Phantastron

Legal Note: Electricity is potentially dangerous, no matter the voltage or current. Although these circuits are designed to be as safe as possible, anyone can still harm themselves with anything electrical, low voltage, medium voltage or high voltage. Therefore, Electric Western and Lorin Edwin Parker (Proprietor) absolve itself/himself of all responsibility for injury incurred while working on these projects.

This is the power supply devised for the Phantastron and other instruments. It is designed to be small, self contained and safe. The trade off is that it is limited in the amount of power it can supply, and requires a 9V or 12V wall wart. Nonetheless, one of these supplies can comfortably supply 2 to 3 tubes (or more), depending mostly on their current requirements (especially the heater current - do not exceed 1.5A of current on the heaters).

Small - Safe - Doesn't make much EMI/RFI - Low Hum. Can be used with my pre-amp below or many of Eric Barbour's, Ken Stone's and others' designs. - You can breadboard with it. Great for DIY, experimenting, synths, guitar pedals. Lightweight. No need to mess with AC mains.

B+ doesn't supply lots of current (ie. no power amps). With high gain circuits transistor & zener filter / regulator introduces a little hiss. Total heater current limited to 1-1.5A depending on regulator used. Poor for audiophile usage.

Zeners should be rated above 5W in total. I use 2 5W zeners for a total of 10W. Other series voltage combos will work (ie. 25,25,25,25...etc)

B+ power is isolated and protected by the fuse, the PTC resettable fuse and the isolation of using 2 transformers. With a PTC that trips at 50mA, if B+ is shorted out or miswired, the circuit shuts down B+ 104V power as the PTC stops flow of current above 50mA. The PTC is resettable, so all you need to do is let it cool down, fix the problem and turn it on again. Since current is held below 50mA by the PTC, the 125mA fuse is unlikely to break, but I added it anyway as a protection against unexpected ills or in the case the PTC did not trip.

Although no power supply or electricity is fool-proof safe, 50mA of 104V is not likely to kill you or even shock you. If it does, it stings, so don't do it, but I'll take the mystery out of it -- it's like a prank "joy buzzer" with a little more kick. Needless to say, always use one hand when handling the B+ supply, that way any shock just hits your finger and is less likely to go through your body to ground. Your body is a giant resistor greater than 10 megaohms, and with shoes, even more... So ohm's law tells us that even if the 104V at 50mA entered your body and went to ground through your feet or elsewhere, the current accross your body, would be 0.0104mA at the most (104/10,000,000). The US government states that 80mA is generally the required current to prove truly hazardous in the body.

Back to the circuit -- The 47uF capacitor smooths some of the ripple out. The 100k resistor is a "bleeder" draining the capacitors when you turn off the supply so they don't store 104V. Next is the PTC explained above. Then, we get an RC filter consisting of the 10K resistors and the other 47uF cap. The resistor and the capacitor combo form a low pass filter, removing more ripple than a capacitor alone. Then, the TIP47 transistor is biased by the other 10K resistor, passing on the DC while only dropping the voltage 1V or so (like a diode). The ripple appearing at the collector of the TIP47 contains about 1V of ripple, however the RC filter reduces this to about 10mV which apears at the base of the transistor. The result is, that the DC out of the emitter of the TIP47 is filtered following the current at the base. The voltage drops about 1V through the transistor. Finally, 2 51V Zener diodes set the filtered DC (which is at about 120V) to 104V.

So, the TIP47 acts as a sort of active filter -- reducing ripple, but without dropping DC voltage and current or creating lots of heat loss. The advantage is that the board can be smaller (smaller capacitors) and lose less B+ from the transformer as heat - the current the small transformer can provide is already small, so it's nice to save it. The tradeoff, and the reason this is NOT an audiophile supply is that the TIP47 and the zener diodes introduce their own noise. The 60 cycle hum may be greatly reduced, but high frequency switching noise is added. Furthermore, if you increase the current draw on this filter, the noise goes up and regulation can suffer.

