Cath-O-Drive

[KD1SH]

  Anyone have any connection info for Kenyon's T-472 "Cath-O-Drive" modulation transformers? Kenyon made a series of mod transformers specifically designed for cathode modulation, from the T-471 to the T-474, and from what I understand, each would have come with a printed spec sheet, much as the UTC types did, but mine didn't have the sheet with it, and web searches have turned up nothing other than Kenyon catalogs. These are universal types, with twelve terminals to allow for a wide range of secondary impedances, typically lower, of course, than a plate-mod transformer would have. Some exploration with an ohmmeter would probably reveal much, but I'd love to have the original spec sheet and connection info.

[KB1VWC]

  Same here Bill. I have a T-471 that I am also looking for the info on. Thought it would be fun to experiment with. Hopefully, a data sheet, is in an attic somewhere.

  Steve

[KD6VXI]

Maybe this will help?

 

Cathode Modulation Transformers
Model Number     Max Secondary DC (mA)     Audio Tubes
 T-471                                   200                             Single 6F6
 T-472                300    PP           6V6s or 2A3s
T-473               450   PP           6L6's (AB1 or AB2)
T-474               500   Universal     (120 watts audio)



--Shane
WP2ASS / ex KD6VXI

 

[KD1SH]

  Thanks. It doesn't quite get me where I need to go, though. What I'm looking for is the connection information to accomplish the various secondary impedances.
  The primary appears to have several taps: two marked "P" and two others marked "P/" (the slash might be a theta symbol—the marks are worn) with a center tap common to both, marked "B". The "B" tap, I assume, would be for the modulator tube B+. The primary is designed for, as the info you provided suggests, either 6V6 or 2A3 tubes, so I'm thinking that the two "P" windings are for the plates of either tube, depending on which tube you use.
  The secondary seems have two separate, isolated windings, one with four terminals and the other with three.
The first winding, terminals marked 1, 2, 3, 4, reads five ohms from 1 to 2; 10 ohms from 1 to 3; and 15 ohms from 1 to 4. The other secondary, terminals 5, 6, 7, reads 37 ohms between terminals 5 and 7, with 6 appearing to divide the two halves pretty much equally, as a center tap.
  I'm going to compare this with other Kenyon universal types, to see if there's some sort of common scheme.

Maybe this will help?

 

Cathode Modulation Transformers
Model Number     Max Secondary DC (mA)     Audio Tubes
 T-471                                   200                             Single 6F6
 T-472                300    PP           6V6s or 2A3s
T-473               450   PP           6L6's (AB1 or AB2)
T-474               500   Universal     (120 watts audio)



--Shane
WP2ASS / ex KD6VXI

 

[KD6VXI]

Trust me, I tried like hell to find it.

It wasn't 'published' in any of their catalogs.  You had to write to them or buy a transformer.  And even buying one sometimes meant writing or calling to find that info.

Sucks.

--Shane
WP2ASS / ex KD6VXI

[KD1SH]

Yeah, me too. Well, someday, when I'm not looking for it, it'll show up at a hamfest somewhere.

Trust me, I tried like hell to find it.

It wasn't 'published' in any of their catalogs.  You had to write to them or buy a transformer.  And even buying one sometimes meant writing or calling to find that info.

Sucks.

--Shane
WP2ASS / ex KD6VXI

[K8DI]

Do you own or have access to an LCR meter?  The primary is easier to figure out, you know the tube type,  look up it’s plate to plate load, put a resistor between the P terminals of that value. Then use the LCR meter on R and measure the secondary taps, you’ll get the values you seek. Use a setting like 1kHz/middle of the audio band. The value at 60Hz or 15kHz isn’t relevant.

Ed

[w8khk]

Ed, that is a good suggestion for determining the expected secondary impedance.

Another option might be to apply a known audio signal level at 1 KHz on the primary, then measure the voltage across the various pairs of pins on the secondary.  From this information, the turns ratio may be determined, and from that the impedance ratio is easily calculated.

The secondary may be more than one winding, not unlike the multi-ratio modulation transformers.   An ohmmeter will reveal which are separate windings, and the relative resistance readings will give an idea as to which sections have more or less turns. 

