Homicidal Maniac - Submini Amplifier

[panterafanatic]

Hey guys, stomped out a basic idea of something to do with 2 6111s and 2 5672s. Basically a pushpull version of the murder one.

I had ordered 3 of each when I got parts for the murder one, which is this amps predecessor, so I had to come up with something based on it right? Same basic opening triode idea, I might play with it once I get my breadboard in (futurlec takes FOREVER). Input transformer for simple phase splitting, small amp = small OT, so I'll hopefully wind both transformers, and just buy PTs.

Hopefully I'll get some cheap transformer stuff and start this. I reckon anyone that has a few left over subminis may wanna try this.

I didn't include powering and heater value stuff since I haven't started and it'll vary per set up. I plan on running the tubes near saturation so it has some power amp distortion, but not square.

Photobucket shrinks it a bit but here it is so far



Any thoughts on it?

[kurtlives]

You need a input grid stopper and grid leak resistor. Otherwise your grid will go positive and you'll get nasty gating and the tube will be mis-biased.

Why mu-followers for the first two stages? The signal will clip very fast, before it even gets to the output tubes.

EDIT: Your output section is messed to. Lots of issues there.

[panterafanatic]

Removed output caps, made it cathode biased, added the grid biasing resistors and well as bypass caps on the cathodes. Kinda messy looking in the biasing for the pushpull stage, wanted the component names to be visible.

[frequencycentral]

The 5672's don't have an external cathode connection, being directly heated tubes, so you won't be able to used cathode resistors or bypass caps on the pentodes. 

[panterafanatic]

Ah, I tried asking that at the other place, but I guess it wasn't phrased quite right. So does that mean I need higher plate resistors?

Or should I just lower B+?

[frequencycentral]

The pentodes cathode bias is set at 1.25 volts by the heater voltage - the cathode is internally connected to the heater +ve (pin 3). No plate resistors are needed for output pentodes, just connect the plates directly to the OT, as you have in the second schematic. You can use negative grid 1 biasing (-6.5 volts) with 5672, see the Murder One thread for that.

[panterafanatic]

I didn't realize that's why you biased it like that, interesting. I've taken off the cathode stuff from the pentodes, cleaned up the other biasing slightly, and added that -6v to -7v option. (I'll be using PTs rather than a charge pump)

[kurtlives]

Your OT CT is wrong. It is not getting power.

The CT should go to the B+. The screens should go to a node with less potenial than B+. B+1.

[panterafanatic]

But the screens have a resistor between them and the B+?  The CT of the output transformer is directly connected to the B+, that's how I read it though.

[frequencycentral]

But the screens have a resistor between them and the B+?  The CT of the output transformer is directly connected to the B+, that's how I read it though.

+1, looks fine to me.

[kurtlives]

No

The screens draw almost no current, the resistor wont drop hardly any voltage. Then your screens will be at the same potential or higher than the plates.

Look at a non submini amp schem You'll see the screen supply has a separate node, that is not the B+.

[panterafanatic]

So I should separate the B+ from the screen with an additional drop down voltage from the PT, yes? As in a resistor drop from B+ to the line going to the resistors to the screens.

[kurtlives]

Yes. Some decoupling with series reistance would be good. Then filter it.

[panterafanatic]

Thanks Chris, I do know to filter each sections of a power supply though. 

Updates tonight with a slightly more sophisticated scheme.

[kurtlives]

Look at any Fender schem for examples.


                                     Voltage 1                           Voltage 2
voltage in-----resistor---cap to ground-----resistor---cap to ground

Often called a "dropping string"

[panterafanatic]

Updated the scheme to support the B+1



Here's a rough power supply info, Would rectifying and splitting the power work with the voltage being brought back up to 120v, Not sure which diodes I'll use for the HT rectification, any suggestions?

My last question on the power supply is will the 1N4148s back to back keep the voltage around 100v or should I find something that can dissipate a little more wattage?

[earthtonesaudio]

Look at any Fender schem for examples.


                                     Voltage 1                           Voltage 2
voltage in-----resistor---cap to ground-----resistor---cap to ground

Often called a "dropping string"

The other thing attached to those cap/resistor junctions is usually a plate load resistor, which does the actual dropping.  If you rely on capacitors alone to drop voltage, you better have some insanely leaky caps.

[kurtlives]

Look at any Fender schem for examples.


                                     Voltage 1                           Voltage 2
voltage in-----resistor---cap to ground-----resistor---cap to ground

Often called a "dropping string"

The other thing attached to those cap/resistor junctions is usually a plate load resistor, which does the actual dropping.  If you rely on capacitors alone to drop voltage, you better have some insanely leaky caps.

There are resistor in line with the B+. The caps are just there to filter.

Of course the voltage then goes to the plate load...

[earthtonesaudio]

I agree with you, but as it is drawn in this schematic, the B+1 voltage does NOT go to a plate load, only to the screen grids.  Like you said, they don't draw enough current to make that voltage drop appreciably, so the B+1 voltage should also be applied to a load that is going to draw some current (such as a set of preamp tubes, like you said).

[panterafanatic]

Relatively final scheme.



Decided to just use two 12v PTs