= mm
coil (s1)= LH closed end, round
coil (s2)= RH closed end, round

Falcon O step 1 = (target 1.75) / / , AVG kgf
Falcon Old O step 2 = (target 4.75) / / , AVG kgf
Falcon Old I+I step 1 = / / , AVG kgf
Falcon Old I+I step 2 = / / , AVG kgf
Falcon Old O+I+I step 1= (target 2.00) / / , AVG kgf
Falcon Old O+I+I step 2= (target 7.00) / / , AVG kgf

I'm done for a while. It is difficult to take readings while simultaneously holding and reading the gauge, the spring guide, and keeping the springs guide in horizontal alignment. I took 4 readings for each step and averaged them. This is a very small sampling, and the spring sets were not equivalent (DSA were new, MSM had 1000 cycles). My conclusion is that the DSA springs are heavier than spec and therfore a new spring - because their old springs were enough ligher than standard that I could Identify them simply by racking the charging handle. Falcon springs are so weak as to be a joke, but eventually I will test them anyway just for the data points. I'll also test the next set of IMBEL springs that show up. I welcome possible improvements to the testing procedure.

But the question everyone is asking is "are the MSM springs good to go?" The answer is a resounding yes!

The DSA springs? I don't know. I will use them in the customer gun and test them again after some use, but right now I'm unsure how the heavier rating will manifest in an assembled rifle. I was definitely surprised by how much more difficult it was to charge, but I need to allow for the lack of leverage doing it on a bench fixture. In any case, they are probably superior to their old ones. It is quite possible that new MSM springs will rate higher than the tested sample as well.

MilsurpMonkey
July 20, 2017, 13:24
All this sure makes me wish I'd have finished 3rd grade math :D

All joking aside I'm really glad someone cares enough to actually test these things in a scientific manner. I'd have to go back to the shot log on the set I sent you for exact numbers, but I know they had around 1k rounds on them before they left my para. I mention this because, in theory, it means they should have taken all the "set" and your numbers should be the actual functioning forces for most of the springs life. I'll have to re-read this one a couple times to get all of your data to sink in, as usual it is very thorough.

gunplumber
July 20, 2017, 13:37
I obtained a new set of DSA springs and the first impression I had was - WTF!!?

They were so heavy I could barely charge the bolt. Now this was with the bare receiver mounted on a magazine, so much less leverage; but still, it was significantly heavier than I'd seen before and appeared heavier than FN/IMBEL standard. I checked everything for binding and there was none. Put this in the context of all the previous DSA springs being noticeably light.

My conclusion is DSA has changed their springs, but won't say so because that would be acknowledging there was a problem with the previous ones they were selling - and what motivated you to make better ones.

As to the springs you sent already having been cycled, that would explain why I noted such a small difference between free and solid OAL. Like they relaxed .2mm since having been in a gun last.

MilsurpMonkey
July 20, 2017, 14:19
Probably released an "overrun" of South African contract springs :beer:

I've got a set of springs I bought at a gunshow as DSA, got them home and they were far greater diameter, longer, and much stronger than anything I'd had on hand including Argentine and Belgian sets. I figured they were FN49. They ran my para alright, but I had issues with them at our fall shoot (random failure to eject even with most ejected cases being thrown well clear, had to pogo stick it twice on our 3 gun event) swapped them out with one of our (then brand new) test sets (first spec, they were at least 3/4" short and the inner springs would go solid) and even that set fixed those issues. Now I'm wondering if they really were DSA springs. Regardless, our current springs have been running without issue for Countershot and I in our homebrews, plus no customer complaints have us pretty happy. If your findings suggest tweaks need to be made, we can address those as well in future runs.

gunplumber
July 20, 2017, 16:23
If your findings suggest tweaks need to be made, we can address those as well in future runs.

I think your springs are fine. DSA's new ones might be. I just got the springs today you drop shipped for my customer. I'll uses those to revise the Lf) in the chart above. But I've spent the greater part of two days trying to put a fair test together and I gotta say I'm burned out right now. I'll return your other para parts shortly.

As you can see, the one MSM FAIL was by 0.01kgf (under spec) on step 1 of the combined spring assembly.

In the same test, the one New DSA FAIL was by 1.34 kgf (over)

With the gauge only measuring to a 0.01 kgf accuracy, and with the built-in error of my testing methods (even when reduced by averaging four measurements), I'd say yours is not statistically significant, while DSA's is.

In either case, both passed the same test at STEP 2 (full compression), with MSM mean being 0.27 kgf under target and DSA mean being 1.02 kgf over.

I noticed your spring diameter was smaller than DSA's by a hair - were you using closest inch equivalent to metric 1 mm? Spec is 1.04-1.06mm (.04" or #60).

As I said before, I'm happy to rerun these tests on other samples and revise the testing procedure as necessary to improve the value of the results, just not right now.

MilsurpMonkey
July 20, 2017, 18:09
The testing protocol is certainly above and beyond anything we figured our springs would be put to. You'd be surprised at how many spring manufacturers aren't willing to do similar testing on end products even for QC. I am glad we prompted DSA to better themselves, even if it is just one small product. Maybe we should look into doing some forged receivers machined to spec... :rofl:

lew
July 20, 2017, 18:32
Excellent effort, Mark, and it's very much appreciated.

Bwana John
July 20, 2017, 18:46
I have a set of used FN49 springs that were sold by SARCO as FAL para springs.

They do function in my para FAL.

I would send them to you to include in the test if you are interested.

pl521
July 20, 2017, 19:26
Outstanding work Mark. You put a lot ot time, money and effort testing the springs and we all appreciate it!

ftierson
July 20, 2017, 19:43
Great info, but you forgot the DSA 'feel' test... :)

Forrest

MilsurpMonkey
July 20, 2017, 19:57
I noticed your spring diameter was smaller than DSA's by a hair - were you using closest inch equivalent to metric 1 mm? Spec is 1.04-1.06mm (.04" or #60).

As I said before, I'm happy to rerun these tests on other samples and revise the testing procedure as necessary to improve the value of the results, just not right now.



I remember Countershot, our spring manufacturer, and I discussing the diameter issue and there was a reason for it, I just can't remember what. By the time our final specs were settled on we'd had to write down everything to the smallest details to keep from confusion. The notes are in Missouri. I completely understand the "burnt out" feeling. Before you were able to provide the FN information from Bruce, we'd already measured every para set we could get our hands on, and came up with specs for over 15 test sets for trial, all those numbers and correlations to function and pressure are enough to drive anyone batty. Overall I think Countershot and I agree we are glad the project landed in our lap and really miss having something to work on. It was a lot of fun, a great challenge, and exceedingly educational.

Thanks for your honest evaluation of the final product, input is always appreciated on something like this. :beer:

ETA: I completely forgot that in addition to the round count on those springs I stored the rifle with the bolt open on an empty mag, we really wanted to make sure there were no surprises when we started releasing these to the folks here on the files.