Ground referenced bias connected to floating circuit

Thread Starter

coinmaster

Joined Dec 24, 2015
502
Here goes me asking about floating circuits again.
I've been trying to figure out a safe way to direct couple stages of a tube amplifier I'm building while still playing nice with external dacs. My specific dac has a direct coupled output with no capacitor buffer so I need to be careful what I connect it to.
I can either keep the input stage ground referenced and have the output of the amplifier be at a high positive voltage
Screenshot_93.png

or have the output be ground referenced and have the input stage be referenced to a high negative voltage but that would require me to float the entire dac which I'm still not sure I want to do.

Screenshot_94.png
The other option I'm thinking of is having the input stage at ground reference and having the amplifier be floating so the high voltage output is not a danger.

The problem I have here is that I do not understand what would happen if I connect a ground referenced dac output to the grid of an input tube in a floating amplifier.
Does this all of the sudden reference the entire amplifier to ground or will the bias just not work? I'm struggling to understand how the electrons will behave in this situation.
 

MrAl

Joined Jun 17, 2014
11,395
Hi,

Not entirely sure what you are doing, but it sounds like you are trying to drive the input to a tube amp directly with a DAC.

Probably the best bet would be to use a buffer amplifier. That would change the output of the DAC to a level required by the tube input. The input and output would be ground referenced, so both the DAC and the tube device would be ground referenced.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
The short story is I'm trying to direct couple the entire chain, using no capacitors or other components to separate the bias voltages.
Why? I don't want to say because I don't want to derail thread.
Currently I just want to know what happens when I connect a ground referenced grid voltage to a floating circuit.
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,987
For the tube to work there has to be a certain DC voltage between the cathode and the grid. For an output tube circuit to have an audio signal on the cathode with 0 Vdc average value there has to be certain boas conditions. For a DAC to work the output signal has a DC bias (usually above ground for a single-supply DAC). When the signal ground of the DAC circuit is connected to the signal ground of the tube amp (as it must be for the signal return currents), the odds of all of these DC operating points lining up in the basic circuit topologies you posted is basically zero. With a single-supply DAC and a single positive DC voltage to run the tube amp, I don't see how things can line up. And even if they did, you'd have to add an external DC servo to keep the output DC level stable if there is any sizeable forward gain. Not exactly AC coupling, but still, a ***capacitor***!!! omgomgomgomgomgomgomg

An alternative is a classic differential opamp topology but with tubes. Ray Dolby did this for the audio channel of the first broadcast quality videotape machine. Input diff amp, voltage amp, push-pull output stage. The whole thing ran on plus and minus 200-300 V (working from memory). 0 Vdc on the input and 0 Vdc on the output.

ak
 

Papabravo

Joined Feb 24, 2006
21,159
The short story is I'm trying to direct couple the entire chain, using no capacitors or other components to separate the bias voltages.
Why? I don't want to say because I don't want to derail thread.
Currently I just want to know what happens when I connect a ground referenced grid voltage to a floating circuit.
There is no thread to derail, as near as I can tell, since there appears to be no viable solution to what you want to do. I suppose transformer coupling is also out of the question.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
The dac I have uses integrator servoes to direct couple its output at ground level. The amp I'm building is not of the topology I posted above, that was just an example to get the point across.

I already use integrators to keep 0vDC offset on the output of the amplifier so that is not a concern. I am still wanting an answer on what happens when I ground the grid of an input tube in a floating amplifier. Obviously from your responses it seems it won't work but in what way won't it work?

I am considering just simply using an isolation transformer and floating the entire shabang, dac and all, and hoping the HV output approach doesn't give me a bad day.
 

AnalogKid

Joined Aug 1, 2013
10,987
If the schematics you posted are not accurate, an answer is kinda difficult...

"ground the grid" is not nearly descriptive enough in a discussion of a system with multiple grounds. An accurate block diagram of the various inputs and outputs would help.

ak
 

benta

Joined Dec 7, 2015
101
If the supply to the DAC is truly floating, and the supply to your tubes is also truly floating, you can do what you want - DC-wise.
Your DAC input/output needs a signal ground reference though, which you do not show.

Your real challenge is designing true floating supplies, as parasitic capacitances and inductances are hard to eliminate.

Benta.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
I'm too lazy to create a readable schematic of the amp I'm building but effectively speaking from the perspective of the question the above schematic is relevant. The actual amp is going to be a cathode coupled input stage into push pull output stage with a DC nulled direct coupled output. It will use a bipolar supply and in terms of the question it might as well be the same as the schematic above.

I am just wondering how a floating amplifier will function with a earth referenced bias on the amplification stage. The way I see it it either won't function properly as a bias voltage since it is referenced to a different ground or it will reference the entire amplifier to earth ground.
Another way to ask the question might be how will a tube or mosfet amplifierbehave if the PK/DS are connected to a floating supply and the grid/gate is connected to an earth referenced bias voltage.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
Your DAC input/output needs a signal ground reference though, which you do not show.
I didn't show it but I did say it was earth referenced. Which is the basis of my question. What happens when I direct couple an earth referenced dac to a floating ampifier? The grid of the amplification triode would be on 0v from earth but the rest of the amplifier would be on a floating supply

Your real challenge is designing true floating supplies, as parasitic capacitances and inductances are hard to eliminate.
Hmm can you explain how parasitic capacitances and inductances effect floating supplies differently then a normal supply?
 
Top