Bringing my WiaD back to life

I'm not completely done bringing my 57 WiaD lathe back to life, but progress has been encouraging, and while the facts are still fresh in my mind I wanted to give at least a partial recount since it may be of use to others who need to debug their DC drive.

I purchased the '57 2-3 years ago, and it showed signs of a complete rebuild and mechanically it was in fine shape (missing a few components that walked while it was sitting). The WiaD on the other hand was a mess. The wiring had disintegrated, many components had shorted and were bad, and it clearly had not run in a long time.

The lathe came from Hughes aircraft, so presumably it had regular maintenance until it developed electrical problems. It appears not to have been running for at least 10 years.

I decided to add ELSR and VSR and to completely rewire the WiaD. I also obtained a set of 220V transformers, and converted the lathe from 440V to 220V. I was fortunate enough to locate the correct wiring diagram and schematic for a WiaD lathe with ELSR & VSR, so the task was a matter of noting the differences between the existing wiring and the new wiring, and of obtaining a few components required by the conversion that I did not have. I purchased a 120V coil for the main contactor, since the control circuitry used for ELSR does not use line current, and a 120V relay for switching the speed potentiometers, used by the VSR. Otherwise I had most of the components needed, cannibalized from various parts lathes.

Adding ELSR meant running a conduit from the on/off switch box to the tailstock area, through the coolant pump cavity, then running wires from ELSR and VSR (which is mounted on the plate at the tail end of the bed), to the WiAD, to the main contactor, to the DC panel. All those wires cross at the On/Off switch box.

Rewiring the WiaD gave me a chance to test all of the components on the board that I would normally not be able to isolate easily, so I found some bad resistors, and replaced the High Speed pot and the compensation pot. The grid transformers were in bad shape so I replaced them with a pair that came from another machine.

Both anode transformers are dual voltage so they were wired for 220V.

When that work was done, the machine sat for almost a year, until two weeks ago when I finally got enough courage to turn on the power and start debugging. Fortunately, nothing spectacular happened when the power did go on, but I ran into my first debugging problem.

The Field Loss relay would not stay engaged when the on button was pressed.

I measured the field (F1-F2), and there was none. That was symptomatic of a bad 6SF5 tube. So I replaced the tube. I also replaced both 3C23's and the EL1C with new tubes.

Now I actually had DC field voltage: 58V to be precise. Since I had just rewired the entire WiaD, I followed every path shown in the schematic for the field, found one loose connection not related, and everything else checked out. Then it occurred to me, perhaps the grid transformer that I had just replaced needed to be in-phase with the anode transformer. OK, I'm no EE, so who you gonna call? Yep, an email to Peter elicited a confirmation on that point, along with the information that the F2 and E1 center tap connections needed to be on the yellow and orange wires from the filament transformer, and the DD and CC connections needed to be on the red and green wires respectively. Turns out that the Monarch circuit diagram had an error. OK, so I fixed that and reversed the polarity of the grid transformer for the field.

Now my field voltage is respectable: 98VDC. I actually got a 110VDC field at one point (I had the F2 and E1 connections reversed and the grid transformer reversed from its current connection), but now it is wired according to schematic. I think I need a scope to verify that the relative phase of the two transformers is actually correct, unless there is another way...(??)

Next problem: the field voltage does not vary as the controls are adjusted.

I haven't solved this one yet. This is what I know: after disconnecting both speed pots and the FA relay coil and the F -25- R path, there is only one device connecting terminal 13 to terminal 15: the high speed pot. I have verified that the resistance varies from 0 to 27K ohms as the high speed pot is adjusted. I measured that right at the connector base for the 6SF5. However, the 6SF5 tube is not doing its job. Field voltage remains constant at 95V, despite the setting of the high speed pot.

Tonight I will try another 6SF5 tube. I'll also record the voltage drop across the 51K resistor connected to the 6SF5 tube. And I will check the resistance between F2 and the grid transformer secondary, which should be infinite.