Knowing the relative voltage across the individual sections of each secondary winding makes it rather easy to achieve the desired output impedance for any project.  Sections with equal voltage measurements may be connected in series or parallel, as desired.

When finished, perhaps a chart could be provided as a courtesy to those also searching for this data.  Just the relative voltage measurements would save others a great deal of time and effort.

[KD1SH]

   Thanks, Rick and Ed, for excellent suggestions on both counts. I do indeed have an LCR meter—the DE-5000—and I'm currently experimenting with using it to test audio transformers, which I haven't done much of.
   Some interesting results: When I use my LCR to measure the inductance of RF coils, as in a tank circuit, measuring across half the coil results in half the inductance that I'd read across the entire coil—no surprise there—but that's not the result I get when using the LCR to measure the inductance of the Kenyon's windings.
   On the primary winding of the Kenyon, for example, I read 484mh from the center tap to either end of the winding, but from end to end I read 1927mh, where I would expect to read 968mh. Looking at simple DC resistance, it's obvious that the center tap really is exactly that: half the total resistance from the tap to either end. What's up with this? Is it an accident that the total inductance is roughly 4x that from the center tap to either end, rather than 2X?
   By the way, I'm using the LCR's 1khz setting, and it reads an 8-ohm speaker right on the money.
   I'm going to gather up a bunch of my audio transformers and experiment a bit, to get a better handle on this, since I'm very likely missing something obvious. Sometimes even a blind nut finds a squirrel (or maybe it's the other way around).

[w1zb]

I picked up a UTC CM-15 and CM-16 cathode modulation transformers a while ago and ran into the same problem of no data sheet. I've seen a couple of old threads on the topic on this site which might help. You might also look at the Frank Jones Cathode Modulation Handbook for more info and possible a diagram of a cathode modulator. If I remember correctly the secondary impedance required is typically 1K - 2K.

73 Jerry W1ZB

[w8khk]

Bill, I believe your reactance test instrument is honestly telling the true story here.  Since the DC resistance of a transformer winding is so small compared to the reactance, it can for the most part be ignored when looking at reactance and impedance.  That said, the reactance will closely follow the impedance (at a given frequency).  But as the frequency changes, the reactance will follow the formula Xl =2 Pi F L.  As frequency is increased the reactance will increase.

The turns ratio of a transformer is the square root of the impedance ratio, thus the impedance ratio is the square of the turns ratio.  This converse relationship applies to the reactance ratio as well.

Take for instance an audio output transformer, that has taps for four, eight, and sixteen ohms.  If you apply a voltage to the primary that provides two volts from the start of the voice coil winding to the four ohm tap, you will also see that from the start to the sixteen ohm tap will be four volts, not eight.  The four ohm tap is the center tap between the start of the winding and the sixteen ohm tap. The voltage distribution is equal on either side of the center tap but the impedance of the entire winding is four times the value of half of the winding.

When studying the primary impedance matches for various single-ended and push-pull configurations, doubling the number of turns provides four times the plate-to-plate impedance.

Another example to consider is the case of the modulation transformer secondary impedance matched to the DC impedance of the final amplifier.   Take for instance a legal limit final running 2000 volts DC at 500 mA, for 1 KW DC input.  The impedance of the load  (the final amplifier) is E/I, or 2000 / .5 = 4000 ohms.   The impedance is "matched" at carrier level with no modulation.  But with 100 percent positive modulation peaks, the voltage is doubled.  Now we have 4000 volts, and as a result of the higher voltage, the final now draws 1000 mA.  The effective impedance of the final is still 4000 ohms, but the output power is quadrupled with just a doubling of the DC voltage.  

Now if we were matching a speaker voice coil to an output transformer, for a given power level from the amplifier, we would want the speaker to produce a power level relative to the amplifier power.  Two volts RMS into a 4 ohm speaker produces E squared /R, or 2*2/4 = 1 watt.  If the 16 ohm output is connected to a 4 ohm speaker, we have 4 volts, so 4*4/4 = 4 watts, four times the power to the same speaker with just a doubling of the voltage.  But connected properly, a 16 ohm speaker connected to the 16 ohm tap provides 4 volts, and 4*4/16 = 1 watt.  Obviously double the voltage, or double the turns, into 4 times the impedance, provides the same power transferred.

