C16J filament phasing

The phasing of the filament voltage with respect to the plate voltage is significant in prolonging the life of the tube. This post presents a method of properly setting that phase relationship in the event that the transformer wiring has been changed. Some explanatory notes appear at the end.

The pinout diagram on the datasheet makes this relationship obvious:



The intent is for anode current to flow through the F- pin. If the phasing is wrong, anode current will flow through the F+ pin, possibly causing excessive heating and premature failure.

WARNING: Hazardous (potentially lethal) voltages exist in this system. If you're not familiar with proper troubleshooting techniques and safety precautions for working on high-voltage control systems, do not attempt to do so.

This procedure requires use of a dual-channel triggered oscilloscope with two 10x divider probes. The probes should be rated for 600 volts peak input. Almost any oscilloscope can be used, since the signal is only 60 Hz.

Because of the high voltage present on the C16J anode during operation, a special 10:1 voltage divider must be constructed for use between the oscilloscope probe and the tube, as shown in the diagram:



The following instructions assume a basic knowledge of oscilloscope operation.

1) Turn off all power to the lathe.
2) Set the 'scope as follows:
....Channel 1: 10 volts per division, AC coupled
....Channel 2: 0.5 volts per division, AC coupled
....Vertical mode: two-channel chopped
....Trigger: channel 1, positive slope, automatic mode
....Sweep speed: 5 milliseconds per division
3) Connect the divider network and 'scope probes as shown.
4) Set the lathe speed control to zero.
5) After confirming all correct connections, turn on power to the lathe.
6) Engage the controls for forward operation.
7) Adjust the vertical position controls so you see two waveforms, one above the other.
8) The waveforms should be in phase, i.e. the rising edges of the two should be inline.
9) If one waveform is inverted relative to the other, reverse the filament transformer connections.

Power the lathe down.

Repeat the above tests for the other C16J.

Notes:
1) The 470K resistors are spec'd at 5 watts due to voltage ratings. Lower power resistors have lower voltage ratings.
2) Chassis is used as the oscilloscope common connection for safety reasons. The probe shield on most 'scopes connects directly to AC power line safety ground. The lathe chassis is connected to control circuit common (line GA2) via a 0.1 mfd capacitor. This connection is adequate for our purposes.
3) The phase of F+ is measured WRT filament center tap (GA2) since this permits use of chassis for the channel 2 ground connection. That phasing is identical to the phasing across the entire winding (F+ to F-).

- Leigh

The real Leigh said:

The intent is for anode current to flow through the F- pin. If the phasing is wrong, anode current will flow through the F+ pin, possibly causing excessive heating and premature failure.

Click to expand...

I think that should read Cathode current, the anode (Plate) is P and only has one place to go, the filament and the cathode are common items, but this prevent tube current from flowing through the filament.
M.

Nope, Leigh got it right.

Brad

How so? you cannot redirect the Anode current, I am aware it flows from filament (cathode) to anode (plate), but I assume the exercise is to ensure that tube current does not flow through the filament of the cathode itself?
I am not saying it is wrong, it is just the terminology I questioned.
M.

Since the tube has no cathode, all of the current flows from the filament to the anode. This is of course driven by the anode to cathode voltage. But there is another voltage involved - the filament voltage. One side of the filament has a greater potential difference with the anode than the other - this is the point. One side of the filament will conduct much more than the other due to the greater potential difference.

Bruce Griffing said:

But there is another voltage involved - the filament voltage. One side of the filament has a greater potential difference with the anode than the other - this is the point. One side of the filament will conduct much more than the other due to the greater potential difference.

Click to expand...

Bruce is absolutely correct.

The voltage gradient across the filament causes an uneven current distribution. The end of the filament which corresponds to the highest differential WRT the anode will conduct the highest current.

This phenomenon has been well-known for over 80 years. It caused a problem with early vacuum tubes in battery-powered radios. The tube emission would drop over time. Reversing the polarity of the filament battery connections would increase emission by concentrating current in the unused portion of the filament.

I expressed the concept in terms of anode current to distinguish it from filament current which also flows through the F- pin. The magnitude of the two are almost the same, which means that the F- pin carries almost twice the current (at peak load) compared with the F+ pin. In any tube circuit (absent a screen grid or similar structure), anode (plate) current = cathode current.

