Modulation Transformer Info

[KE6DF]

I just ended up with a UTC modulation transformer, and wondered if anyone had any more info on it or knew of any schmatics using it in an old handbook or UTC document.

The transformer is model number LS-67

An old UTC catalog shows it as having two primary taps at 9000 and 6800 Ohms impedance, and two output impedances of 2500 and 10000 Ohms. The secondary side has two separate windings which can be connected in parallel for 2500 Ohms or series for 10,000.

The power rating is listed as 260W.

One question I had was, in the UTC catalog it says that the CVM and S series modulation transformers were designed to handle the DC plate current running through the secondary. (Of course they may work better in a modified heising circuit).

There is no comment one way or another on this issue in the description of the LS series mod transformers.

The only schematic I can find using an LS series modulation transformer is in another UTC uplication and it's for a 1000 Watt modulator using a big brother of my transformer. That circuit is a modified heising design with a UTC LS modulation reactor.

Of course LS series modulation reactors are made of an even purer grade of non-obtainium than LS modulation transformers.

So has anyone every worked with one of these LS-67's?

Thanks,
Dave
WB6IYH

[k4kyv]

No, but I do have a UTC LS-103 modulation reactor in my transmitter.  It is rated at 50 Henries @ 500 ma.  I have a LS-691, rated at 1000 watts audio.  I plan someday to use it with a pair of 833A modulators and the reactor  from a Raytheon RA-1000.

I don't think any of the LS series modulation transformers are designed to carry DC through the secondary.

[K1DEU]

Is the core visible ? If it is gaped, it is designed to carry unbalanced DC.

Ah yes, I will accept all UTC Linear Standard (LS) series transformers and chokes gladly.

Gaped would mean that each I and E would be stacked on top of each other. Picture one hundred I's and E's laid on top of each other. If we look at the outside perimeter we will see a perpendicular straight line at the end of each I. So there is a gap formed by each junction of an I end and the outside of each E. We can view two gap lines on two core sides. 
     When we add an ever changing magnetic fields (AC/Audio sine waves) to this (one common core) strong non changing magnetic field (created by DC current) the audio is not completely freely allowed to change its magnetic polarity resulting in a slightly distorted waveform called BH. This resulting BH distortion can be heard by a trained ear with a/b tests on the air.


Cross laminated would mean that every other E+I is laid on top of each other with one I on the right then the next I on the left and we will easily view this on the core out side. With cross laminated the magnetic flux easily shifts magnetic poles with little distortion.

With Low distortion (no BH) Audio transformers or Modulation transformers designed not to carry unbalanced DC the E's and I's would be cross laminated. Sometimes in a 3-2 pattern, or altering every other one. Primary Inductance will more than quadruple if we take an unbalanced (gaped) core and cross laminate it (5 1/2 to 6X for thinner laminations is common). This primary inductance determines our low frequency corner (our bass response). Often chasing low frequency corners we will often cross laminate an existing gaped core and if its a modulation transformer simply add a coupling cap and a heising reactor (DC or Audio Reactor Choke) to carry our unbalanced DC externally, gaining much low frequency response before 90 degree phase shift. Some old articles.

http://hamelectronics.com/k1deu/pages/ham/transmitters/am/pages/modulation_power_mods.htm

http://hamelectronics.com/k1deu/pages/ham/transmitters/am/pages/using_audio_reactor.htm  BH Distortion !

http://hamelectronics.com/k1deu/pages/ham/transmitters/am/pages/negative_feedback_design.htm

Regards  John, K1DEU

Finally posting a few pictures  http://amfone.net/Amforum/index.php?action=gallery;su=user;u=1341

[KE6DF]

Thanks guys,

John,
I looked at the transformer and it does not have a gap. It appears to have every other core piece going the other way.

It kind of figures since the LS series was intended for HI-FI broadcast work, whereas the CG (and the equivalent PA) was designed for voice communications (like police and other two way radios), and the S series was aimed at Hams.

Plus only the LS series offers mod reators. It looks like the LS-102 rated at 50HY and 350MA it the obvious reactor mate for my tranny.

Don,

The LS-691 is the mod transformer used in the 1KW modulatior schematic from the 1951 UTC publication.

FWIW I've attached the page below.

Thanks again, and if any of the others of you out there have any suggestions on how to use this LS-67 modulation transformer, please chime in.

Dave

[k4kyv]

Dave,

Thanks for the UTC schematic.  It is an updated version of the pre-WW2 one I have in one of the earlier UTC books.  Pretty much what I have in mind for the homebrew 833A modulator. 