So it is the square law that applies to both the impedance and the reactance when the turns are doubled, assuming the frequency of measurement is a constant.  Measuring at a different frequency will of course yield a different reactance, but the difference between 1/2 and full winding should still follow the same square relationship.

[KD1SH]

  Yes, I picked up a copy of Frank Jones's book at the Chicopee 'fest last spring; it's what peaked my interest in cathode modulation.
  If you're still looking for the data sheets for the UTC transformers, you're in luck—I've got the CM-15 and CM-16, too, and here's the sheet, scanned into PDF:

I picked up a UTC CM-15 and CM-16 cathode modulation transformers a while ago and ran into the same problem of no data sheet. I've seen a couple of old threads on the topic on this site which might help. You might also look at the Frank Jones Cathode Modulation Handbook for more info and possible a diagram of a cathode modulator. If I remember correctly the secondary impedance required is typically 1K - 2K.

73 Jerry W1ZB

[KB1VWC]

  Had some time this AM, so I took the T471 off the shelf and did some checks.

  This is a single ended transformer with similar a primary to yours Bill.  It has B and P terminals. Two isolated secondaries. 1-4 and 5-7.  Measured DC resistance. Fed 10V@ 1Khz into the B-P terminals and then measured the output voltages.  Calculated the turns ratios, and then the impedance ratio's.  Since this is supposed to be a single 6F6 primary, an assumption of 7000 ohms primary impedance was made.
   Doesn't really help as you could cross connect various terminal to give yourself different impedances. I haven't looked at the Frank Jones paper in quite some time.  Don't they recommend some amount of G1 modulation as well?  Maybe that's why Kenyons have the isolated secondaries? Or it's just the typical multimatch mod iron? Excuse the chicken scratch.


Steve

[KD1SH]

  Great stuff, Rick. Sometimes it seems like the most basic fundamental stuff, especially when we haven't cogitated on it for a while, is what trips us up.
   So, drawing out your hypothetical output transformer, I'm confused, and most certainly misconstruing the whole thing as you describe it. The secondary has four terminals: one is the common, or start of the winding, and then there are terminals for four, eight, and sixteen ohms. With these four terminals, how does any one of them wind up being the center, with equal voltages on either side?

Take for instance an audio output transformer, that has taps for four, eight, and sixteen ohms.  If you apply a voltage to the primary that provides two volts from the start of the voice coil winding to the four ohm tap, you will also see that from the start to the sixteen ohm tap will be four volts, not eight.  The four ohm tap is the center tap between the start of the winding and the sixteen ohm tap. The voltage distribution is equal on either side of the center tap but the impedance of the entire winding is four times the value of half of the winding.

[w8khk]

Bill, we probably spend more time calculating the impedance ratio of modulation transformers, from primary to secondary, than we do calculating the relative impedance of each section of a secondary, such as the voice coil example I provided.  But the technique is the same; the voltage ratio, IE the turns ratio, is simply the square root of the impedance ratio.

For a three-section secondary with four terminal points, from start to point 2 is 4 ohms, start to point 3 is 8 ohms, and start to point 4 is 16 ohms, typical of most audio output transformers.

The ratio of 4 ohms to 16 ohms is 4/16,  1/4, or 0.25; the square root of 0.25 is 0.5, half the voltage of the entire secondary, obviously the center tap!

The ratio of 8 ohms to 16 ohms is 8/16, 1/2, or 0.5; the square root of 0.5 is 0.70710678...  just over 70% of the voltage across the entire secondary.

Say we generate a signal that measures 10 volts RMS across the entire winding of 16 ohms.  We will measure 5 volts from the start to the 4 ohm tap, half the 10 volts.   We will measure 7.07 volts from start to the 8 ohm tap, or .70710678 times the 10 volts across the entire secondary.  Turns ratio and voltage ratios are identical.

