Choke and sag

[Bruce / Mission Amps]

Hey Bruce, that's interesting. What kind of input and output voltages?

jamie

Well it has been a few years since I've even talked about these things but the second to the biggest amp I made was using two Eimac 4-400a tetrode tubes with the plates at around 2500vdc-3000vdc and a 750mA HI-V supply ... the screens were running about 550-600v. Yeah, force air cooled.
Of course the plate transformer was a totally different power transformer then the screen supply tranny.
I think I still have the Hi-V power transformer around here somewhere... it's too big to do anything with though.
The filament tranny was a separate unit too, which was about the size of a small bread toaster, 5v@40A.
The two 4-400a tubes (class AB) made an honest 650-750 watts at 1.85Mhz and 3.75Mhz.
The RF deck was too big and the power supply built in a different cabinet...
I later downsized and went to single 3-500Z triode with pretty much the same kind of power supply, just smaller. It only made about 400-450 watts output.., but that amp was a lot easier to handle.
Both those amps needed a boat load of drive signal to operate though... like 35 watts to 80 watts of drive and ran on a home run, 220v line from my panel to my radio shack.
I often thought about building a big audio amp using transmitting tetrodes or triodes like these.
But who really needs it.

[Steve Conner]

Ha! You don't even have to dredge the ham radio mailing lists, just ask Bruce.

I think that's a resonant filter? The caps across the choke tune it to parallel resonance at twice the line frequency, giving better ripple rejection for a given size of choke. So you can make the choke smaller, saving money and improving the transient response.

I think a few of the big American-made linear amps, Henry and so on, used the choke-input version of it for the main plate supply.

I've got a 4-400C that I bought off a friend, who pulled it out of a RF induction heater. It seems to work, but I've no idea what to do with it!

[Bruce / Mission Amps]

Ha! You don't even have to dredge the ham radio mailing lists, just ask Bruce.

I think that's a resonant filter? The caps across the choke tune it to parallel resonance at twice the line frequency, giving better ripple rejection for a given size of choke. So you can make the choke smaller, saving money and improving the transient response.

I think a few of the big American-made linear amps, Henry and so on, used the choke-input version of it for the main plate supply.

I've got a 4-400C that I bought off a friend, who pulled it out of a RF induction heater. It seems to work, but I've no idea what to do with it!

Right, a 100Hz to 120Hz, resonant tank filter.... high voltage and you don't have to use big ass high capacitance, high voltage caps for good ripple rejection filtering.

Your old gronk tubes:
After finding my toddler son would sleep better with my old Hammerlund AM/HF recvr turned on and glowing in his room, I used worn out triode and pentode power tubes, mounted in an old block of wood and dimly light up with with a 10a 6VCT transformer... to use as a golden glow, night light in my my son's room when he was a baby.
What a waste of power but a very nice soft tubey glow and 24 years later, he still swears he can remember me putting all that old radio junk in his crib. Ha ha it never was in his crib of course.

[tubeswell]

tubeswell,
curious, how does that amp sound w/ the whopping 4k7 screen resistor there as far as sag is concerned?

It sounded pretty good

Also, what's the screen to plate voltage?

the screens are about 20V below the plates

Another choke related question:
An inductor opposes change in current. So when the power tubes in a push-pull circuit are greatly fluctuating in current, does the choke attempt to counteract this?

I don't think so - my understanding is that because the choke is filtered on either end, it is at a fairly stable voltage (and given the resistance in the choke is fairly constant, the current is otherwise fairly constant). In a normal 5E3 6V6, the screens sit at a fairly stable voltage and each screen hence pulls a constant amount of current at that point. The plates are the things that vary (and that is because they are connected to a reactive load). The screens can't vary unless you have unbypassed screen grid resistors. In my case the 4k7 allows for some screen current feedback, but as the phase on either side of the PP power amp is opposing, the changes in screen current tend to cancel out somewhat. Well that's what I reckon anyway. And I like the sonic effect. (I used a 5W resistor BTW)

[lowell]

Bruce, that's awesome... and hilarious (in a sick sort of way) to think of it in his crib. He may have Hot-tubephobia if that were the case.

[lowell]

Ok thanks... this cools the bias current somewhat too right? But maybe you counteracted it after adding the 4k7?

Another question: If you rebiased to 90% after putting the 4k7 in there does this affect the point of clipping on the plate? I'd think the grids would clip earlier as Vgk would be less. I'd also think your plate voltage may be a bit lower after reducing Vgk as there is more quiescent current, therefore earlier clipping at the plates.

[lowell]

Also had another thought that a power amp that's fixed-bias AND cathode bias could provide a good comination of squish AND sag, the best of both worlds. Perhaps in a 5E3 - 200ohm Rk and a bit of negative grid bias to keep things under 100% dissipation. Anyone try this? I'm sure it's been done... maybe in the Music Man amps?

[pdf64]

'a power amp that's fixed-bias AND cathode bias'
See the much reviled CBS TR
http://www.schematicheaven.com/fende...c568_schem.pdf
And the TQR approved Vox AC50
http://www.schematicheaven.com/voxamps/ac50_1.pdf

[Old Tele man]

..."swinging" choke.

[tubeswell]

..."swinging" choke.

Is this just another case of a bad choke , or some technical thing that bears closer scrutiny?

[imaradiostar]

My understanding, which could be totally flawed, is that "swinging choke" was a choke that is used to regulate voltage because it was close to saturation. The inductance decreases as the choke saturates and the DC resistance of the coil "takes over", essentially regulating the voltage.

Bruce's circuit is a tuned C-L-C network that cancels out the oscillations of the line frequency, allowing you to use smaller C values- which makes sense in Bruce's high voltage RF rigs.

A shorter path to the cathode/fixed bias option, and one that probably would be more musically useful, is a cathode biased amp with parallel zener diodes to "clamp" the cathode rise as current draw increases, essentially converting the amp to fixed bias as current draw increases.

You touched on another interesting idea- the difference in tone between a choke and a dropping resistor. I think it would be interesting to do something mentioned above- install a switch to select between a choke and a resistor between the plate and screen nodes. I'd probably want to try a center off SPST switch- one pole would have a choke, the off position would have a large (excessively large) dropping resistor, and the other pole would have a resistor that (in parallel with the "off" position resistor) would have the same DCR as the choke.

With two switches (cathode zeners and screen choke/resistors) you could get a feel for the overall tonal changes from these elements. In a simple amp like a 5E3 or an 18 watt Marshall I think it would be a good study- learning to listen for the sound of screen sag or the sound of cathode rise...and so on.

I was surprised how difficult it was to tell the difference with the parallel cathode zeners switched in or out on one of my builds. It kinda made me doubt my own ears! I'm sure it depends on the amp and how hot the tubes are biased, of course. If an amp is cathode biased hot it won't suddenly sound like a cold fixed bias marshall when you flip put some zeners in circuit- unless they're in series with the cathodes, and that's a whole other thread!

Now I feel like I should go search for threads that talk about that concept.

jamie