However, providing the 1 to 10mA needed for synthesizers, expriments and small preamps, the supply has a ripple of 1mA to 5mA peak to peak, including noise, which is impressive.

The A+ supply is pretty straightforward. A bridge rectifier to make DC directly from the 9V transformer is filtered by a capacitor and then regulated by an LM7806. Since the A+ supply may need to provide up to 1.5 amps, attach it to a heatsink (a large heatsink if current draw for the heaters is over 1 amp).

You can also mod this supply -- like using a 12VAC wall supply and a 12V transformer on B+. You could also use a LM317 for adjustable A+ supply or an LM7812 for 12V heaters.

Buy The Supply Kit From Electric Western -

Tube Power Supply Circuit Board & Parts with 9VAC 1000mA "wall wart" transformer included -- $48.00

Tube Power Supply Circuit Board & Parts WITHOUT 9VAC "wall wart" -- $39.00

The PCB is double sided, lead free and designed with ground planes to reduce electro-magnetic hum. The transformer, fuse and PTC are all selected to work well together. The on-board transformer is an encapsulated toroidal transformer which almost eliminates magnetic leakage from getting into your audio circuits.

It's very "tubey" sounding, uses an old classic pentode (which is readily available and affordable) and adds plenty of 2nd harmonics. It's based around a 6SJ7, but a 6au5 (miniature) or a 6J7 could make nice tubes here too.

With this preamp, the screen is being held rather low in relationship to the plate -- (in my test, screen=14V and plate=63V). This allows a large voltage gain and swing, however the transconductance is low (impedance is rather high). It sounds good plugged into most solid state inputs (the input op-amps can acomodate for the high impedance easily). It would also work well feeding the grid of another tube.

It has a frequency response of about 70Hz (-3db) to 19kHz. It is resonant and peaks a bit in the mid range 200-400. It sounds really nice on vocals, guitar, strings, etc.

If you really crank it you're going to hear some noise - a little 60 cycle hum and a some high frequency EMI stuff. But, at comfortable volumes it is fairly low noise. Plus, I measured the signal to noise ratio on a digital frequency analyzer and it was about 30dB at a gain of 50. That was considered quite good in the days of tape machines and high noise floors -- only with digital equipment do we eventhink that an SNR of 90dB+ is possible. Plus, we're not trying to get clean here -- we're going for warm, tubey, maybe even a little distorted...

Here are some sound samples of guitar recorded directly into the computer (through pre-sonus firebox) - one without the breadboarded preamp shown, and one with it. Please excuse the MP3s, but you should be able to tell the difference. Both clips were normalized after recording (ie. the guitar without pre-amp was very low level, so Ihad to turn it up in software):

Yes, you're going to hear 60 cycle hum. Did I claim this was audiophile? It's an experiment, and all you need to do to lower that hum is add more filtration to the power supply (replace the 47uF caps with, say 470uF and add a choke -- but that's a different subject).

Examining output with the scope, the 60Hz hum is about 9mV rms. Since the output is 10V rms, this is a pretty good figure. I can't even hear hum when I turn off the music playing into it.

I created the entire circuit on a Breadboard and it still sounds really smooth and warm. On a turret board, PCB or directly wired, it could sound better.

Frequency Response (small dips seen are resultant of the digital function generator, not the pre-amp). You can see the curve on this plot with the frequncy anylizer on "peak hold":

The frequency response curve can be modified with a few simple changes... I'll leave it up to you, though to emphasize the lows or the highs or whatever you like. It sounds pretty decent to my ears for a breadboarded circuit with a tiny supply using only small filter capacitors.

DIY is rad! Experimentation with tubes and synthesis doesn’t have to be an obscure, difficult or expensive art. Everyone deserves a chance to make their own instruments with these fascinating methods and learn the ways of antique audio and pioneering sound.