For your cathode modulator transformer project, decide on what tube or tubes you will use as modulators, either single-ended or push pull.   Determine what primary impedance will work best for the modulator tubes, based upon plate voltage, plate current, and class of amplifier to be used.
 
Next, apply a reference voltage to the full primary, and measure the various voltages across each section of every secondary winding.  Depending upon transformer design, the secondaries may be connected in series;  or if there are identical voltages on more than one secondary, they may be connected in parallel to handle more current.  Based upon all the options for various output voltages, you may then define a fraction, consisting of output voltage divided by input voltage, to determine the turns ratio, then square that value to determine the impedance ratio.  

Knowing your target source impedance (defined earlier in the modulator design phase), you now have a list of output impedances from which to choose.   Taken even further, you can determine, based upon estimated audio power from the modulator tube(s), the output voltage swing and watts available to cathode modulate your final amplifier.  Rather simple, but tedious if there are many secondary windings and taps.  But the tedium has a bright side - more options to home in on the optimum impedance ratio for dense, clean modulation.

[KD1SH]

  Yes, Rick, fine business. I fooled myself with my visual image of the transformer. If it had a 2-ohm winding between the common terminal and the 4-ohm terminal, it would have appeared "balanced" to my confused mind. Dumb mistake.
  Turns out I had an output transformer exactly as you described, and a few minutes with my signal generator yielded exactly the numbers you predicted.
  Tomorrow I'll play around with the Kenyon and my function generator, following your suggestions, and see if I can diagram some ratios.

[w1zb]

TNX for the UTC data sheet. Big help.

You may want to look at the article from Dec 1939 RADIO Magazine (page 16) about cathode modulation design and adjustment. I found it a good read.

73 Jerry W1ZB

[KD1SH]

  Okay, I changed my plans a bit, and instead of characterizing the Kenyon, I've been taking a bunch of measurements on my UTC CM-16, since I've got the connection sheet for it, and as an exercise, comparing my predicted values with those stated on the sheet. Great fun, and a great learning experience. Sometimes, when we spend much of our time working through the more complicated stuff, it's embarrassing to be reminded of how we've gotten lazy with the simple stuff.
  Using my signal generator to excite the primary with a known voltage and comparing the voltage across the chosen secondary, I can find the turns ratio far better than I could with an ohm-meter, and from there the math to get the impedance ratio is easy. Once I've gotten the impedance ratio, I can then connect my decade box to the primary and, using my LCR, verify that the impedance ratio transforms the selected value into what appears on the secondary.
  Now, armed with my refreshed knowledge, I can proceed to characterize the Kenyon. This is a thoroughly enjoyable rabbit hole to go down.
  Thanks, Rick, for the encouragement.

[w8khk]

Bill, I am glad you found the measurement suggestions helpful.  The voltage method is so much better than trying to determine the impedance ratio with an ohmmeter.  DC resistance measurements can give you an idea as to which windings have more or fewer turns, but it is certainly not accurate.  First, you don't know if the same wire gauge was used for all the windings, and second, windings closer to the core have less wire length per turn, hence less resistance.  This is why a center-tapped winding may have more resistance on one side of center tap than the other.  Resistance measurements are best for the initial identification of individual windings and tap positions, but not much more.  In my opinion, the voltage method will be more accurate than the LCR meter looking at the impedance reflected by a known load on the other side.  However it is an interesting exercise, and will give you an idea how your instrument behaves when that is the only practical measurement you can perform.

Now that you have the impedance transformation measurements under control, may I throw you a couple more curve balls?   Oops, I am not really a baseball fan, so maybe I should just guide you down a couple more rabbit holes.....   If you wish communications quality audio, you may stop reading now, but if you want your cathode modulated rig to sound as good as when I heard you on 75 meters a couple days ago, please read on.....

