Tuesday, October 8, 2019

Another look inside a transistor voltage and current relationship

I set the generator to produce a 10mv signal riding a 1700mv bias. The 1k emitter resistor will 'self adjust' to about 1 volt producing the bias required to allow a 1 ma collector current. The emitter resistor would drop AC signal and produce degeneration C1 provides a path for AC. C2 provides a path for AC to the load (R3). The .55ua pk-pk signal provided by V2 when set to 10 mv produces 10mv across R3. It appears the amp produces unity gain? Yes 10 mv in produces 10 mv output but look at the currents. I(V2) is .5ua pk-pk, I(C1) is 100ua pk-pk, and I(C2) is 100 ua pk-pk. So the current gain is 200. Next let's increase the input and see what happens.
I simply increases the AC to 100mv. I(V2) is about 5ua pk-pk. The increase created an increase in base current as would be expected. V(out) is 100mv with 100mv input. We have unity voltage gain again. The current gain is more than 200? Why?
Hint: When a semiconductor current increases its resistance decreases. re is 26 ohms at 1ma. re decreases as current increases. gain is the ratio of Rc to re.
It is a good time for a simulation using this circuit. You could try removing C1, increasing and decreasing C1 and C2 and changing the frequency. 

Saturday, October 5, 2019

voltage controlled resistor, voltage controlled current source, or CCCS?

Years ago the transistor was thought of as a voltage controlled resistor thus the name TRANsfer reSISTOR. Then someone decided it was a current controlled current source. The voltage controlled current source is popular today. The debate is clouded with the fact that ß is not a fixed factor it varies with current and voltage and temperature. In this series I'm not going to define the actions within the transistor but will run SPICE programs to demonstrate them. Hopefully the sims will speak for themselves. It would be a good time for you to run some too. You can vary the component values and probe the circuits to see what is happening.


5 ua in the base produces 1ma collector current. I put the 1 ohm resistors in the circuit to make measurements and later they can be changed to see the effect.
Using 656 mv bias gives simuliar results. Ohm's law says voltage , current and resistance go together. Trying to have current without voltage or visa versa just doesn't work. Therefore the debate is it voltage controlled or current controlled. More on that later. 





Stepping the load resistor from 100 to 5.2k ohm gives basically the same collector current. Injecting current into the base controls the collector current BUT there are other factors involved.




Probing the collector voltage displays one of these factor. As we change the load resistor the collector voltage varies and this has some effect on the collector current.




Probing the base current shows it is effected by collector voltage too.




Fixing the bias and allowing the collector voltage to swing displays the current variation. (changing Vce)




Adding a signal shows the current swing responding to it.












adding R1 and probing its current shows the carriers in the collector will produce some current when given a path.




I moved the ground reference to the emitter. In the previous shot it could appear the signal was driving the collector too.




Applying 1 volt to the collector encourages collector current.




With 2 volts it is stronger.




With 6 volts the amp is looking much better.




With a current source a transformer can be used to couple the signal. This transformer is 100mh primary and 50mh secondary. It produces good results as you can see.








100uv in producing 15mv out with PNP germanium general purpose transistor. Time for a build!




10 uv will drive an earbud.




My old ears don't hear well below 800Hz so for me it is ok. You could adjust the bandpass by changing the capacitors.




The gain drops well below my local AM station so that should not be a problem.

EDIT: I assembled this one and it drives my earbud with the AF generator set to minimum. The signal drops out after insert a 20db attenuator.





Monday, September 23, 2019

Adding a second mosfet

Adding a second mosfet produces a large increase in output.
The circuit could be tweaked for more porformance.

Ear bud or small speaker amp

100uv signal will drive the earbud.
1mv signal produces 400ua which is half the ear bubs max rating.
10 mv is near 4ma drive which is enough to burn out my earbud.
My earbud is rated 100dbm and 1ma max drive.
This amp will require some limiting to keep from damaging my earbud.
What if the output stage used 2 mosfets? I have the 2 -200Ohm resistors limiting the output stage current. I tested the circuit without the 200 Ohm resistors and it produced 20 ma signal. That may be another project.

Saturday, September 21, 2019

The pushpull with a splitter feed

The first step is finding a valve that will produce the desired current through the transformer and determine the signal needed to produce the desired output. The generator output is being varied as I try different tubes.








The green is from plate to plate.
The blue and red are the plate currents.
NOTE: These circuits could use some tweaking. I just put them together to see how they functioned. (picking values out of the air.)

Friday, September 20, 2019

Transformer - Z matching - what can we control?

I will make a series of screen shots with the transformer response displaying the circuit response to varying conditions and then pose the question what can we control and what do we have to work around? If we were having a custom made transformer made to our spec's we could just build a circuit and plug in the transformer but that is not the case. I measured a filament transformer inductance and will be using it as a typical example.
This is the reference circuit. The first two readings show the circuit is a reasonable Z match.  It is dropping about 1/2 the signal in the plate circuit.

decreasing the load decreased the internal signal drop.

increasing the load increased the internal drop.

Back to the original circuit. Rp is dropping half the signal so the circuit is matched. The only way to increase the output is to increase the signal (100 volts here).

If I could lower the generator Z it would increase the output.

Lowering the generator more gives more increase.

Lowering the turns ratio increased the output.

If you took the T725 challenge you can calculate the required signals for a 120 volt to 12 volt transformer to match your speaker.
The questions are:
1) What would the valve anode load resistance need to be to match a 10:1 turns ratio with a 16 ohm load?
2) What tube would match?
3) What signal swing would be required for 250mw output?


Saturday, September 14, 2019

6 volt 100 mw amp

A low part count 100mw amp to breadboard next rainy day.
EDIT:
I tacked this one on a board tonight. It drives a 2" speaker to room filling volume. I changed R2 to 200 Ohm. I didn't see any 100 Ohm in the drawer and was not going to the shop to get one.
EDIT:
I connected an ear bud to the output and an antenna coil and diode on the input. It plays radio (weakly). It could use a little more signal than I got with my clip lead 'radio'. That would not be hard to do. It draws around 10 ma with no signal at 6 volts.It will work at reduced output with 3 volts.