Bias Splitter Theory

[alerich]

I am trying to wrap my brain around the function of the bias splitter resistors in the Marshall JCM 800, although this would apply to many different amp circuits. In particular, how the bias splitters interact with the PI coupling caps and the PI itself. I am referencing the Mark Huss schematic: http://mhuss.com/MyJCM/JCM800_2204.gif. In the stock format there are two series 150K resistors that are strapped in parallel after the coupling caps leaving the PI. Negative bias voltage is applied to the junction of the two 150K resistors. My Sovtek Mig60 has a similar arrangement. In an effort to deal with mismatched tubes and just because I like the concept of greater adjustability I opened the junction of those two resistors on my Mig60, constructed a second bias adjustment circuit in parallel with the original and connected it to the other now disconnected bias splitter resistor. I'd like to incorporate the same design in my current JCM 800 build on my bench. Ok - it's really on my kitchen table.

It seems to work well in the Mig60 but I am wondering what the circuit purpose of those bias splitters as designed is. The signal at that point still has a path to ground through bias circuit just as it had before. The power tube grids still also have that same path to ground. However, there is no longer this 300K resistance across the PI outputs or the power tube grids since it has now been split. In other words, how does it impact the output of the PI and the signal in general by making this change?  Not only that but what about its impact (if any) on the action of the grid of the power tube? I have found lots of info about mods to the bias splitters and bias circuit but not much about the theory behind it.

[sluckey]

the bias splitter resistors

That's not a term I would use. Those resistors are the output tube grid return resistors. Those resistors provide a path to the output tube cathode thru the bias supply, ground, and finally to the cathodes. There must be a dc path between the grid and cathode for biasing to be valid.

The grid return resistors do impact the PI tube. Each resistor is in parallel with the plate load resistor and are part of the total load on the PI tube. Smaller resistors will load the PI more. Larger resistors will put a lighter load on the PI.

And those resistors also interact with the coupling caps just like the grid resistor interacts with the coupling cap in any RC coupled amp. There is an RC time constant that is determined by multiplying the resistance times the capacitance. Larger time constants can pass lower frequencies. IE, you can make the bass response greater by increasing EITHER the coupling cap value OR the resistor value. The opposite is also true.

However, there is no longer this 300K resistance across the PI outputs or the power tube grids since it has now been split.

Those resistors are not in series as far as the PI or output tubes is concerned. The junction of those resistors is at AC ground via the bias cap.

Splitting those resistors will have verrry little effect on either the PI or output tubes. This assumes the bias circuit below those resistors does not change drastically and remains the same for each separate bias adjust circuit. A drastic change would be something like changing the 22K bias pot to a 100K or 250K pot.

[alerich]

Thanks for the explanation, Steve. I didn't think it had a significant impact and I had not noticed a marked change in the performance of my Mig60 since the mod. Off to warm up the soldering iron.