In a directly-heated cathode design such as the C16J, the cathode and the filament share the same physical structure, but they perform two distinct and unrelated functions.

- Leigh

Novice question, don't be too hard on me ;<)

Are the C16J's in phase with one another? I would have thought not. i.e. I expected them to be conducting alternately.

However, the F- & F+ terminals are common between the two tubes, so if th C16J's are not in phase with each other, they can't both be in phase with their F-/F+ terminals.

-Dave

"However, the F- & F+ terminals are common between the two tubes, so if th C16J's are not in phase with each other, they can't both be in phase with their F-/F+ terminals."

On WiaDs, with a single filament transformer, that is true, although an accidental mis-wire could place these two at a difference.

The Modular has two independent filament transformers.

In both cases, however the center tap of the filament transformer/transformers is the +'ve terminal of the armature regulator.

The C16Js really can't be out of phase with each other solely by flipping the filament leads. This out of phase condition would only be possible by a bizarre miswire of the grid transformer, which IS common on both drive types.

Peter, the c16J's would be in phase with each other, if correctly wired? What about the 3C23's? Mine are 180 deg. apart (Plate to chassis ground).

-Dave

rimcanyon said:

Are the C16J's in phase with one another? I would have thought not. i.e. I expected them to be conducting alternately.

Click to expand...

Hi Dave,

They should be 180° out of phase, but that doesn't mean they are

The schematics I've looked at show the plate and control grid fed from opposite phases, by virtue of center-tapped transformer secondaries.

However, the filaments are a different story. Some schematics show a separate filament transformer for each tube. But there is no polarity identification on the wiring between the tube and its transformer, so no way to tell if they were phased properly or not.

There are ways to accomplish phasing that are not necessarily identifiable on the schematic, such as using different length leads for the two connections.

I have to go by the way the tube was designed to work. Whether or not Monarch took that into account when they designed the control system I cannot answer.

Thanks.

- Leigh

I could reverse the posts (F+/F-) that C16J #1 & #2 connect to, so that each would individually match the phase of the current flowing through the anode. In that respect the WiaD has as much flexibility as the modular, so C16J tube life would be maximized.

The 3C23's are not marked as to F+/F-, so are they designed in such a way that tube life is affected by incorrect polarity of the two F connections?

Regarding the 3C23's, I found one wiring problem that explains most of the strange results I was seeing(the grid connections were reversed for the two 3C23's). I also found a couple of bad solder joints on the 10K resistor on one side. After fixing that, the 3C23's still did not balance. Then I reversed the connections to A24/A25 from the anode transformer, and now both 3C23's have equal amplitudes, nice sinusoidal waves, 180 deg out of phase, no ripples. The motor is still not running, but I think I am ready to move on towards the field loss relay circuit.

Thanks for the help so far.

-Dave

Hi Dave,

I suspect the filament phasing on the 3C23 is not nearly as significant as for the C16J, due to the much lower anode and filament currents.

Sounds like you're making progress. That's great.

Good luck with it.

- Leigh

Module Machine C16J Phasing

On my machine one tube glow faintly without the spindle engaged. When the spindle is engaged the other tube lights up fully and the one that glowed goes dark. The lathe runs at 1500 RPM with the module removed. Any thoughts on this?

steelz said:

On my machine one tube glow faintly without the spindle engaged. When the spindle is engaged the other tube lights up fully and the one that glowed goes dark. The lathe runs at 1500 RPM with the module removed. Any thoughts on this?

Click to expand...

Yes, start a new thread...

After following all this, I have a question. Is the filament asymmetrical? If not, phasing probably wouldn't matter. In fact, periodically reversing the phase might increase life. Actually, I suspect that it must be heavier on the minus end. Either that, or possibly there is an electrode attached to that end that the discharge migrates to. If you have a filament rated at 31 amps and put an additional 18 through it, don't expect it to be around long. Additionally, if one end of the filament has more influence relative to the grid, a phase reversal can have a major influence on the firing point. In the old battery powered radios, this effect was used to produce a pseudo grid bias. That could be the reason for unbalanced firing in some cases. Two thyratrons phased differently are not likely to match. I only have a schematic for modular units, but a defect is the lack of balancing control, which would be easy to provide.

Bill

9100, according to the spec sheets for the tubes, the phasing of the filaments definitely matters. That is why one is marked F+ and the other is marked F-.