The optimum class-B 833A plate impedance is very close to that of the 849.

I once had the opportunity to buy a half dozen 849's (condition unknown, but they  looked clean with no discoloration on the glass envelope nor loose pieces rolling around inside) at Young Engineering in Salem, MA for $20 apiece.  I thought that price was a little steep at the time (mid-70's), and the guy who worked there was an asshole and wouldn't budge on the price, so I didn't take them, and have been kicking myself ever since.  That $120 bux would have had zero effect on my long term financial situation, but if the tubes had been good, I would still have them and probably be using them now.  But I passed them up for the few lousy dollars that I couldn't talk him into shaving off his asking price.

I had dug them out of the bowels of the unlighted basement, below the main area where all the stuff for sale was stored and displayed.  The guy was pissed off that I had ventured down there with a flash light looking for stuff, and hadn't even known they were there, until I brought them out to the counter.  I went back to Young's some time later and a different employee told me they had kept them out on display for a while, but as things got shuffled around from time to time, he thought the big tubes had all got broken, one at at time, until they were all gone.    

[WQ9E]

Finding modulation reactors today does appear very difficult and that brings up a question.  Would it be possible to use several chokes (say 5 10 henry units) in series to approximate the function of the mod reactor?  Not a very compact solution I realize but I have an old mod tranny out of a Westinghouse transmitter that requires a mod reactor and I would like to put it to use someday.

It was informative to read about the LS series of transformers; I picked up a box of NOS UTC transformers at a small hamfest several years ago because the 3 CVM-0 units on top looked interesting for some peanut whistle rigs and down near the bottom I found a UTC LS-55 and a LS-57 unused in their original boxes.  The whole works (around 50 transformers and most of them very small units) was $5 so I may put the LS-55 and 57 on ebay at some point and use some audiophile dollars to purchase additional vintage radio gear.

Rodger WQ9E

[WA1HZK]

Yes
Dig up some power supply chokes and string them together. Works Pissa!
Keith

[WQ9E]

Thanks Keith!

Rodger

[K1DEU]

Roger et All; Don't forget that gaped modulation Transformers also make good Modulation Reactors. Several years ago I bought several T-368 modulation transformers to use as a audio chokes. Just series the windings in phase. Often we do not series all the windings, use your ohm meter and inductance bridge to check wire size!
Regards John

[KE6DF]

John,

Very good suggestion. I have a Kenyon T-478 modulation transformer rated at 225 watts. Using the whole primary plus part of the secondary in series I measured 38 henries on my inductance meter. I can't measure the primary plus the whole secondary as my meter only goes up to 40 hy.

I suspect that the Kenyon would make an adequate modulation reactor for my LS-67 if I design around the 10,000 ohm output impedence connection on the LS-67. That would give me something like 2200 V B+ at 220 ma on the final and the Kenyon is rated at 270 MA+ depending on which secondary taps you use.

BTW, your articles were very helpful.

Dave

[w4bfs]

you can use a meter/oscope and an audio osc to measure inductance ... series feed the osc thru a known value resistor, say 3.3k to the inductor ... when the voltage drops are equal, then at that freq you have 3.3k of inductive reactance .... xL = 2 pi f L  ---> L = xL / 2 pi f ... a bit of work and is neat to see on 2 ch scope ... 73 ... John

[k4kyv]

you can use a meter/oscope and an audio osc to measure inductance ... series feed the osc thru a known value resistor, say 3.3k to the inductor ... when the voltage drops are equal, then at that freq you have 3.3k of inductive reactance .... xL = 2 pi f L  ---> L = xL / 2 pi f ... a bit of work and is neat to see on 2 ch scope ... 73 ... John

That works only if you have the full rated DC passing through the chokes.  The inductance, with no dc, will be higher, since the rated dc will cause partial saturation of the core and reduce the inductance.

One method I use, which requires a little care and comfort working around  high voltages, is to temporarily wire in the unknown inductor in series with a choke with known inductance @ a certain dc current flowing through it, into my transmitter as a temporary modulation reactor.  I place the two reactors in series, load the transmitter so that the plate current is at the rated current of the known reactor, and using a test oscillator set to somewhere between 500 to 1000 Hz, tone modulate the transmitter about 30 to 50% (to avoid overheating the modulator tubes).  I then use an a.c. voltmeter, usually a multimeter with the a.c. voltage range set to about half the DC voltage on the PA plate of the transmitter. Be sure to use good HV insulated meter probes and be careful not to let anything short together or to ground.  Cut the transmitter off and make sure the filter caps are completely bled while handling the clip leads and connecting the reactors.

