JimL

So, is this the spritzer mod converting from local feedback to global feedback?

Back when Monster cable was first selling their cables with 3 sizes of wire (fat for the bass and thin for the treble), Frank van Alstine offered to wire the thin ones to an AC plug and plug it in while a Monster salesperson held on to one wire in each hand. For some unfathomable reason, the salesperson declined.

For some reason thoughts of flies and sledgehammers comes to mind.

kevin gilmore on 19 Jul 2015 said: "...by the way, so far, I can't even get close to these numbers with a shunt regulated supply, the noise rips right thru the main current source." I wonder if using a cascode current source would help? Just a thought.

Kevin Gilmore said, "yes, buy a bunch." I assume you meant: yes, BY a bunch, i.e. a lot better. On the other hand maybe you did mean BUY a bunch. Oh well.

The schematic seems to be missing the ground output, which I assume is straight line down the middle.

I'm probably being old fashioned, but those look like speakers from a really bad acid trip.

Assuming the shunt resistance is 82k, and that it's a 1000:1 HV probe, the series resistance (between positive terminal and probe tip) should be somewhere around 82 megohms - most standard ohm-meters will show "infinite" resistance.

kevin gilmore said: "this is the new power supply schematic. i like it much better than the previous version." And I assume if you ground the input side and the output side at the junction of the two 470 uf caps you can use this PS for the KGST, KGSS, KGSSHV Carbon, BH, etc - with appropriate adjustment for the output voltage? BTW, shouldn't R8 be 890k nominal? Will need to find a 1 kV rated resistor for that one, or else three 350v resistors in series.

Well, the smallest commonly available trim pots appear to be 10 ohms, so you would have the fixed resistor plus the trim pot in parallel with a small value resistor, e.g. for a 1 ohm adjustment range, 1.2 ohms would give you an overall range between 0 and 1.07 ohms. The adjustment won't be close to linear - about 1/2 the adjustment range would be in the lowest 10th of the trim pot range with hardly any change at all over the upper half of the trim pot range.

If you use a 100 ohm fixed resistor plus trim pot, you can use the 100 ohm resistor as a sense resistor for the current. Then 10 mA = 1 volt.

That looks OK. The only thing I would suggest is to make a test jig to check that the current source actually puts out 10 mA - there is some variation among DN2540s. Or you can use a fixed resistor plus trim pot to adjust the current to 10 mA in circuit.

This reminds me of the nuclear arms race in the 50s between the US and the USSR...except there is no USSR. Comparing KG's amps to any other designer's is like comparing the US nuclear arsenal to, say, France.

Thanks, guys!

Another way to do it without a relay is have the filament switch with one set of contacts to turn on the heaters, and another set in series with the HV switch.

nopants said: "the real kiss would probably be a separate switch" Yup, that's what the old Stax T1 had, a pre-heat switch and a power switch. Now what you want with that system is an interlock so that the power switch won't turn on the HV unless the heater switch is on. Since I've never seen a T1 in the flesh I don't know if it had that built in or not, but it shouldn't be difficult to do, just have an AC relay on the primary of the HV transformer, in series with the power switch, that is controlled by the heater switch. Laowei said: "The 6S4A used in the KGST has a controlled delayed warmup for the heater. The A revision was released with this engineered in, to lower the turn on power surge in TVs." Shows how important tubes were in TVs in the 50s - versions were designed specifically for them. For example, the 6SN7GTA was redesigned to the GTB version which had the controlled warmup. Actually I believe the other big reason for controlled warmup was so that the heaters could be wired in series, e.g. use a 24 volt filament supply and wire 4 6V heaters in series - only works if they all warm up at the same time, otherwise the voltage isn't distributed evenly between the filaments in the string.

I don't think there is anything wrong with the idea of feeding the filaments with a constant current source and using the voltage generated across the filaments to trigger the HV - actually as the filaments heat up their resistance increases, and hence the voltage across them increases, so the idea would be to trigger the HV when the filament voltage rises to a certain value. But, it is probably significantly more complicated to implement than a simple time delay, for which reliable, time-tested circuits are already available. Why go with a more complex method when a simple one does the job, aka KISS?

So I have a similar issue with my SRX Plus, which uses a hybrid tube/solid state rectifier bridge. The positive side is done by the tube and the negative side is solid state with a delay and soft start circuit to roughly mimic the rectifier tube turn-on time. The output voltages go though some gyrations on turn- on, however the voltages on each channel track each other closely so there isn't a big differential voltage (which is what the headphones see as signal), and there isn't a big bang on turn-on - the gyration is subsonic anyway. Tube filaments generally warm up in 10-30 seconds so that degree of delay should be enough - tubers actually reach steady state faster than transistors, which turn on quickly but then take awhile to heat up their heatsinks to operating temperature. That is exactly what sorenb is seeing with is KGST. When it turns on, the output is connected to the positive rail through the current sources, but the output tubes aren't conducting because the filament is cold. Once the filament warms up enough the output tubes start conducting and the output voltage drops and after another 10 seconds the tube is fully operational and the output drops to zero. If you disconnect the servo, you can see that there is a relatively fast voltage change as the tubes warm up, followed by a slow drift as the solid state heatsinks reach operating temperature. Now, if you don't want to be worried about voltage gyrations at turn-on, one way to do it is to have all the filament supplies on one transformer which turns on immediately, and the HV supply from a second transformer with an AC delay so it doesn't turn on until the filaments warm up, then all the high voltages turn on at the same time. Does need two transformers, though. Stax did it manually with two power switches in the T1S, one for heater warm-up and one for HV, which was labeled as the power switch.

I heard from AudioXpress, they plan to publish my article in two parts, the first part in October which analyzes the SRX circuit and modifications (which I've already discussed above in more detail) and the second part which covers the shunt regulated power supply. In my build I used a surplus transformer, hybrid tube/SS rectification and choke input raw PS (mainly because I'm twisted), and a shunt regulator. The shunt regulator is based on ideas from John Broskie's TubeCAD shunt regulator article of July, 1999 and uses his TL431 circuit, but with stacked MOSFETs instead of the pass tube, an RC network for stability of the TL431, and connects the regulator between B+ and B- as suggested at the end of the article where he discusses bipolar supplies. The latter helps keep signal currents out of the ground line. I also used a current source instead of the series resistor. The measured broadband noise between B+ and B- on my build is less than 4 mV and the voltage varies less than 0.2 volts between cold turn-on and warmed up.

Very nice!

Point taken - of course my phono preamp only puts out a few measly volts.

HA! Wimp! My tube phono preamp has over 100 joules in the power supply. Heart defibrillators run between 120 and 300 joules. BTW, here's another enclosure that should work: http://www.ebay.com/itm/141252736876?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Other than European equivalents such as the ECC83 or the British CV4004 and M8137, there is the 7025 which was a premium type designed for low filament hum. There are also rare industrial types such as the 12AD7, 12DF7, 6681 and 7729. Some tube rollers will also use 5751 tubes which were originally designed for the military as 6SL7 equivalents. As for the 12AU7, industrial tubes that can be substituted include the 5814,5963, 6189, 6350, 6680 and 7730. The 6CG7/6FQ7 can also be used other than the fact that they are 6 volt filaments only.

Glad to hear you like it!

There already is a trimmer plus fixed resistor in the circuit, the issue is choosing a fixed resistor so that the trimmer is in range. Whoops, KG beat me to it.