You are here

Home | SA Reference Collection

7.62 X 39 AK-47


Production off the 7.62x39 cartridge in South Africa started in 1987, but testing started in 1986 with ballistic studies by PMP on captured ammunition from either Mozambique or Angola. The 7.62x39 cartridge was never used in any form by the South African military, except in instances where special forces units used captured weapons but even so the practice was frowned upon by the military. Initial development of the 7.62x39 was done to provide reliable ammunition to anti Communist forces operating in Southern Africa, most notably the forces of Jonas Savimbi in Angola who were fighting the MPLA, as well as the Cubans and Russian advisers. At that stage the situation was that many of the ammunition available was booby trapped ammo supplied by the Cubans and included Chinese manufactured ammunition as well.

Despite various requests by a number of collector organizations over the years, we are still refused any access to the Denel archives resulting on a lot of information lacking on various subjects, none more so than on the 7.62x39 project.

The following is a document (V/762/012/86) that was prepared by PMP during the testing of 7.62x39 ammunition for pressure purposes with the goal of manufacturing the ammunition locally. The original document was prepared in Afrikaans and this is a translation of the document.


7,62 x 39 BALL REFERENCE AMMUNITION (“IMPORTED”) LOT 661/83 SHOT FROM 86/12/01 TO 86/12/05


Bullet Lot           : Unknown
Case Lot             : Headstamp 661/83
Primer Lot         : Unknown
Propellant Lot   : Unknown


                                                                   DAY 1      DAY 2       DAY 3         DAY 4         DAY 5

Temperature in firing chamber               25°C       25°C        25°C         25°C         25°C
Relative humidity                                      60%         60%         60%          55%           55%
*Barometric Pressure (mBar)                  869,2       870,2        870,1        873,9        873,9
Time of testing                                           15h00     07h00      08h00       07h00        07h00

*Data received from the SA Weather Bureau


3.1 Barrels

AVL                               321             New/Average              530                      655
AVL                               671               “Old”                 ---------Unknown-----------
AVL                      75110150             New                           ±100               ±1400

3.2 Pressure testing

Pressure testing was done on the case with 3 x 5 copper radials (in AVL-type M10/3 pressure converter)
Pressure and velocity readings were taken separately

Copper Lot number used: 010986

3.3 Device for measuring velocity

Velocity determined by optical light meter
Velocity determined over 2m and 5m.


20 rounds per day were shot at PMP in barrels 321 and 671 and 20 rounds were shot per day at Somchem in barrel 75110150. All rounds were pre-conditioned at +21°C, where 10 round were fired for velocity test and 10 rounds for measuring pressure.


5.1 Individual results graphically illustrated in Annexure A1 and A2

- Annexure A1  : Velocity
- Annexure A2  : Pressure

5.2 Annexures B1 to B6 shows the x̄ and R- charts over 5 days

- Annexure B1 to B4  : Velocity at 2m and 5m
- Annexure B4 to B6  : Pressure with the aid of 3x5 radial coppers (through the case)


6.1 Velocity

A marked difference can be observed between velocities achieved at PMP and Somchem. The Somchem barrel (no. 75110150) reacted in the same way as the “newest” barrel. There is however a big variation in the average velocity and a zeroing value would therefore not be meaningful at this stage according to these statistics. (Seen Annexures A1; B1; B2 and B3)

6.2 Pressure

6.2.1 Pressure is definitely not realistic. The Somchem values are more constant than those from PMP. A zeroing value can therefore not be meaningfully determined. (See Annexures A2; B4; B5 and B6.)

6.2.2 Problems were experienced with the sealant around the pressure holes of the steel cases of the “imported” ammunition. This seems to be the biggest contributing factor of these unacceptable pressure test results. A double layer of isolating tape was used at Somchem to minimize gas leakage, hence better and more constant results (see Annexure A2). Even in this instance there still occurred “out of control” pressures (see Annexure B6).


No meaningful zeroing values could be ascertained according to these results. In the interim, the following zeroing values can be used, namely:

v̅5  : 710 m/s measured separately
P̅5  : 303MPa

The values above were determined as such that:

(i) Velocity were in an acceptable range of ±10m/s
(ii) Pressure correction factor is round about 0 MPa when the velocity correction factor is 0 m/s.

See also memo attached as Annexure C

8. Recommendations

This Section recommends that:

- The zeroing values according to memo, attached as Annexure C be used in the interim.
- 10 000 rounds be reserved that will be zeroed as reference loads after successful testing of the 7,62x39 Ball (PMP rounds). It is imperative, because we foresee a lot of problems if the velocity and pressure of “brass cased ammunition” are corrected with “steel cased ammunition”

*** The annexures are not available - will be uploaded as soon as possible.



Some of the original test rounds used by PMP and Somchem, with later test shown below (not mentioned in original report)


After testing, production of test lots were started with modified 7.62x51 NATO surplus cases.

   Modified 7.62 NATO cases, G denoting tracer, although normal ball ammunition was tested. The two loaded specimens below are both ordinary GM ball.



Further testing was done on imported SAKO cases with South African bullets


Manufacturing dates for 7.62x39 in South Africa were over a two year period, namely 1987 and 1988. By 1989 hostilities were drawing to a close. Namibia gained independence in 1990, culminating in the withdrawal of South African forces and replaced by UNTAG (UN Tactical Assistance Group). Most of the ammunition was sold on the international market by Denel. Below are the three Pressure Test versions.

Both no crimp and three stab crimps were observed on 1987 M1A1 production. On 7.62 NATO ammunition the M1A2 is a designation of ball powder. Three version of M1A2 have been observed, namely no primer crimp, 3 stab crimp and purple PA. As with the 7.62 NATO, the colour of the PA does signify any particular function

New primed empty case, berdan primed

   Unusual dummy with double extractor groove. Bullet is 7.62x51 Ball


                           The 1987 headstamp is the only version of the drill cartridge examined, except for the un-headstamped version below.


Although the PA colour on normal Ball ammunition does not signify anything, the green PA on 7.62 NATO was used in certain circumstances for Bisley loads. As this was not the case with the 7.62x39, the green PA might be REFERENCE loadings. The loads are found with normal Ball as well as GM Softnose as can be seen below.



                            Black tipped dummy (AP). However, the core is hollow on  normal GM jacket. As with 7.62 NATO, South Africa only manufactured AP ammunition for export.

                            Unprimed case with double extractor groove. The significance is not known.


All specimens examined are without headstamp.








info and box labels to follow


                                    This specimen was made by special forces troops from existing cases




Only one type have been seen thus far. No dated headstamps from any period of production have been examined. It uses the same Belgian supplied bullet as was used in later production of the South African 7.62 NATO that replaced the "orange series" post 1983


The specimen below might be an "unofficial" version of a short range. Headstamp is 11 87






These are modified 7.62 NATO cases, in all probability for an export contract.


These were experiments done by PMP in all probability, but the reason is unknown as well as why local cases were not used.





These were reloaded and tested by PMP





Rumoured to be for or by some special forces unit as some kind of safety round with a steel (magnetic) ball bearing in the tip.


   Test round with 147gr. 7.62 NATO bullet

                          Specimen with thin rim and extractor groove




                          This cartridge was featured in RSACCA 204 (2003) and was confiscated by The South African Police

Two specimens loaded for the local sporting market by OPM on LAPUA cases