The initial impedance transformation measurements are "unloaded", essentially "out of circuit", and do not characterize the audio performance of the transformer in your intended application.  A couple more tests will reveal an idea as to what you might expect when your project is finished.  I suggest that after you decide on an initial winding configuration, you feed the audio generator signal into the primary, with a series resistor approximately equal to the plate-to-plate load impedance of your modulator tubes.  This will essentially cut the input voltage in half, and allow you to see how the transformer input impedance varies with frequency.  During this test, also place a load resistor on the secondary, equal in ohms to the expected transformed impedance.  Sweep from your lowest to your highest desired frequency, and look at both input and output voltages, in parallel with the transformer windings, while occasionally verifying the generator output level is stable.  The level at the transformer input will likely be lower as you go lower in frequency, due to the limited inductance on the low end.  Using as many turns as possible for the primary will provide the most inductance, minimizing low frequency attenuation. 

Also, using an amplifier plate impedance lower than the transformer input impedance will minimize the roll-off, while another way around this is to add inverse feedback.  If, at a given frequency, the input level is constant but the output level is reduced, that effect may be lessened by trying a larger load resistor.  If this in fact improves the frequency response, then selecting a lower transformer output impedance than that of the planned load will also improve frequency response.  In general, these two methods of tuning the response can be accomplished if the transformer is oversized, but if the transformer is marginal, you just "run what you brung".   

The last suggestion I offer is to test the response with a DC bias on the secondary, such as would appear when the transformer is in the cathode circuit.  Add a DC supply through a dropping resistance, and apply the rated current through the secondary.  A resistance greater than the matched load resistor will provide the winding DC current without significantly adding more parallel load resistance, such that your measurements are still valid.   With a DC bias on the winding, the response measurements will reveal if partial core saturation is impacting the low frequency response.  If this test results in unfavorable performance, you might consider passing the final amplifier DC cathode current through a separate inductor, and capacitively couple the output of the modulation transformer to the inductor.   

The transformer may have many options as to whether the primary, and the secondary, can be composed of either series or parallel-connected windings.  Using the primary series configuration increases the inductance, reducing low frequency roll-off.  Using the secondary windings in parallel reduces the effects of the DC bias, and provides a higher current source, (lower losses) to drive the low impedance load.  Of course this selection is constrained by the DC current rating of each winding.

Hopefully these simple tests will help you determine whether the project will meet your performance expectations for your next transmitter.   See you when you reappear from the rabbit hole!

73, Rick

[KD1SH]

  Thanks, Jerry—some very cool stuff in there. Interesting to note, too, that Frank, C. Jones was on the editorial staff of Radio magazine. There were giants, in those days.
  Beyond the information on cathode modulation, the old Radio magazines are great fun to pore over: $2.50 a year to subscribe, and not a single mention of FT8 to boot!
  

TNX for the UTC data sheet. Big help.

You may want to look at the article from Dec 1939 RADIO Magazine (page 16) about cathode modulation design and adjustment. I found it a good read.

73 Jerry W1ZB

[K8DI]

Regarding the inductance of the two halves of a winding…this is normal, if you think about the physical construction.  Wind a bunch of wire (many layers), bring out the tap, wind the same number of turns again (to get the same turns ratio). The two coils have different diameters, and the length of wire is longer because the diameter is bigger on the second half.  As you most likely know, as the diameter of a coil increases, so does its inductance. The longer piece of wire explains the slight mismatch in resistance often seen, and tells you which one is on the outside. With certain transformer types (CT both primary and secondary) this can also tell you the polarity primary vs secondary.  Fancy audio transformers are bifilar wound…mostly eliminating this difference, slightly increasing their linearity (and cost).

Ed

[Steve - K4HX]

And Radio had an awesome handbook too (better than the ARRL's for a while IMO). Jones was one of the editors.

I used a rig with this sort of cathode modulation briefly. It was Stancor 110. It used two 6V6s to modulate an 812 triode. This version of cathode modulation is actually modulating the grid and the cathode. It generally more heavily modulated the grid so as to only need a relatively small amount of power to modulate the cathode and achieve 100% modulation. There were a series of article in Radio and coverage in their handbooks on the tradeoffs in the amounts of modulation of the grid versus cathode and the amount of power required. IIRC, there was a chart/graph that would allow one to select the amounts and see the power required.

[WD5JKO]

 Attached a RCA Ham Tips, Jan-Feb 1940, covering Cathode Modulation in detail.