Grid leak/contact bias preamp

[EFK]

In a grid leak biased preamp (this is 12ax7), what is the purpose of a small bypass cap across the grid leak resistor? This amp uses a 6.8M grid leak resistor but is bypassed with a 500 pf cap. The other channel is a more typical cathode bias setup, although it uses a 470K resistor to ground rather than 1M Fender/MArshall value on the input, but interestingly enough, that 470K resistor is also bypassed w/ a 500 pf cap. Is this some way to cut off high frequency oscillation potential, or some kind of RF blocking? I had always read that you "could not" bypass a grid leak resistor. ????

[MWJB]

This 500pf cap is in addition to the gid leak cap between tip & grid leak resistor?

Out of interest, why wouldn't you bypass the grid resistor?

It would shunt a lot of signal & highs to ground, maybe a mechanism to stop the grid leak input from choking under big signals (one of their big failings)?

[EFK]

No, not between the big .05 input cap and grid leak (6.8M), it's *across* (or parallel with, in other words) the 6.8M grid leak resistor, just like a cathode bypass cap - it goes to ground. 500 pf must be shunting an awful lot of highs!

[MWJB]

I get you. Indeed, some harp players have used a shunt resistor accross crystal mic elements to brown out bright mics, values could be anywhere <1000pf, similarly a cap this size might bypass a plate resistor at the input stage for the same reason. How does it sound? What's the amp?

[Enzo]

Well, why not take 500pf and 6.8M and calculate what freq that represents. I'd be thinking that super high impedance is just begging to be a radio receiver, and the cap is there to bleed off any RF that comes knocking.

[Tubebass]

46.83 hz

[Chuck H]

yeah... That's a little below RF, if that's the actual frequency. I don't know the formula but isn't it relevant to that caps pass frequency that the load resistor is in parallel to the plate impedance???

[pdf64]

I think that the a low pass filter turnover frequency would depend on the source impedance, which would effectively be in parallel with the 6M8 (assuming I've understood the layout correctly). And the 500pF would also be in parallel with cable capacitence from the lead to the instrument, which would likely be another 500pF to 1nF.
So at a rough guess, the roll off would kick in around 2kHz.

[EFK]

LAyout is (normal input) guitar cable--->unswitched jack--->47K--->.05 cap--->6.8M/500pf to ground--->12ax7 grid

Treble input is the same, except switched jack and instead of the 47K, there is a .005 cap prior to the big .05 cap.

I think the easiest thing to do is just lift one side of it and see how it sounds!

[martin manning]

What make and model amp is this? Can you post a schematic or a link to one?

[EFK]

It's an old Supro, 1957 serial but no model number. I have a few from the 57/58 period, these have the grid leak front end on channel one but I've nto seen two the same. Have never been able to find schematics - I tried drawing one out (a 1690t) a while back but it's such a rat's nest in there it's easy to miss things.

[Amp Kat]

Think these were some of the first Fenders that used actual grid leak bias and the Cathode was grounded.

[redelephant]

I think that the a low pass filter turnover frequency would depend on the source impedance, which would effectively be in parallel with the 6M8 (assuming I've understood the layout correctly). And the 500pF would also be in parallel with cable capacitence from the lead to the instrument, which would likely be another 500pF to 1nF.
So at a rough guess, the roll off would kick in around 2kHz.

Correct. The cap presents an impedence of 1/(2*pi*f*C), which at 100Hz is a whopping 3,2megohms, and at 1kHz is still a whole 318kohms. Even at 10kHz it is 31.8kohms. Most sources driving this, even relatively highish 10kohms and more won't seem to be much attenuated from this capacitance.

Remember a guitar cable has capacitance due to the shield, and some shielded cable has capacitance of more than 100pF per meter, or every 3feet. So a typical guitar cable has more than the mentioned 500pF, and we don't seem to care about that.

Perhaps such a cap at the input is what gives that vintage rounded jazz tone?

[Chuck H]

Remember a guitar cable has capacitance due to the shield, and some shielded cable has capacitance of more than 100pF per meter, or every 3feet. So a typical guitar cable has more than the mentioned 500pF, and we don't seem to care about that.

Oh contrare. I use to really piss off the counter guys at Gutar Showcase in San Jose because whenever I needed a cable I would take my DMM and test every one before buying. You might be surprised (or not) at how much variance there is even within the same model/brand.

Perhaps such a cap at the input is what gives that vintage rounded jazz tone?

My most recent design uses a 4.7pf cap on V1 from plate to grid to roll off some extreme top end. in the end it's about the same as running fourty feet of cable instead of the typical twenty feet. Not rolled down into "jazz" but it seems to shave the uglies without loss of definition or useful harmonics. The whole point of this was my observation that musicians on the big stage used longer cables... And their amps seem to sound better?!? My amps are built for small clubs, so...