Paraphase Phase Inverter...and other things 5C3

[jmaf]

Jmaf, the cathode resistor predominantly sets the bias (hence the term "cathode biased"), yes the plate resistor has an effect...

It depends.... ThermionicScott told the fella one set the bias, the other gain. That is not always correct. You may have equal plate and cathode resistors, in which case both play the same role(in setting the bias). You may have no cathode resistor at all like in some Gibsons(let's not get into whether it's a good design choice).

A cathode follower for a 5F6-A would be a great example to show how confusing it must be for a newbie to get this sort of advise from a senior member, and then go and see *holy ss@#$ this thing has no plate resistor*....

EETStudent said:

I assume the effect here is to set the bias point, in conjunction with the cathode resistor.

That is a correct description of what that gain stage is doing(in a 1/60 proportion 8-) ) and he got corrected by ThermionicScott, so I just chipped in to say EETStudent is not completely wrong. I may have been too picky?

[ThermionicScott]

LOL! Good question. I don't know - I've always typed BIAS in all caps, I think due to those old cassettes which had "High BIAS" printed on them....

Ha, I remember those!

The cathode resistor has 60 to 100 times more effect in setting the bias for the stage, that is true. Though you can't simply separate things like that, tube transconductance and those 2 resistors are all intimately tied. That is, if you're doing nitty picky preamp design. And I just nittypicked on you for saying one resistor sets bias and the other gain, it's not as clear cut as that.

Yeah, it's interactive and interdependent. What I should have said at the outset was that I would bet LOTS of money that Fender didn't change the plate resistor values simply to alter the bias point, which was EET's guess.

Bringing in cathodyne PI's and cathode followers kind of muddies the discussion that was centered on Fender's preamp stages, but I think we're all in some sort of agreement now.

- Scott

[ThermionicScott]

BTW, here's a decent description of what's happening in those grid-leak biased input stages (according to them, we're using the wrong term ):

"Note: Grid leak bias is often confused with contact potential bias, but the operation and applications behind them are totally different.

Contact potential bias is formed by space charge near the hot cathode of a tube. When electron emitters are heated electrons boil off. They collect in a cloud around the cathode. This cloud supplies all the charges available for normal operation of the tube.

Because the control grid is in the cathode cloud, a few electrons will land on the grid. If the grid has high enough resistance to ground it can develop a small negative potential (excess electron charge) of a few volts. Suitable resistances are in the order of several megohms. This is a very unreliable source of bias. Contact potential bias is useful only in very small low voltage tubes, and it is even unreliable in them if the tube is less than perfect."

Besides saving Fender the use of an expensive (in those days) electrolytic cathode bypass cap, I've heard that this method was less susceptible to hum. Over time, the stability of cathode bias won out.

- Scott

[EETStudent]

MWJB- interesting suggestion about one cathode biased input and one grid-leak( or contact potential...).

I'm running 3 inputs, like the 5C3. However, they all share the same triode input like a 5F11. Would the two input types still work in the situation?

I'm interested in more thoughts on conta ct potential bias at the input. You say it can't handle big signals, but how big? Say I played the amp(stock 5C3 circuit) dimed-how would the different input types affect performance?

[ThermionicScott]

I'm interested in more thoughts on conta ct potential bias at the input. You say it can't handle big signals, but how big? Say I played the amp(stock 5C3 circuit) dimed-how would the different input types affect performance?

The 5B3 layout (no voltages on the 5C3) shows a grid voltage of -0.9V on the input stages. So, a 1.8V peak-to-peak signal will clip on one side, giving nice even-harmonic distortion unless you really wallop the input. I think traditional Fender guitars can only do about 1V peak-to-peak, but it's easy to do more with a boost pedal.

- Scott

[EETStudent]

So the result is no more or less pleasant than a cathode biased input? I play a 50s spec tele, and no more than a mild 808 style OD occasionally. I'd be interested in experimenting if I had an idea of the effect

[EFK]

If you're asking for a comparison b/t a grid leak input vs. a cathode bias input in a like/like comparison (same circuit/tubes), I can relay my experience with a good number of Supro 6V6 and 6L6 paraphrase PI amps. It's not a night and day difference. The only impression I can really offer is that the grid leak inputs sound very slightly more 'open' if that makes any sense, and a bit looser. When you really wind the amp up, and depending on what you are hitting the input with, they seem to take on a more compressed character than a typical cathode bias input. Consider though that it is not just the bias method of the tube itself which is being changed, but now you are running your signal into the amp through a capacitor of a specific value, which certainly must be having an effect on the sound. In the amps I have which are leak biased inputs, I frankly feel that 12ay7 sounds much better than a 12ax7 despite the amp calling for a 12ax7.

[MWJB]

For a given B+ & plate resistor value the grid leak input draws more current & typically has a little less headroom than a typical cathode biased input. No, you can't easily run the 2 methods into the same triode, well, not with out some switching...in cathode bias the cap at the grid would need to be shorted and a cathode resistor between cathode & ground. In grid leak, short the cathode to ground & open the switch accross the input cap. A DPDT should do it.

You could also switch the grid load resistor value (using a parallel input jack?), or pick a compromise? Grid leak designs tend to have larger plate resistor values, but this is just a historical trend (so do many cathode biased circuits of the same era), you can run either circuit with 100K.