Help to get rid of hum please.

[catalin gramada]

Hey. Did not found any comercial tube shields for octals so I had to improvise some. I used in two instances: one ordinary steel sheet 0.025 in one layer and 0.01 mu-metal permalloy sheet in 7 layers. Both are absolutely effective and cut any noise in my preamp to none. I did not found any difference between both shields in my quiet environment.

[Helmholtz]

Often even a non-magnetic shield (aluminum, brass, copper, nickel-silver) works, because it's not magnetic interference.

[catalin gramada]

Often even a non-magnetic shield (aluminum, copper, nickel-silver) works, because it's not magnetic interference.

Hey. You're damn' right ! One 100u out of order cap makes the best choice for my octal. Thanks man !

[The Dude]

That's awesome, and a great use for an old cap.

[catalin gramada]

That's awesome, and a great use for an old cap.

Thanks Dude. Both input sockets are wired in parallel to may use 6n2p (12ax7) or its octal equivalent 6n9s (6sl7). Unfortunately the big one is noisy as hell and cannot be used without a shield...at least in this position. I much preferred this way rather to use an noval-octal adapter

[catalin gramada]

BTW, I don't see a grid leak resistor.

Hey. Does a 1M input grid leak resistor is a really grid leak resistor ? How about a 5.6M input like in Ampegs circuit ? Sorry is very hard for me to figure how a 1M strapped with say 15k dc path by pickup resistance it is still a grid leak in input circuit. I.m talking about a clear situation where dc path from the first grid is switched only from strapped to ground and strapped by dc resistance of whatever pickup it used and no isolated by whatever input coupling caps.
May you enlightening me please ? Thanks.

[Helmholtz]

A cathode biased tube needs to have a DC grid leak path to ground.
If the grid is floating the essential grid-to-cathode bias isn't well defined.

Actually grid leak means a negative grid current out of the grid caused by electrons trapped by the grid.
This grid leak current causes a negative voltage drop across the grid leak resistance.
As long as the grid leak resistance is 1M or lower (with typical triodes), the negative grid voltage is insignificant (say below 0.1V).
A 5.6M grid leak resistance with some tubes causes sufficient negative grid bias even without cathode bias.
Those circuits need an input coupling cap.

And yes, if the source resistance is low, a grid leak resistor isn't necessary. But not all capacitor coupled active electronics (e.g. pedals) might provide the necessary resistance to ground.

Resumee: Add a grid leak resistor. It's cheap, has no drawbacks and may avoid bypass switching pop with pedals.

[Mike K]

Right, you're probably thinking about grid leak bias vs cathode bias.

In terms of input grid leak, it can be as low as 68K as it in on many low inputs (and will probably work even lower). This 1) lowers gain and 2) lower the input impedance, possibly loading up anything out front. I don't know how 1M came to be, but it seems to be a popular choice and provides a high impedance which won't load down whatever it's hooked to, provides sufficient gain in conjunction with the tube to raise the signal over the noise floor, and doesn't (usually) incur too much noise penalty. Grid leaks always contribute to tube noise - higher value = higher noise.

[Helmholtz]

. Grid leaks always contribute to tube noise - higher value = higher noise.

Not really.
Input stage noise is mainly due to total resistance between grid and ground.
The grid leak resistor is shunted by pickup resistance (maybe 10k) + grid stopper resistance.
This makes the grid leak value rather insignificant.
The grid stopper value has a noticeable influence, though.

[catalin gramada]

I thinking for a certain situation. Guitar plug in and out. First function it should provide a DC reference to ground but don't think is matter if this reference is provided by a pickup or a dedicated grid leak resistor. Second AC conditions which due to inductive specs of any coil may provide some resonances in freq response. But also have my own doubts as this resonance should be by default dampened by its own resistance and all guitar potentiometers in circuit. So from this points the input grid leaks worth nothing in my opinion. It becomes compulsory to use just in isolation conditions mentioned by Helmholtz above. This is what I see and the reason I asked for to be sure. Any article related by grid leaks resistors treat the subject like it works only in isolated dc conditions (and the input jack tips goes to nowhere and no switched to the ground when off... ) where the grid reference to the ground should be provided somehow and of course all discussions regard proper chosen values...etc makes sense. My question regard a passive guitar pickups into a switch jack input.

[Mike K]

It's easy to try, but my thought is it is fully functional. But I've always taken it for granted. Someone once gave me this analogy:

A child asks his mother why she cuts the end off the roast before she cooks it. The mother replies, "Grandma always did it this way. It tastes the best." The child goes and asks grandma. "Why do you always cut ends off your roast?" Grandma replies, "I don't, I only did that because my pan was too small for the roast."

[catalin gramada]

I tried and have no problems plugin the guitar on and off into a switchcraft input without any grid leaks in front. Still to be honest feel a difference between a 1M and 2.2M grid leak when use it. The bass seems more dampened,more tight with 1M , but is all subjective , have no ideea how to measure to prove...
A simulation model should be usefully but have no abilities to do it.

[Mike K]

Output taken at junction of C5 and R8.

Values of grid leak, R4: 100 500 1000 5000 10k 50k 100k 500k 1000k 2000k 5000k


As you can see if affects the resonance peak of the pickup as well as the overall output. The output is changed at the input grid as well, but I believe it is a compounding affect in that affects the output of the pickup as well as the gain of the tube. The gain between each is not a constant value.

[Mike K]

Here is with no grid leak - surprising similar to a high value.


I am actually a bit surprised by this.

[catalin gramada]

Thank for that ! Yes, think infinite input impedance as indefinite variable input impedance. No good. We have to declare a impedance-low or high - in respect with pickup characteristics