Open in 86 minutes. But first to the previous postings here.
Oldfast wrote in
viewtopic.php?p=111812#p111812:
This is incorrect. And is a very common belief that I myself hung onto for a long time! LaGard wheels are in fact
fairly concentric. Actually, they're about as 'round' as any other wheel type in any other lock. Extreme fluctuations
exhibited by LaGards (and Ilco's P67 lock) are a direct result of what's known as 'wheel float'. Wheel float exists, to
some extent, in nearly every lock. But LaGard has it to such a degree that it tends to throw most of us for a real loop.
From the graphs I showed here, I think he is very correct.
My strategy for this opening was:
1) Test and write down the right contact point for all possible combinations of w1=0,35,70; w2=10,45,80; w3=20,55,90 and take the combination with the best contact point, or one of the best contact points. My reasoning was that if a wheel is off center I catch a low point with one of three equally distributed points. So even with wild floating wheels I do get a good location for further optimization with just 27 tests. The numbers on the three wheels are shifted by 10, so that they can be dialed without confusion. For actual testing the order can be optimized to reduce work. I start with 0/80/90 which is just 0, then 2.2 turns right and 1.1 turn left. Back to CP then forward to 90 and push w3 further to 20, then test and push to 55 and test. Return to w2@80 and push to 45, then w3@55, 90 and 20. And so on.
2) Optimize each wheel until gates are shown. This avoids identification of wheels and most of all any problem with wheel float when rotation directions are changed.
With this lock and combination I get in step 1 the best CP at 0/10/20 with 8 0/8. (Took me 20 minutes. I am not a quick dialer.) First scan was then L0/R10/Lx. This did show a gate at 40. I did there only halve of a scan, because it was so pronounced. Second scan was then L0/Rx/L40. This did show a strange behaviour at 40 - perhaps a gate - and a low area around 0 to 22 at 7 7/8. I took 20 as a safe and easy way for optimization. Next was Lx/20/40. This many jumps and fluctuation. I think part of the problem was the mounting of the dial. There were some points at 7 7/8 and a bit larger area around 2 to 7. I chose 5 also as a safe number. I did a rescan of w2 with L5/Rx/L40 starting at 45 for w2. The lock did open after 86 minutes at L5/R40/--. From examining the open lock the real combination was probably 6/40/41 or so. It might have been intentional as I made her a complete random number sheet or it was a failure setting the combination.
For this opening it was 1/2 scan of w3, 1 full scan of w2 and 1 full scan of w1 plus the 27 initial combination tests. As I wrote I am not a quick dialer. I should probably reread Oldfast's/altashot's description of how they dial and try to understand it.