Dial Progression on a S&G 6730

[mk5734]

Hello all,

Newbie here. I have a question about test point increments. I see everywhere that during manipulation the test points can be spaced at 2.5 unit increments. That makes sense for a lock with a tolerance of +/-1.25. However, the 6730 has a much tighter tolerance of +/-0.5. Do I have to test at steps of 1 or will some indication be visible at 2.5 or 2 steps?

[MHM]

Hi there. Unless there is something very strange going on, your 6730 is the archetype group 2 lock and so the dialling tolerances are +/-1.25*. Scanning the wheelpack at increments of 2.5 should be fine. That said, I like to go at 2 number increments. Yeah it takes a bit longer but I'm a hobbyist and I'm doing this for fun, so I'm not on the clock.

*Unless I'm missing something of course. What is it about your lock that makes you think it's got group 1 tolerances?

[MartinHewitt]

For a full wheel scan with feeling the CP(s) I have never had the need to do other steps than 2.5 with an R6730 (and at least nearly all other locks), but there are people who like to do steps of 2. The reason is, that the gate forms a flat spot on the wheel and it is possible to see this flat spot even when the fence hangs on one edge and won't drop into the gate. But there are cases where 2.5 is to large. Such a case is when you want to magnify a detail, e.g. determine the exact position of a gate, because 2.5 is to large to hit the gate and get an opening. In this case I use either steps of 2.5/2=1.25 or 1. Another case is when you have all numbers besides one and just need to try the numbers on one final wheel. In this situation you can either read the CP(s) until you see an indication or make smaller steps, so that the lever must fall into the gate. If you are sure with the found numbers method two is faster and the step size will depend on the tolerance of the lock.

[mk5734]

I read that the difference between the 6730 and 6741 is the tolerance. The 6741 is 1.25, whereas the 6730 is 0.5. I can confirm that if I am off by more than half on a unit of the dial, the lock simply will not open. I've tried several times. I believe that I either read it on a post by LockManipulator or heard him say it in one of his videos as well.

I know that it maybe comparing apples and cucumbers, but with the same combination set on the Big Red Wheel CDL-3 and the S&G6730, both behave according to their tolerances; CDL-3 being that of 1.25.

I just found some more information here:

The standard of the industry, this venerable UL Listed Group 2 lock has 0.5 number dialing tolerances for increased security.

Source: https://www.safeandvaultlocks.com/s-and-g-6730-090-3-wheel-combination-lock-class-a-vds-class-1.html

Also:

Capture.PNG

So, should I drop step-size down to 1 or 1.5 in this case?

Thanks!

[MartinHewitt]

I think they have mixed up the information. The S&G Mechanical Safe Lock Guide states, that a newly set combination should be tested by verifying that +0.5 and -0.5 still work. I think they got the information from there.

A 3-wheel Group 2 lock may open up to +/- 1.25 from the true center. The European test standard is fine even with +/- 1.75 for grade A locks like the R6730. Checked on a R6730, which I have lying around here and that has an opening between +/- 1.0 and +/- 1.25. But yes, I have also read somewhere that the zamak lever locks have wider gates than the brass lever ones. Unfortunately I can't find the source of information.

[L4R3L2]

When I AM on the clock, I STILL scan in increments of 2. A missed gate costs time too, and dialing speed can be increased with less regard to precision. I also never 100%assume what lock is on the other side of the door.

Straight from the factory catalog for the following Group 2 locks.
6730, 6731 dialing tolerance +/- 0.5 *
6741 dialing tolerance +/- 1.0 *
* TYPICAL DIAKLING TOLERANCE IS WHAT YOU WOULD EXPECT TO EXPERIENCE IN AN INSTALLED LOCK. THEORETICAL DIALING TOLERANCE IS ALWAYS SLIGHTLY GREATER (USUALLY ABOUT +/- .4 NUMBERS GREATER).

SO, +/-.5 plus +/-.4 is +/-.9 for a gate width of about 1.8 increments. In the real world, I typically find 6730 gate widths to be around 2 to 2.5. In other words, I can usually find at least one full increment on either side of true gate center.

mk5734, what S&G is saying is that when you set a combination, the lock should still be able to be opened when dialing an error of +/- .5 increments. The numbers on the dial are not necessarily truly at gate centers when set, so some latitude is given for the lock to fall within spec. In essence, they are assuring that the edge of a gate will be AT LEAST .5 increment from the number as set on the dial.

