dean markley cd-120 screen grids

[Andy Le Blanc]

Just had as a repair over the bench a dean markley cd-120

The power side is a fender clone 4 6l6 thing.

And suffers the same design as nearly every other fender mimic..
the screen grids are run slightly positive relative to plate

I modded the screens, a 100 ohm then 220k bleeder before the regular
screen grid resistors which placed the G2 V slightly negative with respect to
the plate with the amp biased to 19 w plate dissipation.

I never liked these rigs, nobody I know really cared for them..
the mod was trans-formative, nice amp...

I'm surprised, two extra resistors to resolve historic G2 issues with fenders
fender clones and anything that makes fender based assumptions.

[pdf64]

How much did the screen grid voltage change? Are you sure that you didn't fix a bad screen resistor in the process of doing the mods?
It's just that when I've put a 1k resistor in series with the choke, with a switch to short it out on the back panel, I've been hard pressed to tell which way I had the switch (ie was it 'up' to short it out or put it in circuit?). There was no obvious difference in tone or increase in sound pressure level whichever way the switch was.
I put the resistor in put add some sag to the screen node, like a 5E3, and drop a bit of power output. It did drop the amp's power output on the bench, about 5 watts, but I couldn't hear the difference with the AB switch on a cranked amp.
I think that a 1k resistor in series with the choke will have a much greater effect that the 100/220k potential divider described, so I remain to be convinced that the 100/220k addition to the screen supply would make much difference on the bench or have perceivable benefit to the amp's tone. Pete.

[Andy Le Blanc]

Appreciate the response

this has nothing to do with sag

The power side design adopted by fender and so most other mfg, since
places the screen potential slightly positive relative to the plate.
The voltage difference measured between G2 and P on the tube socket
in the power side of a completed circuit, measured under static conditions at set up.

This voltage difference sets the position of the virtual cathode in beam pentodes.

This also pertains to tetrodes from which beam tubes and pentodes,
pentode meaning a tube with a suppression grid, evolved from.

In the case of a tetrode, as long as the plate voltage is higher than the
screen grid voltage, plate current depends to great degree on the screen
grid voltage and very little on the plate voltage. The fact that plate current
in a screen grid tube is largely independent of plate voltage makes it
possible to obtain much higher amplification.... this is right out of RCA

A beam tube is a dressed up tetrode that relies on a virtual space charge
instead of a suppressor grid, you still have to mind the screen grids
the traditional fender circuit put the screens 250 to 500 mv. positive
a the published bias voltage, to maximize the clean power output.

The design was never meant to be pushed into distortion, fenders were
sold for the power and clean fidelity, distortion pushes the operating class
past the design intentions, with a slightly positive G2 the voltage relationship
between the plate and screen swings, and makes nasty little modulation
the fender fizz, the swing also gets the screen to draw current.

And with a series screen grid resistor the change in current means a change
in voltage which places the G2 and resistor in dire straits, BBQ

which blows first the tube or resistor,
it sounds bad and exaggerates the differences in matching tube sets.

Using a voltage divider instead of a series resistor allows you to set the
G2 voltage relative to the plate and provides a degree of regulation to
the screen grids, it avoids not only G2 issues but also allows you to set
up the power side with out making design assumptions.

The mod allows me to set the screen voltage relative the plate with a
large of certainty and control, the G2 were .250 and .850 positive
before the mod and ended up around -1v the other way or -.150 v
as the nearest potential, without changing the screen grid resistor values
and without the tone issues usually associated with the selection of
a screen grid resistor value.

It takes away the assumptions out of dealing with your screen grids.
it allows you to know exactly what the G2 relationship is with the plate.
it allows you at set up to exactly place the screen grid voltage,
without having to guess about screen grid resistor values

Measure it, set it up, repeat the set up with accuracy

[Structo]

So to be clear, you have a 100 ohm screen resistor on each tube then a 220K resistor from the screen node in series to those?

[pdf64]

Thanks for the further rationale and detail Andy, I'll look into it. However, I need to brush on the theory to do it justice, time for an RDH4 session. Pete.

[Andy Le Blanc]

nope, just the regular 470r screen grid resistors...

but they're sourced from a voltage divider from the usually PS node.

a 100r and bleeder from where you typically connect the screens
then the usual screen grids, it moves the screens negative without
excessive screen grid resistors. By using a voltage divider it gives you
the chance to reduce the screen grid resistor also, but I'd keep a value there.

Ive not wanted to greatly effect tone, but change the G2 plate relationship.

I assume that divider values can be chosen to address design goals instead
of relying on series dropping resistors and the issues that come with them.

I tried a different design approach, instead of basing power side and bias
assumptions upon the tube bias, and all the qualitative (BS) arguments
about tube bias a screen grid resistor values, I turned my assumptions upside-down.

I decided to make an arbitrary choice about utilizing the PT current rating.
I chose to use the bias to meet PS goals instead of basing it on tube type.

with the voltages stabilized with bias ignoring tube type, what changes
with tube type is the G2 relationship in the push-pull circuit.
A voltage divider (screen grid bias supply) makes controlling this possible

And its a different world, tube type is irrelevant with in the practical
limitations of maximum plate dissipation, and its possible to shed all the
BS about tube bias points and screen grid resistor values.

A tube amp is a POWER SUPPLY connected to a speaker.

when you set up an amp the bias controls the PS voltage, unless you
have separate regulated supplies, your noodling with the bias affects
the entire rig, all those carefully measured voltages in your wreck clone
don't mean a thing, its why fenders had balance and not bias control, if you know
that a production run will fall within 5% either side of a projected curve
why spend for the extra parts.

how the heck can you make design choices when your voltages are all
over the place with tube bias, no wonder that marshal eats tube
and the old silver face is all fizzy