"Now my field voltage is respectable: 98VDC. I actually got a 110VDC field at one point (I had the F2 and E1 connections reversed and the grid transformer reversed from its current connection), but now it is wired according to schematic. I think I need a scope to verify that the relative phase of the two transformers is actually correct, unless there is another way...(??)"

I don't have a Monarch, so I can't speak from knowledge of the schematic, but there may be a way to measure the relative phase without a scope. I'll give you a simple example that may or may not work for your case. Suppose you had two transformer secondaries, each connected to ground on one side. If they were in phase, you would measure the difference between the two to be simply the difference in their outputs. So if one was 100V and the other was 50V, you would measure 50V between them. But if they were out of phase, you would measure 150V.

Bruce, I added the field schematic (for a WiaD without VSR) above.

Is there a general convention that the primary and secondary wires of a transformer would be physically parallel where they exit the winding? e.g. if one primary wire comes in on the left side, another on the right side, and on the opposite side, there is a secondary on the left and right, can you assume that the primary and secondary will be in phase from left to right?

"Is there a general convention that the primary and secondary wires of a transformer would be physically parallel where they exit the winding? e.g. if one primary wire comes in on the left side, another on the right side, and on the opposite side, there is a secondary on the left and right, can you assume that the primary and secondary will be in phase from left to right?"

You would think so.

The transformer in question is an audio interstage transformer, intended for public address or other "low-fi" applications, and for these applications there is no need to maintain phase.

OTOH, if someone was making a "stereo pair" of amps, then polarity should be maintained from input to output, and that would certainly include any intermediate transformers, such as an interstage transformer.

The way the Monarch assembly drawing was made, it is clear that the two grid transformers are intended to be installed as a more-or-less matched pair, and that the physical positions of the wires exiting these open frame transformers are the indication if the intended, and required polarity.

And this works, too, provided the two transformers are identical, which one can presume they would be should they come from the same production lot.

Otherwise, there is no convention for such transformers, excepting those intended for "hi-fi" applications, in which case a dot may be placed on the specification sheet, which dot indicates the "start" of the winding.

From the part of the overall schematic you posted, I can conclude a few things. First, you are correct that the phase is important. Second, my proposed method will not work without "bench testing" the parts - ie removing their connections to the circuit and testing them by themselves. You could determine the phase that way, and mark the devices, but I suspect that borrowing a scope would be easier. As to your question about lead location being a clue to phase - I don't really know, but I would not count on anything other than direct measurements myself. Transformers in schematics I see are marked with dots for each winding to indicate the relative phase. Usually, phase is controlled by numbering the leads sequentially. Then if you follow the numbering system in the schematic, you automatically get the correct phase. No such luck here.

rimcanyon said:

...can you assume that the primary and secondary will be in phase from left to right?

Click to expand...

No.

That may be true for some transformers, but not for all.

When transformer windings are specifically phased, this is indicated by a dot at one connection of each winding. The voltage at those dots is in phase.

- Leigh

I've also been speculating that the filament transformer may need to be in phase with the others, due to the relative locations of the F2 and E1 taps. Does this make sense?

Thats a lot to get right, considering how much got replaced here.

However, it is corroborated by some of the strange results I have seen. e.g. swapping E1 and F2 caused the voltage to be 110V.

rimcanyon said:

I've also been speculating that the filament transformer may need to be in phase with the others, due to the relative locations of the F2 and E1 taps. Does this make sense?
However, it is corroborated by some of the strange results I have seen. e.g. swapping E1 and F2 caused the voltage to be 110V.

Click to expand...

The 6-prefix tubes (6H6, 6N7, etc.) have indirectly-heated cathodes which should be totally insensitive to filament voltage phase.

The C16J's are thyratrons with directly-heated cathodes, but they should not be phase-sensitive. Current flows through those tubes from the center-tap of the filament transformer.

The two 3C23's __might__ be phase sensitive. This would relate to the connections at A25 and 73. There's a transformer primary there; its secondary drives the grid of the 6SF5 at circuit point 13. Interconnecting the primary connections of this transformer to points A25 and 73 __might__ affect circuit operation.

