Transmitter Power Supply Design

[KI4YAN]

Phase 2 of Project Timesink:

Build the power supply system.

I have been cruising Ebay for parts and pieces, and much to my dismay, the "Probably a big choke from a radio transmitter, dunno..." actually turned out to be ANOTHER plate transformer...much larger than the one I *JUST* wound. Similar voltage...only slightly higher at 2440VAC...and MUCH heavier. SCORE!

Lucky for me I have another bobbin on the way for my EI175 laminations and can use the same 1.75" wide, 4" tall stack of them for the modulation choke OR modulation transformer. This stack of laminations is capable of being wound into a 900VA power transformer, CCS, or 1.25KVA transformer ICAS.

I have 1.5" wide, 2" tall stacks, multiples...lots of them...that can be wound into chokes as well. These stacks can be turned into 250VA power transformers, CCS rating. I probably have some others in the pile as well. A 2.5" stack is good for 325VA. My STC multimatch modulation transformer and the Thordarson multimatch 125W units are wound on a 1.5" wide, 1.5" tall stacks, so I *should* be able to wind the modulation transformer on one of these cores and get a 150W capable unit. At only 2000V, 150mA on the plate, the 4-125A/4D21 will be sitting pretty with a 225W carrier for 300W plate input, although the modulator will be able to produce 280W of audio with ease.  

I am going to use solid state diodes, each rated 1kV and 6A forward current, for all the rectifiers. I have a box of 100 of them handy.

Picked up some 115uf, 2300VDC paper-in-oil caps for the HV, and have some high voltage 'lytics for the MV supplies, as well as the low voltage systems.

I am going to need the following voltages, PLUS some that I'm sure I'm not thinking about!

+2000VDC for the modulator plates, 220mA (4-65A)
+2000VDC for the PA plate, 150mA (4-125A)
+250VDC for modulator screens, 60mA regulated
+250VDC for exciter/LO, 100mA combined
+350VDC for 10W speech amp, appprox. 150mA combined
+350VDC for RF PA screen, 33mA (screen fed via high inductance choke for less heat)
+350VDC for  IPA plate, 50mA (2E26)
Various heater voltages, 6.0V for the 4-65A's, 5.0V for the 4-125A, 6.3V for the exciter, probably 6.3V for the speech amp

I probably ought to design the speech amp next...I could actually have that up and running in a day or two.

What am I missing, and what kind of problems are likely to crop up in the power supply section?

[KA2DZT]

What about a bias supply for the modulators.  Maybe a low voltage DC supply for relays.  You can run the PA screen from a screen dropping resistor off the PA modulated B+, eliminating the need for a PA screen supply.

Fred

[N2DTS]

I always liked a seperate supply for the screens.
More control, no danger the screen will get to plate voltage.
A big pot allows you to drop some voltage for screen protection, self modulation, and makes tune up much less critical.
A simple overload latching relay circuit protects the screen no matter what happens.
With a variac voltage supply, any tube can be used, things can be adjusted to give the best intermod, class of operation can be changed.

Early on, I decided to build things in a modular adjustable format, so I have an RF control deck.
I can hook up almost any type RF deck to it, and just adjust things.
I could go from a single 6146 to three 4d32, s, to a pair of 4x150a's, or a pair of 813's, or whatever I want, up to the ratings of the filiment supply.

Saves a TON of work when you decide to add a second tube (or a third), or build a new RF deck, or try screen modulation, or use the thing as an amplifier.

For small stuff, you can tap off a low voltage supply and send it to some VR tubes to give whatever voltage you want.

[KI4YAN]

I think I will have the filiment transformer on the rf deck, because I might build a 4cx250 rf deck in the future and I will want to test it in this rack, probably. I have a box of the little tubes but don't have any good way to test.

[KI4YAN]

Made up my main rectifier bridge; using 6A10 diodes in series for each leg of the bridge. Bridge should handle 5kV at an amp continuous, so I don't think i will have much problem there.

I've picked out some power transformers that should handle the current, and I will be regulating the screen voltages for all pentode and tetrode audio stages.

WHOOO! pass tubes!

[N2DTS]

I use a circuit I got out of the Bill Orr handbook to regulate the screen voltage in the 4cx250b bod deck.
It uses 3 tubes, a 6080 (6as7?) dual triode pass tube, a VR tube, and a small tetrode (6sq7?).
Its adjustable, with a control on the front panel, and works well, and has been trouble free for 30 years!

You can also use various combinations of the larger (40 ma) VR tubes, with a big wire wound pot as a dropping resistor.
They made 75, 90, 105 and 150 volt VR tubes, so you can get any voltage you need with some combination.

[WQ9E]

The Tektronix 500 series tube type mainframes used 6080 pass tubes to regulate several of the outputs in their "totem pole" stacked B+ supplies.  My Type 555 dual beam is typical of the overall design.  A VR tube provided the reference for a -150 volt supply (also using a 6080 pass tube) and then this -150 supply provides the reference voltage for 100, 225, and 350 volt regulated supplies.

These supplies were well designed and highly reliable and provide a good reference for anyone designing well regulated very low ripply high current supplies.  Manuals are available for free download on edebris and other sites.  I have used the transformers from parted out Tektronix scopes in various supplies, and no I wasn't the one who parted them out   I don't have the heart to do that to such nicely built equipment.

[KI4YAN]

I had three old 561A's that eventually died, the HV section had a little ferrite transformer that eventually shorted in all of them. I salvaged everything out of them, INCLUDING the power supply schematics. Novel design, and they only needed ONE pass tube to regulate ALL the supplies.

The 6080 pass tube regulated the -150V supply, referenced to ground by the VR tube and error amplifier. All the other supplies were rectified and stacked on top of the regulated 150V supply, so as long as the voltages coming out of the transformer all moved the same degree, they were all regulated by moving the -150V rail.

So, the +350 stayed 100V away from the +250 due to being wound on the same PT, stayed 150v away from +100V, stayed 250V away from the regulated -150V. As long as one end of the stack is regulated, the other supplies will track it!

[WQ9E]

Stan Griffiths discusses that transformer in his Tektronix scope restoration book.  It is a neat HV supply since it uses the rectified output of an RF oscillator to develop the accelerating potential for the CRT.  Given the very high ripple frequency it is easy to filter and if something fails the oscillator simply stops due to the heavy load instead of sending parts up in smoke.   The ZM-11 universal bridge uses a similar type supply for capacitor charging voltage so if an electrolytic under test shorts the oscillator stops dropping the voltage to zero quickly without drama.

My father was a civilian supervisor with the Air Force and worked in the precision measurements lab.  I learned how to calibrate the 500 series scopes at a very early age.  Before one of the engineers came up with using coaxial delay lines it was a  tedious process to adjust all of the L/C constants in the delay line to pass a proper square wave. 

The 555 is probably the least green of any piece of test gear I own and has a very high tube count in addition to two large cooling fans.  There are 15 tubes including five 6080 tubes in the separate power supply, 23 tubes between the two plug-in timebases, and another 72 in the mainframe proper and of course plugging vertical amps into either of the two slots adds to the tube count.  It pretty much offsets the use of CFLs in the rest of the house   I use a pair of more recent 7854 scopes on the bench but I have several of the older scopes including a 544, 545, 555, 556, and the very old 514AD that belonged to my father.