Is this a burnt vacuum tube? Bridgeport Servo ecf series 3109

[JST]

I doubt there are any more components. There are enough to do the job as it is (including the pot, which is elsewhere). and there is really no room for any more under it, that PWB is right down close to the cast panel.

As for the white smoke, we do not know anything about the pot (control). It might have fried when the neon bulb failed, or another part, such as the large power resistor, or a capacitor, may have failed.

What we know is that the rapid works, so the motor is OK. But that probably bypasses the control, possibly by turning the SCR on full. Something in the variable speed has failed, but not the SCR, and not anything that would prevent the SCR from turning on (because of the rapid).

The pot, the diode, and the large resistor are possibilities, along with the neon bulb (an NE2, most likely, as Jim Rozen suggests). Those are in the speed setting portion of the control, which we are fairly sure is not working, per the OP.

 

[sfriedberg]

I am reaching back over four decades in memory here, trying to recall stuff that was obsolete technology even then. The mention of "variable speed" was the trigger (literally). I bet a nickel that, if you drew out the circuit, the neon tube is being used in a stone-simple RC relaxation oscillator, and that oscillator is what triggers that SCR or triac at a variable rate.
There is a 2nd possibility. Super cheap (like the one we're looking at in the photo) motor controls could use an RC divider to trigger an SCR/triac on every AC cycle, but at varying delays into the phase. Longer delay means less power to the motor.
But if that glass bulb is a neon tube, I'll stick with my nickel bet.

 

[JST]

I'm going with your possibility #2.

They are actually almost the same. Your #2 would look just the same as #1, except the timing is such that the trigger occurs within one half cycle. That syncs it to the line frequency, where a free running relaxation oscillator would trigger the SCR at random points. The pot controls the charge-up time, and that controls where in the half cycle the NE2 fires.

Diacs and neon lamps have a characteristic that they will not "fire" (conduct) until a certain voltage across them. But when they do "fire", the voltage drops to a low voltage, which stays low until the current drops low enough, when they stop conducting.

The SCR needs a certain minimum current into the trigger terminal in order to cause it to conduct, after which it stays conducting until the current drops as the AC reverses polarity. The NE2 in this case does the triggering, discharging a small capacitor which delivers a pulse of current that can trigger the SCR.

The pot and resistor setup changes the time required to charge the capacitor to a voltage which will "fire" the NE2 (or diac, in more modern types). That time is set up to vary so that the SCR is triggered at some point during the half cycle. Where it is triggered sets the average voltage, and so the speed of the motor.

It's old-school, with the major disadvantage being that the NE2 bulb is being abused a bit and may not last. They are not really intended to carry more than a few mA at most. The firing voltage for the neon tube is set (in at least certain types) by a small amount of radioactive material on the electrodes or in the gas filling, which pre-ionizes the neon and lowers the breakdown voltage. In those, the voltage eventually rises as the radioactivity is depleted.

 

[Milland]

Speaking as a guy who likes to work materials rather than electrons, I think this stuff verges on voodoo, but it's still interesting to read about the clever design kluges. Interesting thread...

 

[jim rozen]

Is a silicon univeral switch like a diac? Motivated by this thread, went back to figure out this hardinge feed motor drive:

 

[JST]

It is a bit like one, yes. But the breakover voltage is adjustable, so it is a three lead device. The "gate" lead is the voltage setter, generally by a voltage divider, as in the schematic you show. It only works in one direction, like an SCR, which it has internally.

A diac is similar, but has a fixed trigger voltage, has only two leads, and has the same breakover voltage in either direction.

 

[sfriedberg]

Is a silicon univeral switch like a diac?

It's a member of the thyrister/SCR family. I suppose the diac could be considered a two-terminal member of that family. But the three-terminal devices (like SUS) have explicit trigger, while diac is functionally more like a bidirectional Zener.

 

[JST]

They are really both "zener type" action, in that there is a voltage that they break over at. Then they fire like an SCR (because there is one inside)

IIRC, there were SUS' that were settable only.... no inherent breakover voltage (you always set it), but others, including the one in the schematic, have an inherent voltage, but can be adjusted to lower voltages by use of the gate.

 

[jim rozen]

I'm trying to understand the circuit where they're feeding the speed control pot via D2 to the bottom of the armature. I'm wondering if there's a feedback from the armature back emf to provide some sort of torque boost at lower speeds. The original HLVH feed motor control had a variac mechanically coupled to a power pot. As the armature voltage (variac) was reduced, the field current was changed by the power pot - I forget if it was weakened or increased.