One general comment:
The WiaD uses mercury rectifiers. These MUST heat up before you apply voltage to them. Attempting to run the unit before they're up to operating temperature can cause internal arcing which can cause major damage to other components. The unit incorporates a 60-second normally-open time-delay relay (TDR type 6NO60) to enforce this warm-up period. If yours is inoperative, either replace it or make sure you allow adequate warm-up before using the machine.

The TDR normally-open contacts are in series with the START switch, between circuit points B1 and B2. A fault in this relay could prevent your machine from running.

One possible point of concern:
This unit uses a 6X5 rectifier. These tubes are known to have a high failure rate, specifically a heater-cathode short. This will put high voltage on the heater winding. Although the 6-volt heater winding appears to be isolated from other circuitry throughout, it could impress a high voltage on other tube filaments which might cause them to fail.

The most reliable 6X5's are called the "X-plate" type. If you look down from the top, you'll see the two plates at right angles to each other, one above the other. This is a later design, possibly created because of the high failure of earlier styles.

Any of the online tube vendors ( www.vacuumtubesinc.com et al) will understand that description.

At some point you want to check the two speed control pots. These are inside the machine, above the motor, attached to the control knob at the lower left. These are very weird pots. Each changes value through only half of its rotation.

The operator control spans about two and a half turns. Starting with it fully counter-clockwise (zero RPM), the rear 15K pot increases from zero to 15K in the first half of the control range, and remains at that value for the rest of the rotation. The front 10K pot stays at zero for the first half, then increases to max over the remaining adjustment.

Hope this helps.

- Leigh

"The C16J's are thyratrons with directly-heated cathodes, but they should not be phase-sensitive. Current flows through those tubes from the center-tap of the filament transformer."

Thyratrons have an F+ terminal and an F- terminal.

The source might not be phased, and it certainly makes no difference in the dc voltage rectified by the tube, but the tube life is greater when the F+ terminal is at a higher voltage than the F- terminal.

Which is why these terminals are so labeled in the first place.

I would go farther than Leigh - I do think the phase of the grid drive transformer is critical to proper operation. Reversing the 73 A25 connection would accomplish reversing the grid drive phase. Since I don't know what the 73 A25 connections go to, I cannot be sure, but in almost any design you could imagine, you would want to drive the 3C23's on alternate half cycles - but only those with the plate at the correct potential. This is what I meant in the earlier post when I said that the phase is important. I should have been more specific.

peterh5322 said:

Thyratrons have an F+ terminal and an F- terminal.
The source might not be phased, and it certainly makes no difference in the dc voltage rectified by the tube, but the tube life is greater when the F+ terminal is at a higher voltage than the F- terminal.

Click to expand...

Hi Peter,

Yes, I'm aware of the base labeling. I'm the one who posted the datasheets last time we discussed these.

So, given that the filaments are fed with AC from a center-tapped winding, how does one go about putting the F+ terminal "at a higher voltage" than the F- terminal?

- Leigh

OK, real progress.

Swapped A24 and A25. No difference. Put a new 6SF5 in. Still no regulation, but the voltages increased across the 51K resistors, etc. Swapped the grid transformer secondaries again. The pot started regulating the field. Almost had a runaway motor! Called it quits for the night, and for the next week (going on vacation).

Thanks for the help Leigh, Bruce and Peter! I'll start on the armature circuit when I return. I'll be sure to check the 6X5's. I have tested, good C16J's, but I plan on replacing all the small tubes in the armature circuit.

"... how does one go about putting the F+ terminal 'at a higher voltage' than the F- terminal"

The problem is how does one determine this during the conduction portion of the tube, where it actually matters.

Using a triggered-sweep oscilloscope with an isolation-type transformer is my thought.

That means either a Tektronix 485 (and $9,000, when they were last made) or an HP 1725A (about one-half of TEK's price for an equivalent bandwidth, 300 MHz, and the HP already has an iso-type transformer in its power supply, too, whereas the TEK doesn't).

Perhaps there are "modern" 'scopes which can assist in this job.