Leigh, your method of phasing the filaments will certainly work, but it is far more complicated than it needs to be. No oscilloscope is needed at all, just a half decent voltmeter, preferably digital. One of the less than $5.00 model from Harbor freight will work quite nicely, although something higher quality such as a fluke will certainly be acceptable :-)

The way I phased the filaments was quite simple, and only took a few minutes. Here goes my procedure:
1- Turn the lathe on, put the drive in its active condition, but leave the spindle control in the stop position. Setting of the speed pot does not matter.

2- Put your meter on the AC scale, I'll assume it's auto ranging(if not, use lowest scale that doesn't go over range, you'll be measuring around 300V), attach one of its leads to the plate(anode) connector on the tube.

3- Take the other lead from your meter and touch it to each of the two field terminals, noting the voltage reading at each. You will notice that the voltage reading at one filament terminal is a few volts higher than at the other filament terminal. The filament terminal that gives you the higher reading should be the negative terminal(F-), the one that gives you the lower reading should be the positive terminal(+). Switch leads if it's reversed.

4- That's all there is to it, now go and do the other tube.

Ed

The Doctor said:

9100, according to the spec sheets for the tubes, the phasing of the filaments definitely matters. That is why one is marked F+ and the other is marked F-.

Leigh, your method of phasing the filaments will certainly work, but it is far more complicated than it needs to be. No oscilloscope is needed at all, just a half decent voltmeter, preferably digital. One of the less than $5.00 model from Harbor freight will work quite nicely, although something higher quality such as a fluke will certainly be acceptable :-)

The way I phased the filaments was quite simple, and only took a few minutes. Here goes my procedure:
1- Turn the lathe on, put the drive in its active condition, but leave the spindle control in the stop position. Setting of the speed pot does not matter.

2- Put your meter on the AC scale, I'll assume it's auto ranging(if not, use lowest scale that doesn't go over range, you'll be measuring around 300V), attach one of its leads to the plate(anode) connector on the tube.

3- Take the other lead from your meter and touch it to each of the two field terminals, noting the voltage reading at each. You will notice that the voltage reading at one filament terminal is a few volts higher than at the other filament terminal. The filament terminal that gives you the higher reading should be the negative terminal(F-), the one that gives you the lower reading should be the positive terminal(+). Switch leads if it's reversed.

4- That's all there is to it, now go and do the other tube.

Ed

Click to expand...

I apologize to drag up an old post, but I am curious if this method will work to testing the filament phasing. It seems it would, I assume you are testing the AC voltage offset that has resulted from the superimposed DC voltage. The higher voltage should be on the F- as it is designated to be the current carrying conductor. Is my understanding correct?

Right. The RCA manual for the C16J says that you want the F- terminal to be negative during conduction. Since the tube is only conducting on half cycles, the anode will be positive and the filament will be farther away voltage wise at that moment. Because the filament is really only an electrode after firing begins, there may be a metal flag extending above the filament on that side. Once the gas breaks down, the discharge will go to the nearest point at the most negative voltage. I used to wonder why mercury vapor rectifiers and the old Tungar tubes had such low filament voltages. The answer is that the lower the voltage, the higher the current for a given filament area. The higher the filament current, the less percentage increase during conduction from the anode. The heating effect can be so great in some magnetrons that when it is up and running, they turn off the filament voltage and let anode current heat it.

Bill

Thanks for the confirmation Bill. I was going through the more of my wiring schematics, and my understanding is that the F- filament should be at 300 VDC, and I am curious if the center-tap (wire 22) at the T-5 transformer would be -300 VDC. If this is the case wouldn't current flow from the filament F- through the T-6 and T-7 center-taps through wire 51, across the armature A2-A1 then through the Series field S2-S1. I am thinking that this is incorrect, because I can't understand why Monarch would put the CB on the return. I am really trying to learn the drive as I would like to maintain it, and I also think it is cool as all hell.

Also, I was planning on replacing my C-5 and C-6 capacitors and the R-5 and R-6 resistors in the grid circuit.

I was planning on ordering these, but I want to make sure:

150102J630BB Cornell Dubilier | Mouser

and

PR02000202403JR500 Vishay / BC Components | Mouser

What do you think?

Alright well the capacitors and resistors came in today. Wish me luck