Tube preamp power supply

[Analogy]

Maybe i'm missing something here, but what's the problem with using a power supply of around 30 volts for valve preamp equipment? I've previously designed a simple mastering preamp consisting of:
opamp input buffer + balanced MIC input (differential opamp style) -> 2x12AX7 (4 stages) voltage followers class A config -> 10 band active EQ -> opamp output buffer.

Works great, sounds GREAT so what's the problem? A hell of a lot easier for me since decent audio transformers are almost impossible to find here without paying an arm and a leg - not to mention the high voltage safety issues especially when it comes to small time manufacturing.

I'd appreciate any input any of you have so go nuts!

 

[gyraf]

Hi Analogy,

Welcome to "the Lab"..!

- If it sounds Great to you, then there is no problem at all. Really.

The problem for me is that running 12AX7's on low voltage dosen't sound good at all. Yes, you will get tube'y distortion products - but this distortion is certainly NOT why we like tube sound.

Even with tubes specifically designed to work on low voltages (like the ECC86 and EF94) there is sonically a very long distance to the properly high-voltage-fed tube circuits.

I've experimented quite a lot with the low-voltage approach, but had to conclude that you won't get any sonic advantages that you couldn't get by replacing the tubes with FET's.

On the transformerless approach - that has no relation to what voltage you chose to run your tubes at - there is common consensus that decent quality audio transformers are clearly a sonic advantage.

Jakob E.

 

[Analogy]

Thanks gyraf, really should have looked around for a place like this much sooner but glad i found it anyway

Seems like you've got a point, like I said I think it sounds great, but that could be simply due to active EQ a buffering with very selectively chosen frequency response components - honestly as far a valve qualities go it doesn't quite compare to my Fender blues deville :green:

So in your or anybody elses opinion is there any way around the need for input and output transformers? Ignoring impendance matching for a moment (I hope this can be handled by input and output buffering), would a simply resistor voltage divider be enough to drop the voltages to a usable level?

The only very readily available audio transformer i've found here is this one:
http://www1.jaycar.com.au/productView.asp?ID=MM1900&CATID=19&keywords=&SPECIAL=&form=CAT&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceMin=&priceMax=&SUBCATID=378

Possibly totally useless but worth a look?

I've also heard of using standard power transformers as audio input and output stages, any thoughts?

Thanks

 

[gyraf]

That jaycar transformer won't do you any good - it is a 100v-system speaker distribution transformer, ment for low-Z work.

Decent, yet affordable audio transformers are e.g. the OEP range available from Farnell or RS components. I guess that you're located in Australia from your supplier choise? A local her (and there) - Kev Ross - may have some ideas for audio transformer sourcing down under.

would a simply resistor voltage divider be enough to drop the voltages to a usable level?

Yes and no.

You won't need it, if you dimension your tube stages properly, using e.g. cathode followers for output stages. Maybe there's something you're unclear on here: Even though "real tube" circuits run on very high supply voltages, that dosen't mean that their generic AC working voltages are high as well. They're mostly high impedance internally though - that's why low-Z output stages (as cathode followers) are often required for external interfacing.

I've also heard of using standard power transformers as audio input and output stages, any thoughts?

I think that that's a myth. There have been projects flowing around using powersupply transformers as tube POWER AMPLIFIER output trafo's - but the types that I have tried simply dosen't work even close to acceptable.

Don't confuse power amplifier output transformers with mic/line/interstage transformers. They're quite different.

And on line- and mic levels, power transformers will probably do even worse.

So I think you will need to find yourself a source for "real" audio transformers.

Jakob E.

 

[PRR]

> what's the problem with using a power supply of around 30 volts for valve preamp equipment?

It works.

To a first approximation, tubes are linear resistors. Feed them less voltage, they take less current, you get MUCH less power out, but gain stays similar and (allowing for the reduced clipping level) the THD versus output stays similar.

For a better estimate: most tubes' resistance has a kink around 50V. At lower voltages, the plate resistance doubles or triples, further reducing your available output power and generally putting you at an awkwardly high optimum load impedance.

And the linearity changes: they don't have the "tube sound", they start to act more like JFETs.

A tube's plate resistance is set by several things, but the hard-nut is heater power. Lowering the amplification factor lowers plate resistance and somewhat increases the perveyance. But for a real change you need a BIG cathode, which means a BIG heater. So to get a low output impedance you wind up putting much more power in the heater than into the plate or load.

If high voltages are much more expensive than low voltages, this may be acceptable. In 12VDC cars they had some tubes that ate 12VDC at 5mA in the plate to get 5mW of audio power output, while slurping 2-3 Watts of heater. Total power efficiency was awful, but eliminating the 250V supply saved a LOT of cash, enough to pay for a very expensive new-fangled 12V transistor to boost the 5mW tube up to 3 Watts in the speaker.

But as a general thing, best cost balance is with tubes working at 200V and up. Above 300V on most radio tubes, the heater power becomes "small" compared to plate and output power. That does mean transforming the 2K-20K plate impedance down to typical 600-8&3937; loads, but tubes and transformers DO go together.

 

[Kev]

yes
welcome to the lab.

... keep asking questions Analogy and don't let it stop here. You are in very good hands here with these two guys.

Don't worry about components sourcing yet and don't get excited anout a component just because Jaycar have it on the shelf. ( I know that is easy to do ).
Get a design layout thought out first and then worry about components and availability. We can fix that later.

I would like one thing from you ...
... define simple mastering preamp ??

 

[Analogy]

[quote author="Kev"]... define simple mastering preamp ??[/quote]

Yea good question, basically I decided to call it that because it has some nice features (10 band EQ, balanced in and out, valves :grin: ) and anything I put in came out sounding a hell of a lot more powerful - tight bottom end and lots of it if desired, more 'active' in general and quite crisp highs. Not sure if it quite qualifies as a mastering preamp but sounds good to me and those who heard it.

As for the questions, don't worry there's a lot more where that came from. In fact, here's one now:

As a thought, is there any issues with the following configuration:
Opamp input buffer -> HV (~200V) 12AX7 gain stage -> Resistor voltage divider -> Opamp output buffer
*Note: No transformers used

Yes I like my opamps, thankyou. Personally I find them a great way to condition my incoming / outgoing signals and they seem quite happy to operate under any conditions i've put them. Either way, its simply a matter of choice.

After spending some time with a 12AX7 datasheet i've seen that yes this tube will operate much more linearly at a voltage of ~200V with a grid voltage of ~-2V. It seems the tubes response with a plate voltage below 100V is somewhat undetermined and as i'm sure it will work regardless, i'm after designing some really high quality equipment in which case i'd be best off operating them at 200voltsish.

In this case, i'll simply be using capacitive coupling between stages opamp and HV tube stages to block DC voltages - seems ok to me? since opamps were not exactly available in the early 50's did designers simply choose transformers as a means of isolation because of their availability at the time or is there some fantastic qualities i'm not quite aware of?

Cheers! :green:

 

[Winston OBoogie]

...