[mk5734]

mk5734, what S&G is saying is that when you set a combination, the lock should still be able to be opened when dialing an error of +/- .5 increments. The numbers on the dial are not necessarily truly at gate centers when set, so some latitude is given for the lock to fall within spec. In essence, they are assuring that the edge of a gate will be AT LEAST .5 increment from the number as set on the dial.

I see. So, it's a guaranteed minimum which is also close to nominal. I've tried testing the tolerance on mine and it is unyielding beyond the +/-0.5 increment. The lock simply will not open unless I'm within that range.

Thanks!

[L4R3L2]

When checking for the maximum tolerance, you usually have to check one wheel at a time.

[mk5734]

When checking for the maximum tolerance, you usually have to check one wheel at a time.

I just checked it one-wheel-at-a-time for each wheel, where the combinations tested are of the form:

A+1, B, C
A-1, B, C
A, B+1, C
A, B-1, C
A, B, C+1
A, B, C-1

Each of those will not open my 6730. The number dialed must reside within 0.5 of the combination value.

[L4R3L2]

You have a very unusual 6730. I've never seen one that didn't give me at least one increment on one side or the other, if not both. In your case, you'll have to scan in single increments then. I'd recommend finding another 6730 to learn manipulation.

[L4R3L2]

I don't know if this will add anything to the discussion, but the 6730 I manipulated today is one of the tightest tolerance 6730 practice locks I have.

After I manipulated it, I ran some experiments in light of this thread. I ran each wheel for gate width, and found the following.

Dialing in .5 increments:
Wheel three opened at 96.5, 97.0, 97.5, and 98.0, giving an approximate minimum usable gate opening of 1.5.
Wheel two opened at 31.0, 31.5, 32.0, and 32.5, giving an approximate minimum usable gate opening of 1.5.
In each of these two wheels, there were two whole increment numbers on the dial that fell within opening range. Dialing by twos would find one of the numbers or the other. This lock is pretty tight, but wheel one was the tightest of the three wheels.....

Dialing in .5 increments:
Wheel one opened at 46.0, 46.5, and 47.0, giving an approximate minimum usable gate opening of only one. If gate center were shifted either way by .5, it could require scanning by ones to find the gate. So, I rechecked wheel one in increments of ~.25. In doing so, wheel one opened from 45.75 to 47.25, meaning that the actual usable gate opening was really at least 1.5. Even at 1.5, the edge of the fence could end up being just at or near the very corners of the gate when scanning by twos. The chances of this happeing is remote, and note that the usable gate opening was at least 1.5, meaning that a scan by twos would almost certainly pick up one side or the other. I've always been able to crack this lock dialing by twos, but 2.5 may not work for this particular lock, especially for wheel number one.

[femurat]

I've heard that the new 6730 have tighter tolerances. In fact I was not able to open one.
I always manipulate in 2 increments, not 2.5. I found a gate on wheel 3, then a gate on wheel 1. Then I was unable to brute force wheel #2 in 2 increments.

Cheers

[mk5734]

I'd like to thank everyone for responding to my question. It's very much appreciated.

It does appear that they've tightened tolerances lately. My 6730 has a sticker on the lock-body that says it was manufactured in Dec 2019. The box has a sticker that says more specifically that it was manufactured Dec 8, 2019. But the S&G 2019 catalog (the latest catalog of their that I can find) doesn't reflect these new tolerances, just the old of 1.25.

I'm going to try manipulation again at increments of 2 and let you know the results.

[L4R3L2]

That's interesting. Thank you. I don't expect they would change the specs under the same model. It sounds like they may have gotten new tooling is all. Or maybe a new source. Eventually, new tooling will wear with use, until the limits of the specifications are approached and they have to retool again. It's a common pattern in industry.

[mk5734]

So, I redid a troublesome known combination, but this time graphing it with increments of 2 and BANG! Two gates appeared. Those gates are on W1 & W2. Interesting that W3 didn't show up at all.

I rechecked my tolerances and they are much more inline with what L4R3L2 mentioned his was above. Thanks for checking. Mine are indeed shifted toward the direction of programmed rotation.

Thanks again, everyone! I really, really, appreciate all of your help!