• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tubes to avoid

Status
Not open for further replies.
Jump to Latest
#81
After building a linestage using them, I'm no fan of the C3m - I've tried several different "brands", but they were all almost as microphonic as pair of 71A globe tubes (!!). If using, take that into consideration when designing your chassis. A top plate design gave me a loud "donk" "donk" "donk" everytime I turned the Goldpoint stepped volume control.

heh - the 71As were so microphonic, my friends could shout into them and the sound would be amplified to the speakers! Old Tyme Radio sound 😉
 
#82
mach1 said:
Absolutely, there isn’t a single DHT that can’t be significantly outperformed by an indirectly heated counterpart. A clear case of fashion over function dictating consumer choice.

If you are one of the poor deluded miscreants who was suckered into buying directly heated tubes I strongly advise you to list them on ebay for a token sum while they still hold some monetary value.
Click to expand...

Nah, just send them to me for a fitting and proper disposal.. I'll even pay the postage costs to ease the pain.. 😀

Not meaning to mislead anyone and in the interest of disclosure I'm actually a very big fan of DHTs, (SY knows this well enough) including some relatively big ones like the GM70, but I do it for the engineering challenge as much as anything.. They are pretty linear in a lot of cases because many of them were designed prior to the widespread use of global feedback, my system is loaded with them and its not a fashion statement, but fortuitous accident, and yeah they are microphonic as all get out.. I like the way they sound, and I use a lot of transformers too.. 😱

As some of my fellow audiophile friends will attest I am a resolution freak, and don't like overly warm sound...

I was also recently inspired by an ancient SEP design using 6V6/6j7 which I significantly re-engineered and which I think sounds pretty good actually, and am surprised at how little interest in it there was here. I think one forum member might be intrepid enough to build another one.. Could it be there is a whole lot more talk than action around here? (I might also be guilty of that charge.. 😀)
 
Last edited:
#83
Sch3mat1c said:
Do you mean "arbitrarily low distortion", or "distortion that looks the way I want it, not actually low but rather large" (say ~5% at full output)? The first one is easy, especially with multiple stages. The second one will take some looking, especially if you want a single stage.
Click to expand...

So far I have tested with 6SN7 at 6mA and got about 1% at 200Vrms with 800V B+ and CCS plate load driving a source follower. I'm trying to use a triode here for simplicity's sake. I am going to try 6BL7 next. However, 801A curves look beautiful. It is best to look at a type 10 data sheet since it shows low current areas much better if you want a look at what I am talking about. I'm just saying, I have never seen an indirectly heated triode with very little bunching at high voltages and low currents like this.

Sch3mat1c said:
A better solution is to simply avoid such high voltages in the first place (i.e., choose a better output tube). Then you don't run into such problems as rare and expensive output tubes, or equally rare drivers.
Click to expand...

But then I'd be stuck doing the same old boring stuff. I have always wanted to do a circlotron amp (not OTL) and that is where these experiments are leading.
 
Last edited:
#86
SHiFTY said:
I would say the most useless tubes would be the high voltage half wave rectifiers like 866A. These are mercury vapour filled and can be used if you are very keen, but for most intents and purposes are not any use to the audio DIYer..
Click to expand...

i've heard that these Hi-Current tubes work very well,

yet -

they 'must be started from 0 voltage, gently increasing to the desired value.'









SY said:
Unfortunately not true of the much-loved and now expensive 6LF6. I'm on my last set of four.
Click to expand...

how are those workin for you?

6LQ6
 
Last edited:
#87
SpreadSpectrum said:
But then I'd be stuck doing the same old boring stuff. I have always wanted to do a circlotron amp (not OTL) and that is where these experiments are leading.
Click to expand...

Two and three stage amps have been done to death: Preamp, splitter, output. Preamp, driver, output. Splitter, driver, output. Local feedback. Current feedback. Shunt feedback. Global feedback.

So stop doing them. Make a DC coupled amp. Make an interesting circuit, like a triggered oscillator, or class D amplifier, or thyratron sub amp, or... Pick some random transistor circuit, potentially a quite complex one, and see if you can build it in tubes with far less parts 😉

Tim
 
#88
andyjevans said:
I'd avoid any indirectly heated tube. Buy plenty of chokes for filament supplies and some strong shelving and stick to DHTs all through the amplification.

Life is too short to be wasting time with indirectly heated tubes! Except for the 6 months it takes to build a system full of DHTs of course......

andy
Click to expand...

Why?

Some of the best valve amps I've ever heard contain nothing but indirectly heated valves (KT66, KT88)

I'm generally somewhat underwhelmed by the over-rated directly heated amps I've heard. (300b, PX25, PX4)

And I think a Quad405 wipes the floor with all of them..............

And why are you displaying a UK flag and calling valves 'tubes'? 🙂
 
#89
smoking-amp said:
The old SCR inverters had a commutating inductor for each 180 deg. phase pair so that the opposite phase created a momentary reverse voltage spike to turn off the 0 deg. phase SCR, and visa-versa. If some heavy load disturbed the commutation you would get some real fireworks.
Click to expand...

I am familiar with this concept (as used in DC-AC inverters) but I'm a bit skeptical about the application of same method in audio (I believe the LC combination in inverters is calculated for predetermined switching frequency only rather than a span, say 20-80 Hz that subwoofer amplifier would operate in).

On top of that, wouldn't such a complex circuit negate the benefit of using these elements (= more power in same volume) ?

Does anybody know of any actual examples of audio circuits employing thyratrons/thyristors ? Google didn't cough up anything useful, thyristors are usually used as "crowbar" devices for protection in audio, not as the amplification elements and this reflects on search results 🙁 I'd love to see some schematics so I could give this a try.
 