For otherwise, we have to trust to Monarch's assemblers, and hope that they phased the F+ and F- correctly.

peterh5322 said:

"... how does one go about putting the F+ terminal 'at a higher voltage' than the F- terminal"
The problem is how does one determine this during the conduction portion of the tube, where it actually matters.
For otherwise, we have to trust to Monarch's assemblers, and hope that they phased the F+ and F- correctly.

Click to expand...

Hi Peter,

But the relative polarity (the F+ pin WRT the F- pin) changes 120 times per second.

- Leigh

"But the relative polarity (the F+ pin WRT the F- pin) changes 120 times per second"

True, but it is of significance only when the tube is conducting, hence the need for a triggered-sweep o-scope, triggering on the anode voltage (ground to C.T.), and two probes, one measuring F+ (ground to C.T.) and the other measuring F- (ground to C.T.), or a single second probe, measuring a candidate for F+ (ground to C.T.).

Since the trigger is a high voltage, the probe should be designed for this service.

If there was a thytratron plugged in, then the probes connected to F+ and F- would also need to be high voltage, too.

Leigh-
As I imagne this design (in the absence of a complete schematic) each thyratron conducts on only one half cycle. So I think Peter's point is that you need to decide the polarity based on which half cycle the tube is conducting. For that half cycle, you connect the more positive side of the transformer to the "+" connection.
Bruce

Bruce Griffing said:

As I imagne this design (in the absence of a complete schematic) each thyratron conducts on only one half cycle. So I think Peter's point is that you need to decide the polarity based on which half cycle the tube is conducting. For that half cycle, you connect the more positive side of the transformer to the "+" connection.

Click to expand...

Hi Bruce,

I suppose that's possible.

I don't know why there are so many different WiaD schematics. I have at least three versions here. One includes circuitry for the LSR, while an earlier version does not, so that explains these two. Then there's the diagram that was pasted inside the electrical unit cover. Oh well. They all seem to have the same information, but in different formats. Then there's the wiring diagram.

I would be reluctant to probe around inside this thing with any standard o'scope probes. Back in "the day", standard Tektronix and HP probes were rated 600 volts peak. With the advent of solid-state electronics, I believe modern probes have lower ratings.

Even 600 volts would not be high enough for some of the voltages present in these control systems. I would build a separate 10x divider, with a high-voltage 900K resistor to the test point in series with a 100K resistor to ground. Connect a standard 10x probe across the 100K resistor and you should be safe. You'd only introduce a 1% error in the readings, which is not significant.

Thanks.

- Leigh

C16J filament polarity

Perhaps I've found an explanation for the polarity issue.

I found the following comment on the C16J datasheet:

"... circuit returns to filament transformer center tap, filament lead F- negative with respect to filament lead F+ during conduction period..."

The schematic diagram for the tube is drawn rather oddly, with the F- filament connection at bottom center, and the F+ filament terminal at the upper left, around the 10:00 position. The obvious implication is that the F- lead is designed to carry the full anode current (18 amps max) in addition to the 30-amp filament current.

The F- pin is larger diameter, and its connecting lead is heavier, as compared with the F+ pin. If the filament leads are improperly phased, the F+ pin might overheat to the point of causing a pre-mature failure.

I'm going to start a separate thread on the subject of filament phasing, in the hope that it may become a sticky or be included in a FAQ for the WiaD.

Peter was correct, as usual.

Thanks.

- Leigh

The 33 amps at 2.5 volts is equally divided between the two leads.

The 18 amps at 230 volts may be unequally divided, but both F+ and F- carry the pulsating dc voltage which is ultimately delivered to the armature through the F and R contactors.

I suppose the designers of the C16J (and the other CxxJ) tubes at Electrons Inc made an economic trade-off when they designed these tubes, and the F+ internal structure was made heavier than the F- internal structure.

Whatever the reason, the F+ internal structure, and its external lead takes more of the load than does the F- internal structure and its external lead.

Hi Peter,

It's actually the F- lead that carries the heavier current.

I just posted a new thread on filament phasing. http://www.practicalmachinist.com/vb/showthread.php?p=858582#post858582

I'd be interested in your comments.

Thanks.

- Leigh