Ferroresonant transformer issue

[OperaHouse]

A friend built a plasma pip cutting machine, pretty impressive device that is over 5 tons and can cut a 26 foot pipe up
to 30 inches in diameter into sections with V cuts ready for welding. I gave him a Torque Systems ferro DC power supply
that had been in storage for 30 years to run the stepper motors. The machine gets quite a bit of use and has been in
operation for over three years (those discussing capacitor storage life should take note of this). I get a call that the
drive modules are indicating over voltage fault. This is not surprising since the supply normally runs about 84VDC and
the drive is rated for 80V with the spec saying the maximum allowable voltage is 85V. I don't know at what voltage the
actual fault triggers at but it must be close. He explained that once in a while he would get an OV fault, but this time
all three drives showed OV for half a day. Shop voltage can go quite high at times. His half brother who just graduated
EE designed the system. Impressive software but the drivers are on the edge. They burn out a driver on average every year.
power supply is split so it could be run at 42V.

Anyway, that's the background. I'm leaving town for a while and thought I would just throw in a buck transformer to reduce
the line voltage and maybe that would drop the DC just a tiny bit. Then I get another phone call saying everything is running
fine again. The new information is that when he was getting the OV fault the DC was at 95V instead of 84. My understanding
of ferro regulators is that the voltage couldn't jump that much, just a couple percent. Also that these are capable of +-40%
input voltage. Find it hard to believe the resonant cap is failing like that.

Second question. If I add another small cap to the resonant cap could I reduce the voltage to 80V by shifting the resonant
frequency? Has anyone tried this? Seems like a more elegant solution than adding a couple 25A bridge rectifiers in series
for a voltage drop.

 

[Comcokid]

In the old ferro-resonant DC supplies, the resonant capacitor is a AC capacitor the same as an AC motor run capacitor. The output is followed by a full-wave rectifier and large electrolytic capacitor. My experience has been that the output voltage accuracy is rated at a constant load. One old ferro-resonant I use is 24 volt output, but always varies by load from 26 to about 30 VDC.

You didn't say what the current was, but for simplicity, I would just add a few diodes and heatsinks on the output and just use diode drop to lower the voltage some.

 

[OperaHouse]

Yesterday I added a buck transformer that dropped the 125V AC to 105V. As expected, it didn't drop the DC voltage of
84.7V, I was surprised that it didn't even drop it a tenth of a volt. It was mostly an experiment and to show I was
at least trying something. I dug through the scrap pile and found a ferro 12V supply and measured the secondary as 15V
AC. To simulate an open capacitor I removed a wire from the resonant cap. The voltage dropped to 11V. That indicates
that an open cap would not cause the voltage to rise to 95V and that adding a parallel cap to detune the transformer
might reduce the voltage to the desired 80V. This problem may be imaginary. When the stepper motor drives shut down
the unloaded power supply may naturally rise to 95V. I tried this with the 12V ferro supply and just a 1.5A load. That
supply dropped 1.4V or about 10%. That makes it a relay race, can the drives draw enough current before the voltage
sensing kicks in. I'msupplying two bridge diodes in series to drop the voltage about 3V to my friend for the next time
this happens.

 

[mikekilroy]

TAKE A DEEP BREATH, relax, and step back and look at the macro picture.

You have a ferro xfmr that will try to keep its secondary voltage THE SAME over a large input voltage swing... why then try to put a buck boost in and lower the input when the ferro will try to make it the same?

WHAT is the pri and sec of the ferro xfmr? 120 in and 120 CT out? If you cant live with its output then DUMP it. put in a regular xmfr.

Or put your buck boost on THE SECONDARY OF THE FERRO?

Or by a $ 100 DB regen circuit to dump the peaks.

Or....

http://www.KilroyWasHere<dot>com

 

[OperaHouse]

I couldn't be more relaxed, I'm a thousand miles away and on vacation for four months. As I stated the reason for the buck transformer was to maybe drop the voltage just a few tenths of a volt to get it away from the 85V module shut down voltage. The other hope was it would drop the maximum voltage when the resonant capacitor opened up and save a module. Later tests indicated an open capacitor actual lowers the voltage and that the load/no load voltage swing is the problem. The diodes should take care of that.