can please someone explain this to me?

[krusty75]

Hi there,
sorry for this question..maybe can be silly for many of you..but I'm freaking out trying to understand how this circuit works..
the idea is to use it for bias scaling..with screen scaling
here is the circuit..and then will follow the question ;-)

                      NPN
Vin _____________ _C__     __ E_ ______________ Vout
                 |    \___/     |
                 |      |B      |
                 R4     |       |
                 |______|       R3
                 |              |
                 R5             |
                 |     PNP      |
                 |__C_     __E_ |
                      \___/     |
                        |B      R2
                        |       |
                        |       __________________Vscr
                       GND      |
                                R1
                                |
                               GND

Vscr = from 200V to  400 V 
Vin = -100V 
Vout = ??

I tried to simulate the circuit with PSPICE and to change Resistor values in order to understand how it works..
Vout follows Vin with a coefficient related to Vscr...
I think that the two transistor works in their linear region of operation...
Suppose that I want to get Vout in the range -90V to -10V (the -90 volts is obtained when Vscr is 400V and the -10V when Vscr is 200V)
So now I have to calculate the resistors value in order to obtain this result!
I think that R4 and R5 are used to make NPN transistor works in linear region but I don't know how to calculate them ..
The same applies to R1 R2 and R3...
How is the formula I have to use to calculate them (I think their relations.. i.e. R1 is half of R2 and R3 is double than R2)..
I actually don't know how to start
I found that if R1 is very small related to R2 and R3 the Vout lowers (I mean MAX VOUT goes toward zero volt while if R1 is 2 or 3 times smaller than R2 I can get -100V on Vout when Vscr is 400 and Vin is -100V)
Is there anyone that can help me to calculate the proper values for R1 to R5 in order to obtain , let me say a value Vout (in a specified range) when I have fixed Vin and the range of Vscr?
Thanks a lot in advance and sorry for my english
Krusty

 

[smoking-amp]

Looks like the E and C terminals on the NPN transistor need reversing in the diagram.

Vscrn and R2 set a current thru the PNP. This current controls the current in the NPN via R5, which then acts like a current mirror with gain. The NPN output current then feeds back as negative feedback thru R3 to the PNP. The PNP E terminal acts like the inverting terminal on an Op. Amp.

R2/R3 will set the gain ratio approximately (finite loop gain) between Vscrn and Vout. R2/R3 = Vscrn/Vout

Don

 

[krusty75]

yes you are right, Emitter and Collector are swapped in my diagram on the NPN transistor..
But what about Vin...
I mean...in this circuit Vout should follow Vin...
you wrote R2/R3 = Vscr/ Vout ..but what the relation with Vin?
Thanks again.

 

[smoking-amp]

Vin just needs to be a few volts or so more negative than the maximum -Vout so the NPN can function. Other than that, higher Vin just makes the transistors run hotter. All other resistor values pretty much just follow from how many volts to drop between Vin and GND and how much current you want to run in the transistors (max diss. in the transistors sets the limit or max. current available from the -bias/Vin supply).

Don

 

[krusty75]

Thanks a lot!
Now seems to be clearer to me! Now I will make some tries with PSPICE!
I really apreciated your help!
Krusty.

 

[smoking-amp]

Hi Krusty,

It's an interesting auto bias circuit, I haven't seen this approach used before. Is someone using this in an amp design? I guess it could be used to avoid regulating the screen supply.

Presumably the R2/R3 ratio is set to somewhere near the g1/g2 Mu for the tube. (I think one will still have to adjust this initially for a new tube though.) AC power variation will be automatically compensated by the tracking bias.

Interesting things may happen under clipping conditions though. The g2 voltage would be dropping, if derived from the same xfmr as the plate voltage, so this will be avoiding excessive g2 current, but the g1 bias will be decreasing too, maybe entering grid 1 current mode with signal peaks. Might want to put a small protection diode across the g1 to cathode to stop positive grid swings.

(if you think about this scheme a bit, it really is the equivalent of NO regulation on the screen AND the bias supplies coming off of the same xfmr., so they are tracking, but more accurately here time wise)

Is only one of these circuits used for both output tube's bias?

Don

 

[krusty75]

Hi Don, Actually my idea is to use this circuit to have a "power scaling" in my guitar amp...
so in a fixed bias power amp if I reduce the voltage on the anode and screen I have to change bias voltage in proportion to these to voltages to have the almost the same transfer function..
so probably I can use R2 and R3 to have a voltage drop in order to choose the right PNP transistor (no sense to have a PNP transistor with a max Vce = 400V or so on right? if my output is at most -100V)
I don't know if it's used in some production amp...many have power scaling feature but I don't know how it's done!
Hope this helps!
and thanks again.
Krusty75