Discussion in 'Amplitude Modulation - AM Fans' started by W2BTK, Jan 11, 2019.
tnx for the Crown amp info; plus clarification on Stu's role.
Don't overlook the possibility that the raspiness might have been caused by diagonal clipping in the receiver...lots of vintage receivers suffered from that with ultramodulation. (The synchronous detector took care of most of that problem). You might have been transmitting a perfectly clean signal.
Bell Labs engineered and WECo built loads of toroidal audio transformers for broadcast and telephone service. Wide bandwidth, nearly indestructible (I've seen a 111C repeat coil used as a 120V isolation transformer in a pinch) and smaller per Watt than C or EI iron.
Kudos for making them work for you.
Surprised to see my name mentioned in this thread. Yes, I did it 35 years ago with a Harman Kardon Citation 16S amplifier with the left and right channels in bridge configuration. And I used a Harman Kardon Citation II audio output transformer reversed connected through a 4 ufd oil capacitor and (2) 20 Hy chokes in series as a modulation reactor. I used this lash up to modulate my Globe King 500C transmitter. I ran it like that for at least a couple of years. The 60 Watt reverse-connected audio output transformer never complained about being operated well beyond its intended power limit.
I should point out that I am not by any means an innovator of the technique. I had heard about it long before I even got my ham license. The idea has been around for many decades before I knew of it.
The idea of using toroid power transformers as modulation transformers is a lot newer and solves the problem of obtaining correctly-sized audio output transformers that are now harder than ever to find.
I had a QSO with Dave, W2VW on 160m last night. He was running a highly modified Johnson Kilowatt with outboard solid state modulator and a lash-up similar to what was described in the OP. Sounded good, with full modulation, except that he was lacking in highs and that made the signal hard to read whenever he faded into the background noise (just white-noise hiss, no thunderstorm static last night).
Not too bassy and the mid-range was about right, but not sure if the lack of articulation had anything to do with the frequency response capability of the toroidal power transformers, or if it was merely the tone-control settings of the amplifier.
Great post, I enjoyed your presentation and what you are doing with toroidal power transformers in audio applications. I am currently working along similar lines in two applications, more on that later. The Crown Xli-800 amplifier is a two channel power amplifier at 300 watts per channel into 4 ohms with the ability to work in bridged mode at 600 watts into 8 ohms. That being the case you may want to consider using just one transformer, with all windings. The impedance calculations would be something like:
(800+800+6.3+6.3) / 115 = 14.023 to 1 voltage ratio, hence 196.63 impedance ratio (using all windings) With an RF load of 3000 ohms on the secondary, the reflected load at the primary would be about 15.3 ohms. I assume that in a bridged mode, both channels driven, the Crown would be rated at 300 watts into 16 ohms. A good match for this transformer load combination!
Crown amplifiers are classic, an excellent choice for this application, good frequency response, robust power and well built. By utilizing just one transformer and the bridged amplifier mode, you would be taxing the Crown less, (and we all want lower taxes!), with lower transformer losses in total, and a slightly better frequency response. Am I missing anything here? A Heising; i.e. impedance coupling, would be essential to keep unbalanced DC from the transformer, again an excellent choice. Values for L and C are fairly well defined. One item to consider is the exact configuration of coupling audio to the RF section; the diagram as posted uses the power supply filter capacitor in the audio loop. The loop would be: modulation transformer secondary, coupling capacitor, modulation reactor, power supply filter capacitor, and back to the modulation transformer thru the ground connection.
With respect to the ultra-modulation system I concur with your analysis. It is a means by which one can install a measure of negative over modulation protection, while at the same time piling on more audio! With my Globe King 400, I do precisely the same thing, (OK … you caught me I over do it!), using the three diode method of ultra-modulation with a 200 vdc keep alive voltage. Often times when I am sitting peacefully, looking out from the back porch, having a frosty mug of scotch, I ponder having one horse power of audio, and hitting modulation peaks of 200%!
My attempt to use toroidal power transformers in audio applications is in two applications. First is using a 115/230 vac to 12 vac 50 VA torrid as a driver transformer for my push pull parallel 805 modulator. The frequency response is quite good and the ratio is a good match when using a PA amplifier with an 8 ohm output to the 805 grids. The modulator seems easier to drive with good audio peaks.
The second application, (which I am just now working out on paper), is to use an AnTek AS-5222 or similar in “turbo-connection”; i.e. auto-transformer, coupling my push pull parallel 805’s to the Globe King 400 replacing the current CVM-5 modulation transformer. The RF load of the Globe King is about 4700 ohms while the required plate to plate load of the 805’s is 3350 ohms.
Again, thank you for the post hope to work you on 75.
Bruce – WA3JVJ
Hey Bruce. That sounds like it would work pretty well, it makes sense. I am just not sure how you would get the crown to 16 ohms, you said drive both channels? I'm not quite sure how that works. I just assumed I could use one channel for 4 ohms and both channels for 8 ohms bridged and that's it. I'll post an updated schematic of my modulator below that is drawn better and shows exact values.
Bruce, I have a pair of the antec dual 800 volt transformers I have no use for if you want them.
I would like them to be gone...
How would 3B28's work for the diodes?
Are they fast enough?
The original 1956 QST article on ultramodulation used 866-As. 3B28s are probably just as fast.