Modulation Transformer Mods

[KI4YAN]

I read here about ungapped modulation transformers, and how they are used in higher power equipment. I have a SNC 5P325 multimatch modulation transformer, something similar to a Stancor A-3894. This Stancor model lists as a 125W modulation transformer, and is rated for 225mA primary and secondary current. Obviously, it's a gapped mod transformer. I recently was recommended to save my pennies and go with a single 812A as my RF final, and managed to pick up a number of 807 tubes to use in the modulator. The original modulator-skeleton that the modulation transformer came from has just been identified as none other than this one, set up to run the pair of 807's at 750V plates, for 120W output in class AB2:

http://viewoftheblue.com/misc/807_1.png
http://viewoftheblue.com/misc/807_2.png

My question is this: IF I take the time to relaminate the transformer in a cross-laminated fashion and use a modulation reactor, will it handle more audio power? If so, how much? Will it improve the sonics enough to bother doing even if I don't exceed the 125W guesstimated rating?

I have four NEW 807s, and three used ones. With a single 812A or 811A final, my DC input to the final will be 175W. it is my understanding that to fully modulate this, I'll need ~88W of audio. Is this a correct assumption?

[WBear2GCR]

Perhaps the story is backwards?
The gapped transformer is used in higher power apps.
The ungapped in lower?
Although I seem to recall that both are used in low and high power situations...

The gapped one will not saturate/shift the low freq F3 point as the result of DC offset or significant modulation as much as an ungapped transformer will. Otoh, for the same physical core and windings the gapped transformer will have less inductance at low levels, and so not go as low in frequency.

In practice, one can gain a benefit by taking the DC off the core of any transformer, but especially one that is ungapped, via the use of a "modified Heising" scheme.

In p-p operation, there ought not to be much DC offset, which obviates one of the reasons to have a gapped core.

But, if you are not needing extra LF response, and you are not running the transformer at or near its power capability, then there may not be any real benefit to building in the extra components required for modified Heising.

The short answer to your question is that going with modified Heising and ungapping the core ought to buy you some LF extension, not more power handling (at least not much) - and the need for that LF extension depends on your intended use and how "good" that transformer actually is in the first place.

At least this is my limited understanding on this subject...

              _-_-bear

[k4kyv]

The lack of DC might buy you some more power capability in terms of the range of magnetisation of the core before saturation, but you will still have the same limitations of voltage breakdown of the transformer insulation. The lack of DC will give additional a.c. current carrying capacity to the wire in the windings, but that is usually not a limiting factor in a modulation transformer.  With a proper reactor it will definitely give a big improvement in low end response.

I once read an article in a circa 1956 issue of Radio and Television News in which a ham claimed he could get enough audio through an ART-13 modulation transformer, using a 10 hy power supply choke as a reactor (without modifying the transformer), to modulate 700-800 watts DC input without (immediately?) blowing the transformer.  This was an emergency fix after he had  lost his regular modulation transformer.

I cross-laminated the mod xfmr in an Eico 730 modulator and added a 40 Hy reactor (a real Electro-Engineering mod reactor rated at 120 m.a., not a power supply choke), with some changes in coupling capacitors in the speech amp, and the thing is flat from 20~ to over 10,000~.  I use it to modulate my Eico 720 10m rig. I normally feed my 500-ohm balanced audio  line into the phone patch input and bypass the built-in mic preamp altogether.  I did not attempt to run any more power with the rig beyond its stock design.  I removed the tube and all other components from the speech clipper circuit, which left enough room on the chassis to mount the reactor.  But the latter is too tall to allow me to re-mount the top cover back on the modulator.

[KI4YAN]

So realistically, my 125W mod transformer will hold up to 125W, and that only if it's a good one. I am getting ready to test the transformer again and verify the connections I'll need to make for the required impedances, is there anything other than insulation test and impedance ratios that I would need to know?

[WZ1M]

When you say "insulation test", I assume you mean testing from winding to winding and winding to ground using a dc hypot tester. Another very important test that is overlooked and its probably because of no test equipement to do it is, a PI test. This tests the insulation for breakdown. Test takes about 10 minutes to do. This test measures the insulation resistance at 3 minutes and then at 10 minutes.
Regards,
Gary, WZ1M

[WA1GFZ]

Gary, I would think you could use a power supply or hi pot in series with a microamp meter to do the PI test. I never heard that term but have run the test as part of a hipot isolation resistance test. I worked on isolated systems and we had to put 1000 volts on all connector pins and check for current through the chassis. 

[W1ATR]

I would just hipot it. A PI test is going to be hard on those old windings. If this tranny is the ONLY one that will fit the application you want to use it in, then you may not want to risk it. A PI test will absolutely find a fault if done right, but if your dealing with old insulation and mod iron is getting harder to come by.


Gary, If I recall, it's supposed to be 1 minute, and 10 minutes.

[WZ1M]

Yes, 1 minute and 10 minutes. Senior moment. I have to disagree with you on "hard on the old windings". I would rather have it breakdown now then under a full load.
Gary...WZ1M