 

[JST]

The back EMF is a good measurement of the motor speed. The motor has a generating constant at any field setting, and voltage is related to the constant x rpm. Many schemes to measure it have been invented for motor control.

Polarity of back EMF is same as the incoming voltage that it opposes, plus opposes plus. And it is available after the SCR cuts off at the end of each half cycle, until the SCR fires again.

So, the higher the back EMF, the less voltage is applied to the pots to charge up the capacitor and fire the SUS to turn on the SCR. The back EMF on the lower motor terminal will be negative relative to the rectified "positive" voltage, more negative as the motor runs faster. That delays firing of the SCR and applies less voltage to the motor, slowing it.

The pot adjusts the speed to a balance of off and on time, which is actually then based on "measured" motor speed. So if the motor speeds up, back EMF is higher, voltage on the pot is lower, the SCR firing is delayed, and the motor slows. And, if the motor slows, back EMF is less, so the pots get more voltage, and firing of the SCR is earlier, speeding up the motor.

Relatively few components make this happen. Really quite elegant.

Since the adjustment to the SUS firing voltage is dependent on the incoming voltage (R1 and R2), there is a measure of compensation for voltage variations as well.

Edit: I should add that the diode keeps the trigger system from discharging during the time the SCR is "on", when the voltage input to the pot drops to near zero. Without it, the pot could discharge the capacitor at a rate depending on how long the motor is "on".

 

[jim rozen]

Interesting analysis, thank you. This was reverse engineered from a replacement setup a buddy purchased for his HLVH. He originally had the small cylindrical potted module that hardinge had during one period, which tend to go bad, which his did. They shipped him a bundle of parts but no real instructions, and when he assembled it, at least one of the resistors fried as they had stuffed the board incorrectly. They sent him a replacment gratis, which was correctly wired. But not before I reverse engineered the broken one. I think the value for R1 and/or R2 was incorrect. The "S" on the diagram means "Smoke."

 

[JST]

R1 seems possibly low in resistance, for two reasons. If it is a half watt resistor, it will dissipate a little over half it's rating at nominal voltage for a 240V AC input, higher than the normal de-rating. It's still OK (less that 1/2 watt) even at maximum mains voltage, though. Of course if it was stuffed as a 1/4 watt, that would easily explain the problem.

The other reason is that the divider ratio goes over the SUS' internal zener voltage near mains peaks. That may be intended, however.

 

[jim rozen]

I seem to recall that divider string was stuffed with the wrong values. And wrong in a 'sensible' way - somebody goofed up on the color codes.

 

[DaveKamp]

NE2's were actually great gas discharge tubes... they'd fire off very predictably at what... 82 volts or so. Using them as a 'pilot light' was actually more of a secondary function than a primary.

 

[JST]

They were used a lot as voltage limiters in tube electronics, to ensure a voltage never went too high before the tubes warmed up.

They work great....... until they don't. Not sure if they are the type with some radioactive material in them to set the voltage, but they eventually fail.

They stop lighting up reliably, usually blinking, or not lighting at all unless there is a spike voltage, etc.

 

[sfriedberg]

I ran into a weirdo neon bulb behavior a few months ago. I have an extension cord with a neon bulb molded into the receptacle end, presumably to let you know power is available (i.e., other end is plugged in to a live socket). However, the bulb doesn't work in the dark. (Really!) I spent 5 minutes flipping the shop lights on and off, and it was very reliable: Turn the lights on, the bulb in the cord (laying on the floor) turns on. Turn the lights off, the bulb turns off.
I figured it was not a well-made bulb, and needed just a leeetle bit more stimulation to fire, and that the shop photons were enough to push it over the edge. In the dark, you could see it sort of stuttering a dim glow. In the light, it was definitely full one.
And before some wiseacre asks, no, the shop overhead light switch is not connected to the circuit the extension cord was plugged into.

 

[JST]

Yep, I've seen that before.

If it was one that uses a radioactive material to set the firing voltage, it may no longer have enough activity to reach the critical ionization at lower voltages, and does not fire reliably at 120VAC.

The light may be just enough to push it over the edge to fire at a lower voltage.

 

[john.k]

I have a house light ,mini fluoro i think ,it wont light unless the TV is on ,and even then ,it lights after about 10 minutes when the program on the telly emits a bright screen......yes its doubtless Chinese .

 

[jim rozen]

Discharge lamps that only fire on when illuminated by external light are showing the photoelectric effect.

Photoelectric effect - Wikipedia

en.wikipedia.org

 

[Scottl]

Could be the trenule pipe. Check your differential girdle springs.

More likely an issue with the Fetzer valve.