Split Cathodes

[shaun]

Hi All.
I have come across some older posts dealing with separate cathodes on the power tubes, with the discussion revolving around power attenuation. The Carr Vincent is mentioned. They're very interesting, and I'd like to give it a try in one of my builds, however, the graphics are too grainy to make out, and I cannot find them in the archive or elsewhere. Seems the subject went quiet for some reason.

I'd be one happy camper if anyone could post jpegs or gifs of those diagrams. Geezer seems to have had considerable success, as did others.

Many thanks, and I hope you're having a good building day.

[CascoSieg]

I've been thinking about this for my next build as well. FWIW, here's a draft of how I understand the basic idea, taken from what I've read here: split cathodes, with a ground lift on one of the pairs AS the output secondary is also switched to present the proper load to the remaining active tubes/plates.   This is just draft - ignore component values.

[shaun]

Thanks Cascosieg.  Very interesting.

[passaloutre]

I actually did this today on my Cathode biased 6L6 plexi. Pardon the hand-drawing, but hopefull it makes sense. The switch parallels the two cathode resistors and adds the 25 uf cap in parallel. The 100k drains the top of the cap to prevent popping (it should probably go across the switch but it works fine this way).

Oh, it sounds great BTW!

[PRR]

> it should probably go across the switch

Yeah, or even to the resistors (same thing). So the cap holds the average DC even when idle.

But 25uFd is very small for a cathode cap; maybe the tube hardly notices. Especially in a push-pull amp.

And of course if it works for you, that's all that matters.

[passaloutre]

It’s 25u because I originally had 6v6s in this amp and largely copied the Tweed Deluxe power section. When I moved to 6L6 I changed the resistors, but not the cap (it’s rated for 50v). would I get more output power with 47u, or just more bass?

I swear I had a reason for wiring the anti pop the way I did but yours makes a lot more sense.

While I have your attention, I’ve been thinking about elevating the heaters in this amp, and the cathode resistor is a natural choice usually. Does the switchable split-cathode bypass make this a bad choice? I’m guessing that DC voltage is a lot less stable than it would be in a traditional shared-bypassed configuration…. I can just elevate the heater center tap elsewhere.

[passaloutre]

A day after installing this switch I was playing the amp pretty extensively, and I get kind of a burny smell in the low power position. Is there anything about this mod that would generate more heat? Do the big wire wound resistors need to “burn in”? It seems to be coming from the power end of the amp, and I’m not sure if it’s the new resistors or possibly a transformer or something?

I’ve got two 6L6s with 405 vdc on the plates and 27 vdc on the cathodes. Nothing *looks* burnt or overheated inside.

One weird thing I noticed is I get 283 vdc on the output of the rectifier (diodes) with the standby switch *off* (i.e. open), and 410 vdc with the standby *on* (closed).

The cathode resistors are 10w rated

[shaun]

Burney smell isn't good. Smoke is the obvious tell-tale to look for. Hot smell is okay to a point. I find that 5watt resistors get hotter than I'd like them to, even when they're only dissipating 2watts of heat according to Ohms Law, so I tend to go for 10watt resistors whenever there is real estate, for things like cathode resistors and/or power dropping sometimes.

Anyway, I'm attaching a schem of a split cathode system I installed this week, and it works very very well. In addition, I found that with split cathodes, when the bypass caps are lifted from ground (by insertion of a 47k res on each cap's negative side, using a DPDT switch), it reduced gain by a little more than half. Which is good because the amp must be somewhere around 25watts, which is kinda loud at home. And the tone doesn't shift too badly by removing the caps either, which surprised me. A handy power dropping circuit. And it doesn't seem to pop in any noticeable way. Lots of surprises with this relatively easy circuit.

Hope that helps.

[sluckey]

I bet those 470K cathode resistors are really 470Ω.   

Da Geezer was a moderator and active amp builder quite a few years ago. He came up with this simple circuit that works well so I put it in my amp scrapbook. I've attached it here for your convenience. Take a look. Y'all may find it interesting or even useful...

[passaloutre]

Definitely no smoke (I left it running for 15 minutes or so on the bench after checking things out). I kinda feel like it might just be a hot smell (cathode bias runs hot obviously) from new parts or something (maybe cooking off residual flux from my rosin core solder) but just asking. No evident effects to the sound, and the low power option works very nicely (I’d agree with the half power estimation).

Quick estimate from the oscilloscope: I get around 17v rms across the 8 ohm speaker at the full power setting (36 watts), 11v rms from the low power setting (15 watts). This is with a 1kHz 500mv signal on the amp input and max clean output (visually on the scope).

[PRR]

...One weird thing I noticed is I get 283 vdc on the output of the rectifier (diodes) with the standby switch *off* (i.e. open), and 410 vdc with the standby *on* (closed).

410/283 is 1.448.....

We might expect 1.414 (why?) but rounding.   Axtually, meter rectifier.

[passaloutre]

Not sure I understand, this is the rectified dc voltage (without the reservoir cap). Should the reservoir go before the standby switch?

[sluckey]

Not sure I understand, this is the rectified dc voltage (without the reservoir cap). Should the reservoir go before the standby switch?

Since there is no filter cap connected directly to the rectifier, all you have is 120Hz DC pulses. The peak voltage of these pulses is about 410V, but your meter is averaging these pulses to display a lower average voltage of 283. Once the filter caps are connected via the STBY switch, the caps can charge to and hold the peak voltage of those pulses. Now your meter can more accurately measure the smooth filtered DC.

It would be much better to move those two parallel 20µF caps to the other side of the STBY switch, ie, connect directly to the rectifier.

[passaloutre]

Roger that. This is what I get for copying old schematics

[shaun]

I bet those 470K cathode resistors are really 470Ω.   

Yes! Absolutely. Thank you! My mistake. 470R cathode resistors. They're giving me around 65mA zero-signal per 6L6GC, which peaks (according to my mA reader) around 72 per side with signal running through.