Dave J. said:
If you don't have the "inside dope", it'd be easy to think that Fisher was shipping great stuff from Los Banos and when they went on the auction block, First Texas bought a great company and proceeded to screw everything up. But that is not what happened.
When FTP bought Fisher, Fisher was dying a horrible death. I saw the factory, I saw the books. Same building, mostly the same people, but a completely different company from the one I'd left in 1995. Manufacturing was in total chaos, sales on everything were falling like a rock, profits were a negative million dollars a year, and they were selling some products below manufacturing cost just to make it look like they still had a business. I told the boss that I may still love Fisher, but as a business matter he shouldn't pay a dime for it. Fortunately he didn't listen to me on that one, and here we are nearly 10 years later shipping more "Old Fisher" product than the "Old Fisher" was shipping.
Part of what killed the "Old Fisher", was that under the Roger Cimino regime there was an effort to reduce manufacturing costs by cutting stuff out without first understanding why things were the way they were. In the case of the CZ's, they replaced the original labor-intensive calibration procedure I wrote, with a Cimino procedure to make it real easy to pass one final step provided that you didn't then exercise the machine to do everything it was supposed to do. In short, it was a fake calibration that fooled only the person doing the calibration, not the customer. The result was enormous variation, with no machines actually doing what they should. The situation created an aftermarket calibration business for Dankowski, who'd been sufficiently involved with the CZ3D to know how it worked and how to calibrate it.
When FTP bought what was left of Fisher, they didn't tell us that the correct calibration procedure had been scrapped and replaced with a Cimino procedure. When the operation was moved to El Paso, the procedure that people were taught was the Los Banos procedure. Since machines were passing that test, we were puzzled why so many problems were being reported by customers. Finally I got dragged into it, discovered that the calibration procedure being used was completely wrong, and plowed through the archives looking for a copy of the original procedure. Fortunately it was there. We resurrected it, made a couple minor modifications to it, made sure that all copies of the Cimino procedure were destroyed, and deployed the correct procedure.
Even with the correct procedure in hand, the CZ is a difficult machine to tweak up. If you're inexperienced or in a hurry, the result is a sloppy tweak (although still far superior to a Cimino tweak). Things got a lot better overnight, and with experience and with emphasis on not being in a hurry, fairly soon complaints of lousy tweak just about vanished.
--Dave J.
