Hifi, why bother?

[Kalee20]

For some reason, the attachment from previous post, didn't load!

[Amie]

Hello Peter

I have a pair of old 6v6g, they're definitely larger structures than the noval. I also have a few Brimar 6v6gt pulls in dads box of come in handy junk valves that look like later production, and although they are also coated glass do definitely look the same inside with just larger mica?

I suppose if you were going to start making a smaller valve with the same spec., then you may consider using the smaller structure for both to save some money.

Circling back to Andy's amplifier ideas, that push pull guitar amp that I had played with recently that had the 6BW6 op valves, had a decent sized output transformer and was an "ultralinear" but had a couple of controls that altered the bias on the output stage that were on the front panel, one allowed the amp to run into starved distortion at low level, the other altered the balance a little. They did alter the character of the sound considerably, so weren't just a sales gimmick for guitar use

Thanks,
Amie.

[Mike Watterson]

The heater power is related to the electron emission needed. The anode and even g1 is too far away. Sometimes the visible outer "can" isn't the anode but a screen. Also to a small extent the insulation, some valves can only be used near the earthy end of a series chain. A power rectifier on the same heater supply as the rest of the valves or the mad 115V heater rectifier valves used in some USA mains/battery sets needs more insulation. Compare warm up times of a parallel and series mains radio. Some series sets are nearly a minute. A direct filament needs much less power, hence older rectifiers using a filamentary cathode and a separate winding. Most direct Triodes started off as battery valves, at the output stage the hum issue isn't so severe on a direct triode. Hence battery valve sets are basically instant on, when on battery, but ones with valve rectifier for mains can seem slow.

You see the variation in filament current needed with anode current required most clearly on the Telefunken Y8 metal can parallel only battery valves from the 1940s. They were far more "efficient" than the RCA B7G, because the 1T4 filament current is artificially high at 50mA to allow series operation.
Some mains signal valves are deliberately 6.3V and 300mA, or two 6.3V @ 150ma (for 12V cars OR 300 mA series chains to suit series or parallel.
It's obvious on the audio power out and power rectifiers, over 1A on 6.3V for some (1.5A for EL34 and 1.27 A for KT66), The UL84 is 100mA and 45V (4.5W), the EL86 is 6.3V 0.76 A, surprisingly 4.8W, apparently the same valve with a different heater. I'd have thought the UL84 would need more. Comparison of heater power of a selection of series and parallel power valves with otherwise the same spec would be interesting. You'd imagine the UL84 is the 100mA version of an EL84, but allegedly not.

I'd thought a stereo amp with no mains transformers and PL509 or PL519 in push - pull was an interesting idea to save a big transformer. Four of those alone is 160V @ 300mA for the heaters.
You could have switches to triodise, feed screen grids from audio transformer or not, vary cross-over and feedback, maybe even parallelise for more 2nd harmonics instead of push-pull (which cancels most of the second harmonic). Lots of scope for experimentation.
Use a voltage doubler to have about 600V HT, or a tripler for 900V direct from the mains?

I shall dismember and photograph more dead valves

[Kalee20]

Amie wrote:

I suppose if you were going to start making a smaller valve with the same spec., then you may consider using the smaller structure for both to save some money.


Seems a good idea, economies of scale - assuming everything equal.

On the other hand, the 6AQ5, with its piddling little anode, is going to run hotter and have a shorter life than the 6V6 electrode structure. Unless... The 6AQ5 anode is made of a purer, more expensive material, that exudes less gas when heated (or it undergoes extra heat treatment to remove absorbed gas) so it can run hotter without problems  - either of which would shift the cost balance. I guess we'll never know   

Amie wrote:

Circling back to Andy's amplifier ideas, that push pull guitar amp that I had played with recently that had the 6BW6 op valves, had a decent sized output transformer and was an "ultralinear" but had a couple of controls that altered the bias on the output stage that were on the front panel, one allowed the amp to run into starved distortion at low level, the other altered the balance a little. They did alter the character of the sound considerably, so weren't just a sales gimmick for guitar use.

Interesting! Did the amplifier have NFB around it all - which would have reduced the distortion up to a point, but also made the onset of severe distortion more acute? Or was it open-loop? I'd guess the latter.

Increasing bias to increase distortion has the benefit that standing current will be reduced, giving a somewhat easier life for the valves. Win-win!   

