- Reaction score
- 13,478
We have this Swiss-made Dixi horizontal borer from 1957, very nice machine still with its original DC variable voltage drive. In those days there were no VFDs, variable speed machine drives all used DC motors or Schrage AC motors, with the best DC units capable of better performance than VFDs at low speeds; this one goes all the way down to 70 rpm at 100% torque. In 1957 there were really two options for producing the variable voltage to feed the machine spindle motor itself: A motor generator set that runs at constant speed with variable excitation on the generator (the Ward-Leonard system) or a thyratron (gas-filled valve) controlled rectifier bridge. Our borer has a Ward-Leonard drive, with a 2900 rpm converter set sitting next to it with a 3-phase motor, DC generator, exciter generator and a cross-field servo-amplifier machine all on one shaft. This is in turn under the control of a magnetic transductor amplifier which gives closed-loop speed control.
All of this has been working pretty reliably for 65 years but yesterday it wouldn't stop working when told to. The spindle motor would stop and start but when the machinist went to stop the converter set, as soon as he took his finger off the button it fired back up. The E-stop button isn't supposed to stop the converter: it stops the spindle and the hydraulic pump, but leaves the converter running because that is needed to dynamically brake the spindle. It only stopped when he turned off the isolator.
The 3-phase motor of the Ward-Leonard converter has a star-delta starter within the panel which is more or less conventional, just with a few interconnections to tie it into the machine controls, e.g. an auxiliary proving contact on the delta contactor enables the DC spindle controls once up to speed, and the converter overload relay trips the spindle drive first, which then trips the converter off. I expected there would be a short somewhere in the control interlocking but the reason was simple and obvious once the panel was opened: the line contactor was jamming closed mechanically. With a finger on the converter stop button the holding circuit was broken, releasing the delta contactor and de-energising the motor. But when the button was released the circuit was still present through the main contactor holding contact which pulled both it and the delta contactor back in.
The star-delta starter is the group of contactors at bottom right of the panel. Ignore the arrow, that's the transductor.
Opening the contactor revealed a broken part of the moving contact carriage. The mystery is how long it has been broken. It's a strong phenolic moulding that was showing wear on the cracked surfaces. I suspect it was damaged when the panel was moved 14 years ago, and the parts have been rubbing together since, but just yesterday the broken bit moved into a different position and jammed the contactor. These are the original German-made Klöckner-Moeller DIL3 contactors, from the days before they were DIN-rail mounted. Vertical operation, gravity return, exposed auxiliary contacts. I had to decide between replacing or repairing the contactor. For reasons explained below, replacing it would be a little complicated, so I chose the repair option. I have some very high performance epoxy resin handy so I carefully glued the parts together and they have held very strongly. The contactor is back together and now starts and stops the machine as intended.
I took the complete starter sub-assembly out of the panel to repair:
One interesting feature of the starter is the thermal star-delta changeover timer, which uses an expanding hot-wire element instead of an element coiled around a bimetal strip. This keeps the thermal mass low so that it cools and resets very quickly. Once an auxiliary contact on the delta opens, it resets ready for another start in one or two seconds. The wire is short and thick and requires only a very low voltage to heat it because its resistance is only around 170 milliohms. The control system of the machine is 36V AC, so the hot-wire is energised not from the 36V but from a secondary winding on the line contactor. Over the ordinary 36V winding you can see the secondary of 16 turns which develops about half a volt at 3A, fed via the delta aux into the hot wire of the timer. To avoid having to alter this and replace the timer, had I chosen the replacement route I would have replaced another contactor with a modern one, and used the liberated contact assembly to replace this one and retained the original line contactor.
Star-delta timer: The hot wire is strung vertically below the spring. which keeps it under tension As it expands the spring shortens, pulling up the projection on the grey lever. The bottom end of the lever moves to the left pulling the paxolin lost-motion link with it, and once the lost motion is absorbed, the tail end of brass contact arm which opens the contact and releases the star contactor, allowing the delta to close. Time is set by adjusting the eccentric pivot of the contact arm, which varies the amount of expansion needed to take up the lost motion. Pic of coil shows secondary winding (visible turns) over primary (under wrap.)
One day (!) I will do a walk-through video of how this machine works. There can't be many still in use that have not been updated to VFD, or at least electronic VVDC.
All of this has been working pretty reliably for 65 years but yesterday it wouldn't stop working when told to. The spindle motor would stop and start but when the machinist went to stop the converter set, as soon as he took his finger off the button it fired back up. The E-stop button isn't supposed to stop the converter: it stops the spindle and the hydraulic pump, but leaves the converter running because that is needed to dynamically brake the spindle. It only stopped when he turned off the isolator.
The 3-phase motor of the Ward-Leonard converter has a star-delta starter within the panel which is more or less conventional, just with a few interconnections to tie it into the machine controls, e.g. an auxiliary proving contact on the delta contactor enables the DC spindle controls once up to speed, and the converter overload relay trips the spindle drive first, which then trips the converter off. I expected there would be a short somewhere in the control interlocking but the reason was simple and obvious once the panel was opened: the line contactor was jamming closed mechanically. With a finger on the converter stop button the holding circuit was broken, releasing the delta contactor and de-energising the motor. But when the button was released the circuit was still present through the main contactor holding contact which pulled both it and the delta contactor back in.
The star-delta starter is the group of contactors at bottom right of the panel. Ignore the arrow, that's the transductor.
Opening the contactor revealed a broken part of the moving contact carriage. The mystery is how long it has been broken. It's a strong phenolic moulding that was showing wear on the cracked surfaces. I suspect it was damaged when the panel was moved 14 years ago, and the parts have been rubbing together since, but just yesterday the broken bit moved into a different position and jammed the contactor. These are the original German-made Klöckner-Moeller DIL3 contactors, from the days before they were DIN-rail mounted. Vertical operation, gravity return, exposed auxiliary contacts. I had to decide between replacing or repairing the contactor. For reasons explained below, replacing it would be a little complicated, so I chose the repair option. I have some very high performance epoxy resin handy so I carefully glued the parts together and they have held very strongly. The contactor is back together and now starts and stops the machine as intended.
I took the complete starter sub-assembly out of the panel to repair:
One interesting feature of the starter is the thermal star-delta changeover timer, which uses an expanding hot-wire element instead of an element coiled around a bimetal strip. This keeps the thermal mass low so that it cools and resets very quickly. Once an auxiliary contact on the delta opens, it resets ready for another start in one or two seconds. The wire is short and thick and requires only a very low voltage to heat it because its resistance is only around 170 milliohms. The control system of the machine is 36V AC, so the hot-wire is energised not from the 36V but from a secondary winding on the line contactor. Over the ordinary 36V winding you can see the secondary of 16 turns which develops about half a volt at 3A, fed via the delta aux into the hot wire of the timer. To avoid having to alter this and replace the timer, had I chosen the replacement route I would have replaced another contactor with a modern one, and used the liberated contact assembly to replace this one and retained the original line contactor.
Star-delta timer: The hot wire is strung vertically below the spring. which keeps it under tension As it expands the spring shortens, pulling up the projection on the grey lever. The bottom end of the lever moves to the left pulling the paxolin lost-motion link with it, and once the lost motion is absorbed, the tail end of brass contact arm which opens the contact and releases the star contactor, allowing the delta to close. Time is set by adjusting the eccentric pivot of the contact arm, which varies the amount of expansion needed to take up the lost motion. Pic of coil shows secondary winding (visible turns) over primary (under wrap.)
One day (!) I will do a walk-through video of how this machine works. There can't be many still in use that have not been updated to VFD, or at least electronic VVDC.
Last edited:
