Attempting manipulation of a S&G 6642-001

[MartinHewitt]

Have my new 6642 mounted on beech and got the combination changed by my CCG. The 6642-001 is a Group 2 lock. (Edit: It seems it as Group 2, not 2M lock) it has false gates on each wheel. The false gates are different on each wheel. The false gates look randomly positioned, i.e. they are totally different on all wheels, but as other S&G locks have non-randomly positioned false gates they are possibly not. Let's see how the manipulation goes. My aim is obviously below 2 hours. I believe it is doable. When I have achieved this, I will mount my 6630-090. This has also false gates, but additionally a lever nose roller. The false gates do not look randomly positioned. Wheel 1 and 3 have false gates at 1/3 and 2/3 of the wheel. Wheel 2 has 5 false gates. Is it known where they are?

[Dean Thatcher]

hi

do you have any pics of the insides of these locks?

[MartinHewitt]

Not at the moment, but I will make photos, when I have it manipulated open. It is basically an el-cheapo 6730 (aluminum wheels, zamac lever) where fake gates with the width of true gates, but a depth of maybe halve a millimeter have been made.

After two and a halve graphs I have something which might be the true gate of wheel 3. Next to this gate the wheel is largest, the gate goes most down and halve a turn from this gate it is the general lowest area. That is just like a wheel on my 6730. Besides this gate (25 notches further) I see another clear gate showing not as much of a height and not as much of a depth. So I believe it is not the true fate, but it would be a candidate. Otherwise I see only a very slight indication of a gate which is perhaps nothing.

[MartinHewitt]

I have missed my target time and it is not yet open. 2:02 so far. The combination I have at the moment is 22/99/70. I optimized the left contact point by 0.5 notches and the right contact point by 0.75 notches. On w1 22 was the only significant drop, but it did not look like a real gate, i.e. the drop was to smooth without any lips. I did only a halve scan of w2 because 99 looked relatively good. So I will probably finish that with w1@22. w1@70 really looked like a gate of a 6730 with lips and step drops. Besides the mentioned locations I did not see really good gate indications. Very different from what I did expect. The dial is slowing me down because the notched area is rather narrow, I believe. The AWL scan took this time 16 instead of 12 minutes the last time. That the wheels have a higher turning resistance is not that bad. It slows dialing also down, but I clearly feel every pickup and can hence optimize dialing with more confidence.

PS: Because I do not see reliably multiple gates per wheel scan known, but irregular gate positions would probably not improve opening. But if the distance would be regular, e.g. 1/3, then that knowledge would help.

[femurat]

Even if you are not within your target time, I suggest you continue this manipulation.
Also, I'm curious to see graphs pictures.

Cheers

[MartinHewitt]

That can be arranged.

I worked on this sheet from the outside to the inside. Red line is an AWL scan. With the three short blue lines I tried to optimized the low area around 15. The blue graph in the second column is R15/R15/L scan. The halve red graph in the third column is L16/R/L70 graph. The red graph in the second column is L/R0/L70. Finally the full blue graph in third column is L22/R/L70. The gate at w2 is probably around 99. Neither 22/99/70 nor 22/99/95 will open the lock.

[Jaakko Fagerlund]

When I have achieved this, I will mount my 6630-090. This has also false gates, but additionally a lever nose roller. The false gates do not look randomly positioned. Wheel 1 and 3 have false gates at 1/3 and 2/3 of the wheel. Wheel 2 has 5 false gates. Is it known where they are?

Just checked my 6630 and it has the lever nose roller and two false gates per wheel, positioned 120 degrees apart from the gate. This is the main fail of this lock, as you only have to graph 33 numbers from any point to find a true or false gate and then tryout 27 combinations maximum.

But I have never seen 5 false cuts on these on any wheel, always the same 2 in addition to the actual gate and always 120 degrees away from it.

[MartinHewitt]

This is the 6630. The wheels of the 6642 look the same, but the false gates are at different positions. In this 6630 the green arrows show the positions at w1 and w3 and the red arrows these at w2.

I attached above the graphs' scan which I forgot.

[Jaakko Fagerlund]

This is the 6630. The wheels of the 6642 look the same, but the false gates are at different positions. In this 6630 the green arrows show the positions at w1 and w3 and the red arrows these at w2.

I attached above the graphs' scan which I forgot.

If someone knows better, I say that the wheel is either from a different lock or something has been changed in production.

[MartinHewitt]

The lock is a 6630-090 with production date Februar 3rd of this year. So it has to be a change in production.

PS: My estimation of the false gate positions of w2 take from the photo are: 20, 37, 47, 69, 92.

[MartinHewitt]

PS: My estimation of the false gate positions of w2 take from the photo are: 20, 37, 47, 69, 92.

This is probably a standard 5-false-gate wheel. Measuring the other way round would be 8-31-53-63-80, which is very much like the numbers I found in another forum for a 5-false-gate wheel: 0-10-31-53-63-81

[MartinHewitt]

Dumb! Dumb! Dumb! (The smiley banging its head onto the floor is missing.) The gate at w3 is 27, not 22. Now I see much more, i.e. three gates at w2 and 95@w3 seems to be deeper than 70@w3.

PS: w2 is a standard 3-false-gate wheel with positions at 33, 66, 99.

[Oldfast]

Neat! When you get to the other lock - it'll be interesting
to see how much trouble the eccentric roller gives you.

[MartinHewitt]

I have it, but not yet mounted. Anyway, I want to get the 6642 open first.

[MartinHewitt]

Phew! 4:21. w1 showed only 17 and 27, w2 showed 33, 66, 99, w3 showed only 42, 70 and 95. I started to try out possible combinations, ordered by probability on w1 and by ease of dialing on w2 and w3, i.e. 27/99/42, 27/99/70, 27/99/95, 27/66/42, ... . While I dialed I tested the right contact point. Numbers which led to higher contact point where discarded, e.g. 27/99/70 and 27/99/95 were higher than 27/99/42. So I discarded 70 and 95 for further tests. That really did not matter, because at 27/66/42 the lock was open. The 4:21 I could have easily improved by 42 minutes by not wrongly using 22 instead of 27.

About the position of the false gates I will write soon, when I had a look.