Michael Watterson wrote:

The heater power is related to the electron emission needed.

True, but it boils down to keeping the cathode at the right emitting temperature. Cathodes emit so many milliamps per square mm of area. You want more current capability? Then you have to increase the area. And more area means more heat loss, so you have to up the heating power.

However, a nice snug anode, shiny on the inside, will reflect radiant heat back to the cathode, meaning less is lost - so the heater does not need so many watts to maintain temperature. Hence my thinkings!

Michael Watterson wrote:

The UL84 is 100mA and 45V (4.5W), the EL86 is 6.3V 0.76 A, surprisingly 4.8W, apparently the same valve with a different heater. I'd have thought the UL84 would need more. Comparison of heater power of a selection of series and parallel power valves with otherwise the same spec would be interesting. You'd imagine the UL84 is the 100mA version of an EL84, but allegedly not.

The different heater power, UL84 and EL86, which have same characteristics, is a surprise!

It's true that UL84 and EL84 are different, quite apart from the heater. It's a mistake to think that if only the 'heater' letter changes, then characteristics are the same apart from heater. This is a good counter-example, others are DF91 and EF91, the first being vari-μ and the second, straight. Or (even more extreme), DM71 and EM71, which aren't even the same base!

[Amie]

Hi all

Interesting! Did the amplifier have NFB around it all - which would have reduced the distortion up to a point, but also made the onset of severe distortion more acute? Or was it open-loop? I'd guess the latter.

I will have to look at the notes I made, however it did have NFB but the thing was quite clever as the "bias" pot's were each ganged with a switch, (just as a volume or tone control would be with a mains switch), the switch would shunt the NFB when you wanted to increase the bias. Same also applied to the AC drive balance, (I will call it), with both controls turned off it operated as a normal UL amplifier.

I did find out that it was a manufactured item that had been re cased, (I had a friend re-re-case it with two cheap 60W 5" "soundlab" twin cone drivers to replace the knackered 10" job  ) German made, (Kohler?) as a universal instrument amp actually, not just for guitar.... As you know Peter, I was persuaded to sell it.

Thanks,
Amie.

[Kalee20]

...you could add a control that lets you mix between the "enhanced" sound from the valve amp and the "clean" sound before.... It would give you a convenient front panel control which you could label "Niceness", allowing you to choose your poison, according to the recording...

Circling back to Andy's amplifier ideas, that push pull guitar amp that I had played with recently that had the 6BW6 op valves, had a decent sized output transformer and was an "ultralinear" but had a couple of controls that altered the bias on the output stage that were on the front panel, one allowed the amp to run into starved distortion at low level, the other altered the balance a little. They did alter the character of the sound considerably...

I will have to look at the notes I made, however it did have NFB but the thing was quite clever as the "bias" pot's were each ganged with a switch, (just as a volume or tone control would be with a mains switch), the switch would shunt the NFB when you wanted to increase the bias. Same also applied to the AC drive balance, (I will call it), with both controls turned off it operated as a normal UL amplifier.

Looks like the amplifier could easily have satisfied Mark's suggestion as a 'Niceness' control, in fact if you had 3 controls, one for bias, another for feedback, and a third selecting tapping points on the O/P transformer for screen-grid feeds (or even a sort of twin-ganged Variac type thing, for continuously variable from straight pentode to triode operation), there's enough combinations available to suit anybody - a knob for 'Niceness A,' 'Niceness B,' 'Niceness C.'

It's less benefit with beam-tetrodes like 6BW6 than pentodes like EL84, but a side benefit of UL is the added contribution to the output power from the AC component of the screen-grid current.

With feedback control turned back a bit to give less NFB and higher gain, bias control turned up to give more bias hence crossover distortion, and UL tapping appropriately set to give more power and gentle clipping, there'd be soft fuzz, hard fuzz, and plenty of loudness. So, Andy, how about giving it a try and report back?

[Amie]

Hi.
I don't know if the 6BW6's were originals. Someone here linked me to a circuit diagram of something similar, (the German made amp), I'll try to find it and see, would have thought the valves would be Telefunken's or Valvo make, not Brimar 

Amie.

[Mike Watterson]

At least after WWII:
The German Schaub and Lorentz (spelling?) were owned by ITT-STC who I think owned Brimar. Mike / Crackle would be the expert on that.
Valvo was owned by Philips.
Indeed many Austrian and German radio makers used Telefunken.