The Old Timer of Royal Oak, Mich., was a regular contributor to H&P years before we ever even heard of the internet. But most of his advice is just as useful — and interesting — today.
So rather than leave his wisdom printed on pages archived in our storage room, I pulled out issues from the late 1980s and early 1990s and have been reproducing relevant entries in this blog. Here is my 23rd entry, which was originally published in the November 1989 issue:
Some years ago, our engineering department sent down a design for a light-duty transfer system using a 6-ft pneumatic cable cylinder operated by a 3-position directional control valve with blocked center for jogging. The design looked fine on paper but turned into a first-class disaster in production after assembly. A major contributor to our problem was the fact that our shop had standardized on valves with metal-to-metal seals — which were relatively new at that time — because our ancient pneumatic distribution system was full of rust, compressor oil, and other debris. These valves digested contaminants fairly well, but to do so, they inherently left some leakage paths open.
When we first turned the air on for the transfer and energized solenoid b, the piston accelerated suddenly to the right, and the transfer bar slammed into the left-side end stops, breaking them off. Because there was no pressure air in the right end of the cylinder, the flow control had no effect on speed.
We had our electrical people interlock the solenoids so that the last one energized stayed on until the other was energized. This took away our jogging capability, but it also ensured that there was always air pressure in the cylinder to resist the next stroke. Things worked fine until lunch time, when the electricity was shut off for some minor adjustments. Springs centered the valve spool, and enough air leaked down from both ends through the blocked center so that when we restarted and energized solenoid a, the part on the transfer was launched about 40 ft across the plant floor, and another wreck occurred.
A quick call to engineering reported these incidents, and within 24 hours we had some new valves. “If blocked center won’t work, let’s use Y center to put line pressure on both ends,” they said. This new valve did work — for a few weeks — until the transfer wore in and loosened up a bit. Then the unit tended to drift off the end stops and limits, especially if the transfer wasn’t level or the laod was uneven. Attempts to jog an inch or two wound up moving a foot or two — or three. (I believe the debris inside our air system was a major contributor to this erratic performance.)
Our production people demanded that we fix this locally, without waiting for another guess from engineering. We came up with our own circuit using a dual-pressure center in the main valve and upstream check valves. Everyone was gun shy about the transfer by now, especially when the action proved very shaky on the first startup with this new circuit. But after a couple of cycles, things smoothed down and operation became pretty solid. We rigged things up so the unit did a few dry cycles before they started loading parts, and this worked well. We had tamed the beast, but no one really trusted it. They still tended to step away from the ends of the equipment during the first couple of loaded transfers.
Since those early days, spool valve designs have improved dramatically, and we’ve learned a lot about air preparation. I guess the main lesson of this reminiscence is that you have to consider the system — rather than a single component — during both design and troubleshooting.