With the transmitter on (preferably into a dummy load or on the air on a dead band) turn on the test oscillator and adjust the audio level to get a good reading of the a.c. audio voltage across each of the two chokes, and record the two voltages.

The a.c. voltage drop across each inductor will be proportional to the inductance.  For example, if the known inductor is rated at 10 Hy @ 500MA, and the a.c. voltage reading across it is 300 volts, and the meter reads 900 volts a.c. across the unknown reactor, you should get a total of 1200 volts across the two reactors in series.  That tells you that the inductance of the unknown reactor is three times that of the known reactor (10 Hy), or 30 Henries.

My transmitter uses a 50 Hy, 500MA mod reactor, so I can also temporarily wire the unknown reactor in series with it, and compare voltages.  At less than 500MA, the mod reactor is likely to be more than 50Hy, but I can still put the unknown reactor in series with it, measure the ac voltages and get a pretty good idea of the inductance, or at least how well it compares to a known good modulation reactor.  If I am measuring a choke at 100 ma, I can get the most accurate results by inserting a known choke rated at 100 ma in series with it, while shorting out the mod reactor in the transmitter or disconnecting one end of the winding, and taking the measurements.

I have measured the inductances of unknown modulation reactors and power supply chokes using this method many, many times over the past 4-plus decades.

[w4bfs]

thats a good point, Don ... I use the scope method just to get a rough idea (and smaller chokes)... what do you find as typical change of inductance with addition of rated dc current ? ... does 2:1 seem reasonable ? I'm sure it will change with inductor construction ... seems if I ever get over that way I might have to find some of that cheap Nashville microbrewery beer and buy you a round or two ....73 ...John

[W9GT]

Good discussion. Thanks guys!  Lots of good info here.  Also good parallel discussion in another thread on Modified Heising Modulation.  I checked the schematic for the Bauer 707 BC xmtr.....looks like they used a 5 KV cap for audio coupling with a nominal 3 KV plate voltage on the RF final.

Thanks and 73,  Jack, W9GT

[Tim WA1HnyLR]

Hi Dave , I used to have an LS 67. I used it as a driver transformer for the pair of triode connected 4-1000 s in the modulator . I turned it around using the secondary as the primary. The full 9.7 k drove the screens and the 6.8K drove the grids. I tied the 9.7 k winding terminals together as the centertap and used the so called centertap connections (spilt winding) to drive the screens of the 4-1000 modulators. This arrangement provided a much greater voltage ratio between the grid and the screen of each modulator tube. I used a pair of 845s as the audio driver tubes. I had performed tests of the LS 67 in regard to various load impedances in regard to the rated load impedance . The LS 67 blew away all other transformers I tested. It is without a doubt the BEST mod transformer to use in a medium power transmitter. Unfortunately I lost the transmitter in a fire in 1992. I have been looking for an LS 67 ever since. If you want to sell it  or swap it for something ,I would be interested. If you want to use it. I would use it in one of two ways. Either a single 813 or4-125 modulated by a pair of 811s or 805s.Class-b.  The 805s are preferred. One may also use a pair of 845s in AB1 resistance coupled.Use a plate supply voltage of 1250-1500-DC. Being that the LS-67 is such a great mod xfmr it would be a shame to use it in a circuit with the conventional driver transformer circuitry. One should use a direct coupled cathode follower scheme. One should wrap much negative feedback around the modulator as well. The secondaries should be in series for 10K Impedance operation. Due to the high impedance of the PA stage a heising reactor value of 50Hy minimum should be used . Preferrably 100Hy. Being that the transformer presents a slight step up ratio it will be possible to modulate 150% positive no problem . You would have a very fine 150-200 watt transmitter with this piece. Another use would be a modulator for an existing transmitter like a Valliant or DX 100 etc.Use a pair of 811 tubes with 1000-1250V on plates . Parallel up the secondary windings for 2500 ohm operation. About 30Hy would be a good value of inductance for the reactor. I would be very curious what you wind up doing with this piece of iron . Good luck.
Tim WA1HnyLR

[KE6DF]

Thanks Tim,

I had been also thinking that I don't want to use the LS-67 with a standard driver transformer driving modulator tubes. I have a UTC CG-53ax and also a PA-53ax which I believe are pretty much the same in different cases. They would make reasonable driver transformers for a CVM-3 or 4 or an S series based modulator, but they would just put me back into communications quality level audio.

