HYBRID POWER AMPLIFIER - Please Help!

[gianmaria]

HELLO!
I need help to solve a problem.

The joint principle schematic is about an hybrid power amplifier made by two stages:
- all voltage gain: vaccum tubes
- current gain: BJT.

The "rest" current is regulated by diodes, as you can see.

Now I need a solution to enter from the tubes stage to the bjt stage but please first read the following points:

- The first solution will be to enter by two capacitor on points (equal) A and B.

- The second solution will be to enter by only one capacitor, for example on point C.
This solution is not very good because the signal will pass through the diodes, with some non linearity. But of course it works.

NOW, the question is (number three):

1) I don't want to use capacitors to couple the two stages.
2) I don't want to use transformer to couple the two stages.

I already solve this problem!

Infact I have used a special vacuum tube stage, with a dual voltage, that comes out only with the signal and NO dc voltage.

In this case I can direct couple the vacuum tube stage to the BJT stage, entered on poit C, without capacitor... OK! It works.

BUT BUT...

as I already said, the non linearity of diodes makes the sound bad, especially low frequencies.

SO and however, I don't want to enter on point C but in the same time on points A and B.

BUT...

in the BJT schematic, I CANNOT JOIN POINTS A AND B OTHERWISE I WILL BURN THE AMPLIFIER!!!

SO WHO HAVE THE SOLUTION TO THE PROBLEM WITH PLEASE A LITTLE SCHEMATIC TO EXPLAIN ME BETTER??

I think is so simple that I cannot see it.

Thank you very much, sorry for my english, if anything is not clear please just ask.

G.M.

[anatech]

Hi gianmaria,
How do you deal with the offset before the tubes are wared up and stable? What happens if there is DC drift? I can't see your tube stage, but I can't imagine getting around those shortcomings. You need a coupling cap.

-Chris

[ilimzn]

It would help if you post the schematic of the tube driver stage... it may be suitable for putting the bias voltage generator in series with it's tube(s)...
Otherwise, here are some pointers:

1) Your bias (standing current generator) is potentially thermally unstable (you do not give values of the resistors so I cannot be sure), if the resistors are small in value, you cannot compensate 6 diode drops (Vbe of your tripple darlington) with 3 diodes (hopefully they are on the heatsink with the output devices!). Use a Vbe multiplier instead, or even two complementary Vbe multipliers (divide the bias voltage into two parts), then inject signal in the middle, which would be similar to your point C. The only way to ensure the Vbe multipliers do not act as an extra nonoinear stage, is to bypass them with a suitable capacitor or capacitors.

2) If you join points A and B your bias will drop to zero - and unless your amplier is unstable, absolutely nothing bad will happen except for very high distortion.

3) Your tripple darlington makes all 3 stages work in class B which generates excessive distortion. Your transistors can turn on quickly but not turn off as you do not provide any way of the excess base charge to escape (i.e. a resistor between base and output of the amp or between emitters of transistors in each stage). If this stage ever goes into oscilation, it will likely destroy itself.

4) Your bias generator is fed from resistors conencted to + and - supply rails of the BJT portion of the amp. This quite unnecesairly load the preceeding tube stage, which, unless you are using a large triode or pentode, increases it's didtortion. Using current sources instead would be a better deal, and also prevent the idle current from being dependant on rail voltage. In your schematic bias current is heavily dependant on rail to rail voltage as any change there produces a varying voltage on the resistors in series with the diodes between the bases of the BJTs. Bias is also dependant on the input signal (thoug this is less of a problem) since that also changes the current through the same resistors.

[ilimzn]

Originally posted by anatech
Hi gianmaria,
How do you deal with the offset before the tubes are wared up and stable? What happens if there is DC drift? I can't see your tube stage, but I can't imagine getting around those shortcomings. You need a coupling cap.

-Chris

Bot if you have a DC servo setup. If so, it needs to be fed from the output of the amp, not of the tube stage. DC servo is not that much more difficult for tubes than for silicon...

[anatech]

Hi ilimzn,
If the tubes are not conducting yet, your DC offset = supply voltage. A couple hundred volts anyhow. There is no DC servo that will help in this situation. Same for a bad heater contact. Bye Bye speaker(s).

Direct coupling a tube stage to a solid state output stage is never a particularly intelligent idea. Sorry.

-Chris

[ilimzn]

It is quite typical (and of course expected) that the DC servo is coupled with output protection, which disconnects the speakers until the proper operating conditions are reached.
There are several direct coupled amps that use a similar approach, the most popular was produced in thousands of pieces - the Luxman LV-103/4/5(u) series (too bad the direct coupling is one of it's few virtues...)
Of course, if you do not want to have any series elements on the output, it is not at all difficult to disconnect the driving signal until proper conditions are present (again, you use the output of the DC servo to sense for this).
Of course, if you want to keep it as simple as possible and avoid any switching of the audio signal, (and there really is no reason why this would be a lesser approach), you can always capacitor couple as you say.

[anatech]

Okay, the Luxman units are not the same topolgy as gianmaria is thinking from the sounds of things. I am very familiar with these units. They do incorporate a SS front end, the tubes are not run from extremely high voltage (high enough I guess). The tubes are almost in circuit as a curiousity. I do agree they are nice sounding amplifiers.
I fixed many of these over the years and under warranty.

I would like to see what gianmaria is thinking about for the rest of the circuit. Otherwise there is no helpful info I can give.

-Chris

[gianmaria]

Hi to everybody and thank you all.

Here you have a less principle (for the BJT stage) and a principle schematic for the tubes stage I'm using at the moment.

As I repeat, now I'm entering from the tubes stage to the BJT stage on point C of the circuit, of course without any capacitor seen that from the tubes stage we have only AC voltage output (i.e. musical signal).

It works, but I think is not the best solution.

I repeat my question:

I don't want the signal to pass through diodes, but I want it to be connected directly on points A and B, WITHOUT ANY CAPACITOR OR TRASFORMER.

But I cannot join points A and B otherwise the amplifier will work very very bad.

Thank you very much, G.M.

[gianmaria]

here you have the tubes stage

[jacco vermeulen]

Gianmaria,

why go for a triple darlington instead of driving the output devices with mosfets ?
You want the SS section to be all BJT?