Page 3 of 6

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 2:29 am
by femurat
MartinHewitt wrote:Amplify around 22.5 with w1/w2 still parked at 85. I do it with halve steps, i.e. 20, 21.25, 22.5, 23.75, 25.


Is it really necessary? I mean, this is a creative way to take 5 readings instead of 6, but do we need it?
I prefer to take readings every increment, in this case 20, 21, 22, 23, 24, 25. It avoids confusion and takes me a few seconds more.
And, even if I have some experience with manipulation, I take all the readings again. Remember that consistency is key here.

No hard feelings here, I'm just thinking out loud.

Cheers :)

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 2:52 am
by ratlock
Thanks for the advice, I'm thankful for all the information, and input I can get.
I need all the help I can, never graphed a safe before..
Thanks.
The Rat

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 3:07 am
by femurat
If you have never graphed a safe, you may enjoy this tutorial.

Cheers :)

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 4:01 am
by MartinHewitt
femurat wrote:
MartinHewitt wrote:Amplify around 22.5 with w1/w2 still parked at 85. I do it with halve steps, i.e. 20, 21.25, 22.5, 23.75, 25.

Is it really necessary? I mean, this is a creative way to take 5 readings instead of 6, but do we need it?
I prefer to take readings every increment, in this case 20, 21, 22, 23, 24, 25. It avoids confusion and takes me a few seconds more.
And, even if I have some experience with manipulation, I take all the readings again. Remember that consistency is key here.

The sheet I use has lines for 20, 22.5 and 25. So graphing 20, 21.25, 22.5, 23.75, 25 is easy while graphing 20, 21, 22, 23, 24, 25 is not. As I had never issues because of this I never had a reason to change it.

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 9:14 am
by femurat
MartinHewitt wrote:
femurat wrote:
MartinHewitt wrote:Amplify around 22.5 with w1/w2 still parked at 85. I do it with halve steps, i.e. 20, 21.25, 22.5, 23.75, 25.

Is it really necessary? I mean, this is a creative way to take 5 readings instead of 6, but do we need it?
I prefer to take readings every increment, in this case 20, 21, 22, 23, 24, 25. It avoids confusion and takes me a few seconds more.
And, even if I have some experience with manipulation, I take all the readings again. Remember that consistency is key here.

The sheet I use has lines for 20, 22.5 and 25. So graphing 20, 21.25, 22.5, 23.75, 25 is easy while graphing 20, 21, 22, 23, 24, 25 is not. As I had never issues because of this I never had a reason to change it.


Now that you mention it, it makes perfect sense.

I use graph paper, and draw the scale so that every number has its own line.

Cheers :)

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 11:11 am
by MartinHewitt
femurat wrote:I use graph paper, and draw the scale so that every number has its own line.

If you make a scan, which numbers to you graph? Every second? Also every second and a half?

Martin Hewitt

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 3:24 pm
by ratlock
Had more of a look around number one. Maybee its me getting better at looking and feeling what I am doing, with a bit more finess. I m going for 23.5ish for first.

Heres my latest graph dialing 23.5 in number 3 wheel. Parking wheel 1 and 2 testing, and repeat at 2.5 increment intervals.

Im going with 92 for the next number. Going by the graph, does it look like 70 is the third number.?

More graphing, (or maybee try some numbers tomorow)

Image

PS just noted Ive written wrong number for right contact in the margin, Im one whole number written down wrong, (its been a long day). Sould read 3.3/3.2/3.3

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 4:26 pm
by MartinHewitt
The gate candiates you have here are very wide. In my locks the gate is much more defined, but this might be due to my brass lever and your Zamac lever. But even considering this the valley around 70 looks to undefined to be a gate.

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 4:55 pm
by Jaakko Fagerlund
It might be a gate, it might not be a gate, but all you can say for sure is that it is one very low spot that you can leverage as a next parking place if your 92 doesn't give anything more.

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 6:20 pm
by MartinHewitt
And two notes:

1) I do always graph left and right contact point unless the left contact point is just a flat line. Sometimes the right contact point says it is a gate, but the left says it is not (i.e. both graphs going to the left and not together). In this case it is most probably not and it helps to discard wrong indications.

2) I use multiple colors for multiple graphs on one sheet, because it is helpful to see the change in the graphs. Was there really a significant change in the graph with w3 positioned at the suspected gate? At the moment it might be to soon for this, because it appears you are still looking for your way to test for contact poins and to draw the graphs. If I compare your second and last graph, it seems to me they are very similar (unless I interpret them wrongly as I don't understand the "3.3/3.2/3.3"). So it might be that either the gate of w3 is not at the selected number or you are doing something wrong. (Edit: See attached graph. Red is your second graph, green your last one.)

And one more: I do have problems understanding what you are exactly doing. Oldfast is describing his operations nicely, but using your own description is fine. Which wheels are moving? Which wheels are staying at which number? In which direction is the movement?

I wonder ... What we are doing is an optimization. What is the lowest point? There are a lot problems like this in science (lowest energy of a protein structure and such stuff). They have their optimizations algorithms for this. Can we use them for opening locks quickly or at all?

Re: Chubb combination lock fail

PostPosted: Thu Mar 23, 2017 11:49 pm
by ratlock
Thanks​, keep me right on this, as I might make mistakes.
The last graph was me trying to figure out the second number, here is how I tried it.
Assuming 23.5 is correct for wheel 3.
I dialed starting all wheel left, (anti clockwise)I think I choose around 27 to start graphing.
Then one turn clockwise to where I think the gate on wheel 3 might be,(23.5). Then check contact points.
All left again, this time up 2.5 graduations more than the last. Spin wheel 3 back one full turn to 23.5 check contact points, and repeat..

I like the idea with the two colors. It makes it easier to see comparisons.
Once I make less mistakes and get more accurate, that's the best way forward.
Yesterday I threw 4 graphs away because they were garbage. The wife, and kids were doing things in the kitchen, and dumping stuff on the table. With a big contact mic, and headphones on through a practice guitar amp, it was hard to concentrate with all the noise. Funny thing is, no one usually wants to make conversation with me, untill the head phones go on.

Re: Chubb combination lock fail

PostPosted: Fri Mar 24, 2017 5:25 am
by oldlock
When learning 2.5 numbers gives no margin for error.

Look every number.

Re: Chubb combination lock fail

PostPosted: Fri Mar 24, 2017 8:31 am
by ratlock
Thanks. I think I will start again. I tried all the near combos and failed. Took the back off to see if I was close. Off by a country mile, in every direction. will revisit this again.

The combination was 3-38-25.

Image

Combination now changed Im going back to the drawing board on this one.

Re: Chubb combination lock fail

PostPosted: Fri Mar 24, 2017 9:03 am
by MartinHewitt
But the 23.5 was then about right.

Re: Chubb combination lock fail

PostPosted: Fri Mar 24, 2017 9:41 am
by ratlock
Yep but it went down hill from there. Ive reset the combo, got some 0-100 graph paper and will try again later when waiting for Daughter if daylight holds out.

Sargent& Greenleaf 1 Ratlock 0