Herring Hall Marvin Manipulation

[gzasada42]

My sister and I have had a fun summer project of attempting to manipulate this old herring hall Marvin safe that was found in the wall of a speak-easy, we are really looking forward to finding a century old paperclip. We have probably dumped 30 hours each into this project. We borrowed the Guide to Manipulation book and also found a similar safe online and asked for pictures of the lock. The learning curve has been steep but we have had just an absolute blast trying to crack this old thing, but now we need some guidance from the experts.

Summary of what we think we know:
L - R - L - R(drive wheel)
#1-70 wheel
3 digit combo
On the last right turn the safe will stop automatically
1 Contact point felt clockwise (60) during our test of the wheel (L2.5 - R60, L5 - R60)
When the stack is turned clockwise, resistance at 41/42
Contact point is estimated to be 6 numbers wide and to the left (?) of where we feel the contact point
Spring Fence lock, but not a yale HE, lock manufacturer ID difficult to determine

Method:
We charted the data initially for every 2.5 numbers and then going back to the contact point logged where the needle fell. We went back and then charted every number in one sitting and came up with a few possibilities of what could be a point on a wheel. The book is using a yale HE as an example so we figured everything is reversed: we chose the numbers that shaped a gate that erred larger verses smaller. This might be incorrect as the book chose numbers that created a gate shape that passed the contact point and were lesser.

We tried to isolate the wheel that this number might be on and didn't have great results or anything consistent. Ex: 42 was a number we chose based on our graph so we attempted to isolate the wheel via 52-42-42/ 42-52-42/ 42-42-52 then did the same thing with minus 10. We spent quit a bit of time on this and just couldn't get anything conclusive. The book says the displacement should be the greatest. I do not know if that should be more indicative if it is less than 60 or greater than 60-- could someone elaborate?

Anyway we tried the other numbers that could be potentials from our chart and just really struggled with the isolating the wheels stage. A youtube video took 3-4 points from the initial chart and those were actually numbers on the combination. Maybe we are working too hard?

The actual safe

A similar safe online with the following pictures




2 Minds might be worse than 1...

[MartinHewitt]

I am impressed, you have found a lot of correct answers. So your 30 hours invested time was not wasted. The summary of what you think what you know is mostly correct. You err only on the lock identification, which has consequences on the CP. The lock was made by HHM. It is not a spring fence lock, but a friction fence lock. The fence in a friction fence lock is moved by the gear and will make a CP only in opening direction, which you have correctly identified as right. So only a right CP, no left CP. I can't tell you where it should be located approximately. The National Locksmith Guide to Manipulation (if this is what you have) has Chapter 5 about friction fence locks. The lock they have as an example has a final left turn to open, so it is reversed to your lock. Maybe this information will help you to reevaluate what you have done so far.

The "52-42-42/ 42-52-42/ 42-42-52" is not wheel isolation, it is called the Hi/Lo test and is done to locate the wheel of a found gate when you have moved multiple wheels at the same time. Wheel isolation is the name for when you move only one wheel for a graph and leave all others at a fixed position.

"Maybe we are working too hard?" No. Persistence is key.

[Kaesekopf]

Summary of what we think we know:
L - R - L - R(drive wheel)
#1-70 wheel
3 digit combo
On the last right turn the safe will stop automatically
1 Contact point felt clockwise (60) during our test of the wheel (L2.5 - R60, L5 - R60)
When the stack is turned clockwise, resistance at 41/42
Contact point is estimated to be 6 numbers wide and to the left (?) of where we feel the contact point
Spring Fence lock, but not a yale HE, lock manufacturer ID difficult to determine

I'm not quite understanding the 6 numbers wide part, you're trying to estimate how wide the contact area is? If as Martin says it's a friction fence lock with 1 contact point, I wouldn't worry about it. Just record the contact point you do have access to, and work with that.

Also, that resistance at 42 may be a number, friction fence locks can wear on the cam/fence surface, giving access to other numbers

Method:
We charted the data initially for every 2.5 numbers and then going back to the contact point logged where the needle fell. We went back and then charted every number in one sitting and came up with a few possibilities of what could be a point on a wheel. The book is using a yale HE as an example so we figured everything is reversed: we chose the numbers that shaped a gate that erred larger verses smaller. This might be incorrect as the book chose numbers that created a gate shape that passed the contact point and were lesser.

Because you're feeling a contact point when turning clockwise, the numbers that record as higher on the dial at that contact point will result in "lower" readings, meaning the fence is able to drop in slightly further into the wheelpack.

Question: what precision are you reading the contact point at? I read in 10ths, e.g. 60.3 60.5 etc.

We tried to isolate the wheel that this number might be on and didn't have great results or anything consistent. Ex: 42 was a number we chose based on our graph so we attempted to isolate the wheel via 52-42-42/ 42-52-42/ 42-42-52 then did the same thing with minus 10. We spent quit a bit of time on this and just couldn't get anything conclusive. The book says the displacement should be the greatest. I do not know if that should be more indicative if it is less than 60 or greater than 60-- could someone elaborate?

If you're feeling the contact point when turning the dial clockwise (right) the readings at 60 indicate "better" readings. 42 being promising makes sense, and I would expect it to be wheel 1 from experience. NOT definitively though

In your test, because you're moving 1 wheel away from the test area at a time, you are looking for the reading that becomes worse. In other words, when you move away the wheel that the gate is on, the reading should immediately turn bad. This tells you which wheel to leave there from now on.

Anyway we tried the other numbers that could be potentials from our chart and just really struggled with the isolating the wheels stage. A youtube video took 3-4 points from the initial chart and those were actually numbers on the combination. Maybe we are working too hard?

Certainly not working too hard, unless you were expecting payment Also, graphs? would love to see pictures, or spreadsheets since that is what you seemed to be using. It's much easier to assist if there is data to look at.

If you are able, it would be worthwhile to make sure this lock does not have a 4th wheel. To do this, turn dial right to 25 around 5 times (to be safe). Then, turn back left and count how many times you pick up a wheel around 25.

-A safecracking Cheesehead

[billdeserthills]

Is it just me or am I seeing a possible entry hole in picture #1?
The hole is just outside the dial near the top

[Kaesekopf]

I think you’re right. Top left

[billdeserthills]

I think you’re right. Top left

Now you can learn how to transfer the gates around till the fence drops in!

Actually after looking at your inside pic, that hole may just be right near the fence now...

[bitbuster]

Removable dial yet the hole was put outside the dial .