I could use an LS-49 or LS-47 in a driver stage for the LS-67, but that puts me back into the mode of obtaining unobtainium.

I do have a few 828 tubes in my junk box. I was thinking of using a pair of those AB1 and then some kind of transformerless driver with negative feedback. 845s or 805s might be better, not sure.

I played around with my Kenyon T-478 some more, and I believe that with the primary and full secondary in series I would have around 60 HYs. Of course, this is with no DC flowing thru the transformer. I'd have to do something like what Don suggested to get an inductance reading with a DC load.

The Kenyon, is however, intended for use with the final DC flowing through the secondary. The catalog specs it for from 270 to 400 ma DC on the secondary. If I shoot for a plate voltage in the 1500-2000 range I'd be running DC through it well below it's max ratings.

I can't tell if the core is gapped as it's a potted unit, but assume so give the specs and intended use (communications and amateur, 1930's vintage).

Dave

[k4kyv]

I have a UTC CG-53ax and also a PA-53ax which I believe are pretty much the same in different cases. They would make reasonable driver transformers for a CVM-3 or 4 or an S series based modulator, but they would just put me back into communications quality level audio.

I could use an LS-49 or LS-47 in a driver stage for the LS-67, but that puts me back into the mode of obtaining unobtainium.

The CG-53ax and the PA-53ax are identical.  The case is the only thing that is different.  UTC also made similar changes in the style of the LS series, but did not change the type numbers.  They also rather drastically changed the electrical specs on some of the LS series transformers over the years, without changing the type number. One I recall is a power transformer; the older pre-WW2 version has a 2.5 volt filament winding, but the newer version has a 6.3 volt winding.  Both transformers carry the same type number. The catalogues rate some of the later LS transformers at substantially lower power levels than the earlier versions with the same type numbers.  I'd feel sorry for some poor slob who tried to replace a crapped out older xfmr with a new one, thinking he had an identical replacement transformer. I don't know why they didn't at least add a suffix or something to indicate that the electrical specs were changed.

The CG/PA series transformers are midway between the LS (broadcast quality) and the S-series (typical "ham radio" quality).  But I have run tests and found that many of the LS-series, particularly the older versions,  fall far short of the 20-20,000 Hz +/- 0.5 dB specifications listed in the catalogues.

I looked for an LS-49 for many, many years before I finally obtained what was probably the very last one in existence.  It was new old stock, still in the original UTC box.  It is designed to match exactly what I have in my transmitter, a quad of 2A3's in pushpull-parallel driving a pair of triodes in class-B.  Chicago Transformers made a very similar transformer, the type BD-2.

The UTC LS-48 is rated at about the same power level as the LS-49.  It is designed to match a pair of 845's to drive a pair of class-B triodes.

I believe Tim lost several LS-47's in the fire.

[WBear2GCR]

To drive the modulator tubes there is also the "cathode follower" type driver. The advantage is that no coupling iron is required. There are a few variants on this design, I think that Guy, W1FRM favors these, and uses a resistor to ground from the cathode in his designs... fwiw.

That might be a way to go...

           _-_-bear

[K1KFI]

A somewhat related situation...

I have built, almost there, a modulator with 805's, an Ls-33 and a cvm-4.
The primary of the vm-4 is doing what it should. With the secondary setup like the UTC chart shows for 5200 ohms the mod transformer only gives 100 watts, should be 300 plus. Since no schematic of the vm-4 exists to my knowledge I'm scratching my head! Has anyone had this experience and can share some information on how to make it talk like it should?

Cliff
K1kfi

[KE6DF]

Cliff,

Attached is a schematic from a 1930's UTC publication showing modulators using the VM-3, VM-4, and VM-5. See the last one on the page as it's a 805 based modulator with a VM-4. The schematic of the VM-4 is shown, although they don't number the pins. But perhaps you can figure it out from this plus the connections table you have.

Dave

[K1KFI]

Thanks Dave!

 

[Steve - WB3HUZ]

Connection info for the VM series trannies here:

http://www.amwindow.org/tech/htm/modtran/utc/cvm.htm

[WBear2GCR]

RCA calls for 6700 ohms plate to plate @ 1250vdc on the plate and about 6-7 watts to drive it into grid current. If you have 1500 vdc on the plate then the plate to plate load is up to 8200 ohms.

FYI.

          _-_-bear