Traditional tube amp energy provides are outdated school–relatively excessive voltage, with large “iron,” and customarily not regulated. Typically, they provide a variety of voltages for various purposes–a present supply for the output transformer, voltages for the preamp tube plates, and generally (on this case) a separate voltage for the pentode screens.
Unlike regulated provides, the totally different provide voltages are created with current-limiting resistors. These are sometimes known as “voltage-dropping resistors,” however their operation is determined by the present draw of every stage.
Designing an influence provide
The first step is choosing the proper energy transformer (see the “How did this venture get began?” part.) To choose the correct transformer, take a look at the info sheets for the facility tubes.
The 6DG6GT tubes have a max plate voltage of 200V. Theoretically, an AC RMS voltage is ~zero.7 of the height voltage, and the height is approx 1.414 * the RMS. In apply it is lower–the transformer is below load, there are losses within the caps, and many others. So one thing lower than 1.four is extra real looking. (Gotta dig that loopy sq. root of two…that 1.414 quantity pops up in so many locations!)
I am not sure in regards to the availability of PTs with secondaries within the 125-150V vary. But possibly the 6DG6GT can deal with considerably greater than 200V. Another various is to make use of a “choke enter” energy supply–that’s connecting the choke FIRST, earlier than any filtering cap. A choke enter ought to drop the secondary voltage to zero.9 of the RMS (vs the 1.414 for the standard filter), so a 225V RMS AC secondary yields 202.5 VDC, additionally glorious.
My “recycled” transformer was ~140V (142) RMS AC, which, when rectified, (in an excellent world) turns into 200.788 peak (VDC)–perfect! (in practice–rectified, filtered and loaded, it is about 190V, nonetheless glorious.)
The solid-state rectifying bridge was chosen over a tube rectifier to retain as a lot of that voltage as doable. That’s OK–the a lot vaunted “sag” impact of tube amps would not apply to single-ended, Class A amps. They draw the identical quantity of present whether or not there’s an enter sign or not… Also, the PT would not have a centertap, so except I used two tube rectifiers (or went with a half-wave design), solid-state was one of the best answer.
These voltages have been wanted by the circuitry:
B.1 : 190V — Max voltage for the facility tube plates/output transformers
B.2 : 180V — A faucet for the preamp tubes (Added throughout construct)
B.three : 120V — Screen voltage for the 6DG6GT energy tubes (between 115-125V, relying on the info sheet)
I did the preliminary design utilizing a wonderful (free) design software: Duncan Amps PSUD2 Designer
The closing consequence diversified fairly a bit from the simulation in PSU designer, nonetheless. That may very well be associated to the unknown current-suppling potential of the TV transformer–but I am starting to suspect that the 6DG6GT screens draw a lot much less present than famous famous on the info sheets…
A Redesign Partway Through the Project…
The design developed. Initially the primary filter stage was an RC (Resistance-Capacitance) filter, however that modified shortly. To get a clear sign, I might have to insert one thing like a 50 ohm, 20 watt resistor. But after I noticed the quantity of present wasted, I balked, and altered to an LC (Inductance-Capacitance) filter design.
Also a vital change–there was no B.2 provide in any respect, initially. I had deliberate that the preamp would run from the decrease display voltage (120V.) For the 12AX7, that is a fairly low working voltage. So the preamp provide was added.
The Inductor for the LC Filter
It helped that the gutted TV additionally included a (mighty large) inductor. It’s an unknown worth (inductors are measured in Henries), but it surely was matched with the TV energy trannie, so I used to be certain it could work–and it did. And actually, an LC filter does a way more environment friendly job of smoothing out the availability ripples in a single stage than an RC filter does.
Incidentally, it was the addition of the the LC filter (pi filter) that prompted me so as to add the standby switch–the preliminary inductance spike exceeds the 200V max of the 6DG6GT’s, by a good quantity. But through the testing section the swap wasn’t wired. There have been no unfavorable penalties and I am unsure the standby will probably be wired in. It’s sort of foolish, really–NOS tubes have been typically run at 150% of their rated voltage, so a brief spike at startup would not quantity to a lot…
Also changed–originally, the preamp plate provide was slated to run on the identical voltage because the screens. But it made sense to run the preamp at a better voltage. So a further RC stage (B.2) was added:
Preamp provide (B.2): As famous, this part was inserted AFTER the primary model was constructed. I began with a 220 ohm resistor for the RC filter, however settled on a 1K worth for a smoother provide. 1K did not drop the voltage a lot in any respect (which had develop into apparent earlier than when constructing the display provide.) It could be good to run the preamp tube straight off the B.1 provide, however preamps want one thing much less noisy…
Screen provide (B.three): Originally the second part of a two part PSU; in actual operation it did not match the Duncan PSUD2 software program very carefully. The simulator estimated the resistor for the final RC filter at 2.7K – three.3K. But through the construct the display voltage was a lot too high–over 170V. with substitution, the eventual 15ok worth was chosen, which positioned the screens at a pleasant 120V. A 20Okay resistor would most likely work simply as effectively… Surprisingly, the amp nonetheless functioned (poorly) with the preliminary excessive display voltages, and the tubes weren’t broken. Vacuum tubes are amazingly forgiving of abuse…
The PS voltage-dropping resistors are all 5 watt, though a three watt sort would have been tremendous for the B.three part (15ok.)
Regarding capacitance values, maybe 4 100uf caps are overkill, however they do the job. 100uF could be too excessive for a tube rectifier, however is not an issue with the SS bridge.
No “bleed resistor” has been put in. One quirk of this amp–the PS caps appear to empty by means of (cathodes to screens) the 6DG6GT tubes, probably as a result of very popular filaments. They maintain the tube internals heat sufficient after power-down that the tube retains functioning for a second or two. I do not know this for certain, however after I was experimenting with “triode mode” for the facility tubes (screens not linked to most important B.three), the caps have been NOT draining.
Regardless, ALWAYS examine the filter caps earlier than touching the internals.
Like the entire construct, the facility provide’s look is a bit inelegant, but it surely was modified a number of instances through the venture… Eventually it must be disassembled and reassembled in a smart vogue.
I’ve included a PDF on toroid transformer development, for the adventurous…