Why was grid-leak/contact bias used in guitar amps?

[Rhodesplyr]

I've read a number of online discussions of grid-leak/contact bias where the focus is on the problems it presents: too dependent on the characteristics of the particular tube, lack of a stable operating point, etc...

I'm not thinking of building one, but with regard to vintage amps in which they were used, did they have any particular advantages that made them attractive to circuit designers--or was it just a matter of cost and parts count?

I guess it would save a bit of time in terms of wiring up a cathode resistor and having a terminal for it, but that seems a relatively small gain. You also need an extra capacitor at the input.

I understand that "contact potential bias" is the more accurate term but that "grid-leak bias" has become the de facto terminology.

[R.G.]

Maybe it was for a lot of voltage gain in one stage. Old pickups were not nearly as hot as modern ones.

[Gingertube]

Yes,
R.G. is correct. It was for maximum gain without the need for a cathode bias resitor and hence an electrolytic bypass capacitor. The need for a series capacitor on the input (NOT an electrlytic) was a small price to pay. Trade off is that this arrangement is susceptible to hum pickup from the heater.
It does not work very well with old tubes as the voltage developed across Rg1 from grid current from gas ionization, opposes the bias established by grid current due to the "grid leak", that is, the POSITIVE and NEGATIVE grid currents oppose each other and for the same reason it does not work at all with power tubes.

Cheers,
Ian

[Leo_Gnardo]

Whatever the reasons, and the drawbacks, there's a very nice squashy compressed response in grid leak inputs - I like it! (Of course you have to have your guitar dialed all the way up.)

[pdf64]

My guess is that in the early days, every component was expensive (probably even more so with electrolytics) and unreliable, so introduced more failure modes; best to reduce the count wherever possible.
The 5E3 volume arrangement may be seen as another example - no mixer resistors needed!
Pete

[Rhodesplyr]

What's interesting to me is that a few of the guitar amps I've worked on with grid-leak/contact bias of the input stage also have lower input impedance (in guitar terms). For example, the Supro Super that inspired this question turns out to have a ~200k impedance, set by resistors, when only one input is used. A Premier 120 has ~300-400k on the Instrument input. So, were these inputs designed to load down the input voltage swing to a level that the input stage can handle? Just a thought...

[Leo_Gnardo]

What's interesting to me is that a few of the guitar amps I've worked on with grid-leak/contact bias of the input stage also have lower input impedance (in guitar terms). For example, the Supro Super that inspired this question turns out to have a ~200k impedance, set by resistors, when only one input is used. A Premier 120 has ~300-400k on the Instrument input. So, were these inputs designed to load down the input voltage swing to a level that the input stage can handle? Just a thought...

I'm thinking more likely to "round off" hi frequencies for a mellower tone. This is an area open to experimentation so slap in a larger resistor for a brighter tone if you please. I've seen up to 2M2 (2.2M for those who write that way.)

IIRC the old Crown mic preamps are grid leak too, and have a plush tone & just a bit of compression built-in because of that.

[uneumann]

Just a comment - an example of new'ish amps that still use grid-leak input stages is the THD Flexi-50. It has 1M input resistor, 68K grid stopper, and 6M grid leak (behind a 6.8nf cap).

I have one and it works great with a number of different 12ax7s I've tried. Even the ECC803 works well and I currently use it in that amp. It seems robust to newer tubes - I don't use any NOS. The sound impact is hard to pin down. I can't say I hear anything that I can specifically tie to the grid-leak bias configuration... but I also can't say there is no effect.