#90
Back in my running disco days (late sixties) I built several high power amps round 807s, basically because they were dirt cheap at the time. I would definitely not recommend this for optimum audio quality, but we did put them out on display, where they made a very decent light show at the same time, with the gently red glowing anodes and the blue ionic glow modulating with the signal (or the RF instability)

And you can get a lot of power out of a pair, though they really needed fixed bias.
 
#94
Re: Arnulf

I've never heard of using thyristers for audio. Only seen 60 Hz motor speed contols using them. The turn off time would be critical for the high frequencies needed for audio. Ferrite inductors would be essential. Would have to check the thyrister data sheets. If one were going to try, I would go for a fixed square wave frequency set, using two phase controlled square waves versus each other, H bridge type setup. That avoids the crossover distortion inherent in PWM designs and alleviates the turn off time problem some too. I think Crown Audio uses such a config. with Mosfets. Patented too.

I would also use a linear corrector amplifier summed with any such switching concoction to get better results. Or use the switchng stuff for variable rail voltages for a linear amplifier. Maybe keep the switching stuff for a Sub amp as well.
 
#95
Sch3mat1c said:
Two and three stage amps have been done to death: Preamp, splitter, output. Preamp, driver, output. Splitter, driver, output. Local feedback. Current feedback. Shunt feedback. Global feedback.

So stop doing them. Make a DC coupled amp. Make an interesting circuit, like a triggered oscillator, or class D amplifier, or thyratron sub amp, or... Pick some random transistor circuit, potentially a quite complex one, and see if you can build it in tubes with far less parts 😉

Tim
Click to expand...

Go easy on me, Tim! This is my first amp, after all. There will be plenty of time for a sub amp later. Thyratrons do look pretty cool...
 
#97
smoking-amp said:
Found it. It's a little different than I remembered it, but this avoids crossover distortion in a switching amplifier. Crown's "Balanced Current Amplifier"
Basic diagram is top of page 4.

http://www.crownaudio.com/pdf/amps/bcapaper.pdf

Getting this to work with Hydrogen Thyratrons though............
Will need some Mosfets I think, just to turn them off.
Click to expand...

Would mercury be too slow? Hydrogen thyratrons look pretty cool during operation, but not as cool as mercury.
 
#98
SCR Inverter

smoking-amp said:
Re: Arnulf

I've never heard of using thyristers for audio. Only seen 60 Hz motor speed contols using them. The turn off time would be critical for the high frequencies needed for audio. Ferrite inductors would be essential.
Click to expand...

Ferrite? That'd be hugemongous! We need good old IRON for this one! 😉

See, I told you the junkies would approve...

Here's a basic test I've done:

Schematic:
http://myweb.msoe.edu/williamstm/Images/SCR_Inverter.png

http://myweb.msoe.edu/williamstm/Images/SCR_Inv1.jpg/img]

The SCRs are on their original heatsinks, actually. They came from a motor drive. Probably inverter grade 800V 20A, t_q < 20us or so. The circuit generates alternating pulses for each edge of the input (rising pulses the top SCR, falling pulses the bottom SCR), so you simply feed it PWM and away it goes. As you can see, the control circuit is very simple -- it would fit in a few square inches if surface mount parts were used. The SCRs and chokes dominate the physical size.

I had this running up at a few kHz, which would have to be very well filtered for audio use, but that's just more iron and caps -- an engineering issue, not a fundamental problem.

[IMGHTTPDEAD]http://myweb.msoe.edu/williamstm/Images/SCR_Inv2.jpg[/IMGHTTPDEAD]

I believe this was no-load, so the SCRs aren't actually staying on -- they act more like BJTs, because the load is too little to keep them latched on. Capacitance keeps the voltage stable until the next SCR fires, pulling it back, etc.

This is the important one:

[IMGHTTPDEAD]http://myweb.msoe.edu/williamstm/Images/SCR_Inv3.jpg[/IMGHTTPDEAD]

Under load (an amp or so), it stays on, and the crummy PSU shows lots of ripple. The supply choke shows commutation bounce, as it should, and the commutation snubber is working exactly as intended. Bottom trace is snubber current: it's a single cycle sine wave, actually pulling current negative (about -5A peak), ensuring both SCRs are reverse biased during commutation.

Transient zoom:

[IMGHTTPDEAD]http://myweb.msoe.edu/williamstm/Images/SCR_Inv4.jpg[/IMGHTTPDEAD]

(And yes, my current transformer was wired backwards. Oops. The initial peak is supposed to be positive.) Output voltage is shorted to zero, while the SCRs/snubber diode carry the resonant current. This provides enough time in reverse-bias to turn off the other SCR. The gate drive pulse must be longer than this event, otherwise *both* SCRs will end up off, and the output goes open circuit!

When current goes positive again (at about 3.2 div from the left), load current forward-biases the circuit again and voltage begins to rise. The supply choke charges the snubber capacitor.

The limiting factor in using this inverter for audio is, each SCR has to be on for a minimum of the commutation time (~100us in this case). This means an absolute maximum operating frequency of 5kHz, at 50% duty cycle and no room to vary. At 1kHz, duty cycle can run 10-90%, which is fairly reasonable, and only reduces your peak voltage output by 20%, without impacting ultimate power output substantially.

Downsides to common thyratrons include the very slow deionization time. Xenon and mercury are both typically 1ms. The commutation network has to ring at least this long, to ensure turn-off. This is fine at line frequency, when the whole thing is reverse biased for 8.3ms (or longer in some countries). I don't know if they ever made intermediate grade thyratrons; I don't think I've heard of any besides xenon, mercury and hydrogen. Hydrogen thyratrons are still pricey!

Tim
 
Status
Not open for further replies.
Top Bottom