Wednesday, February 27, 2019

AF Amp with gain control

I put a coil. V-cap and diode feeding the amp and found the volume to high. I added a gain control.R15 is a 2K rheostat. It could be placed in  Q4 collector circuit. This would limit  the low end.
With the rheostat in Q4 circuit it swings the output from 2 mv to 18mv.

Sunday, February 24, 2019

AF Amp circuit operation

Q4 is a preamp.
Q3 is a phase splitter.
Q1 and Q2 are a push-push amplifier.
* connecting a center tapped transformer in the output circuit would produce a push-pull amplifier. This is how I was taught 50 years ago. Today the terminology may have changed?
Anyway the key is Q1 and Q2 bases are being driven out of phase. Q1 and Q2 operating class B (conducting 50% of the time) feed the load a full cycle with one transistor biased off while the other conducts. (or AB with one being throttled down while the other is driven harder.)

When Q2 is driven into conduction current flows DOWN thru C1 and R3 to ground. Charging C1.
When Q1 is driven into conduction current flows from C1 DOWN thru Q1 to ground and UP from ground thru R3. Discharging C1.
Here we see the phase relationship between the signals.
The one thing to watch here is not to use to small a capacitor. A large capacitor will receive a large charge and provide the power on the next half cycle.
I set R1 and R2 to 4 Ohms with C1 at 1000ufd. 5ma battery drain and 180mv output with the same input as before.
It is a good earbud driver as I made mine but you can adjust the bias and add more power as desired.
Good luck!

Friday, February 22, 2019

another attempt at a low power low Z amp

The transistor output Z is to high to feed a speaker or headphone. One way to overcome this is to raise the collector current. A higher current will reduce the output Z but it cost power. I want a battery powered set that will not drain the battery to quickly. A BF2040 stage will feed a 50 Ohm load with 10ma current drain. I want to try and do a little better than that. We could use an output transformer but I want to use parts available off the shelf. While winding a transformer is not extremely difficult is not beginner level. I need 1 mv to drive my earbud. This circuit works faily well.
Driving it with 25uv gives me the 1mv out.
100uv gives  over 4mv. So far it looks good but what about the power drain?
1.5ma current drain at 9 volts is not bad. This could be my circuit for my new radio build.
It drives my earbud with the output set to min on my AF generator. I connected a speaker to it and could hear the tone across the room with the generator on about 5%.

My favorite hearing aid amp is very good but it requires transformers.
This one works well. I have some transformer in the parts been that work in it. If you don't have any transformers and don't care to wind any the circuit above is a good alternative.

Monday, February 18, 2019

100 mhz amp I -O and script

I set  the generator to 100 uv at 75Ohms. Half the signal is being dropped in the generator.

 I moved the generator and battery on screen to show the whole picture.



You can cut and paste the spice script below.

************************************
Version 4
SHEET 1 880 880
WIRE 128 80 128 64
WIRE 400 112 400 64
WIRE 32 144 -32 144
WIRE 336 160 128 160
WIRE -112 176 -128 176
WIRE -32 176 -32 144
WIRE -32 176 -48 176
WIRE 32 176 32 144
WIRE 128 176 128 160
WIRE 592 192 528 192
WIRE 624 192 592 192
WIRE 400 224 400 208
WIRE 528 224 528 192
WIRE 624 224 624 192
WIRE 80 240 32 240
WIRE 160 240 80 240
WIRE 32 256 32 240
WIRE 80 256 80 240
WIRE 160 272 160 240
WIRE 528 336 528 304
WIRE 576 336 528 336
WIRE 624 336 624 304
WIRE 624 336 576 336
WIRE 32 352 32 336
WIRE 400 352 400 304
WIRE 400 352 368 352
WIRE 416 352 400 352
WIRE 576 352 576 336
WIRE -32 384 -32 256
WIRE 80 384 80 336
WIRE 80 384 -32 384
WIRE 160 384 160 336
WIRE 160 384 80 384
WIRE 368 384 368 352
WIRE 416 384 416 352
WIRE 80 400 80 384
WIRE 368 512 368 448
WIRE 384 512 368 512
WIRE 416 512 416 464
WIRE 416 512 384 512
WIRE -656 528 -656 512
WIRE 384 560 384 512
WIRE -656 624 -656 608
WIRE -560 656 -656 656
WIRE -656 672 -656 656
WIRE -656 768 -656 752
FLAG 80 400 0
FLAG -128 176 RF
FLAG 32 352 +Vcc
FLAG 128 64 +Vcc
FLAG -656 768 0
FLAG -656 624 0
FLAG -656 512 +Vcc
FLAG 400 64 +Vcc
FLAG 384 560 0
FLAG 576 352 0
FLAG 592 192 OUT
FLAG -560 656 RF
SYMBOL npn 128 240 M270
SYMATTR InstName Q1
SYMATTR Value BFR193
SYMBOL res 64 240 R0
SYMATTR InstName R1
SYMATTR Value 820
SYMBOL cap 144 272 R0
SYMATTR InstName C1
SYMATTR Value 10n
SYMBOL res -48 160 R0
SYMATTR InstName R2
SYMATTR Value 7k2
SYMBOL cap -48 160 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL res 16 240 R0
SYMATTR InstName R3
SYMATTR Value 1k4
SYMBOL res 112 64 R0
SYMATTR InstName R4
SYMATTR Value 7k2
SYMBOL voltage -656 656 R0
WINDOW 123 24 124 Left 2
WINDOW 39 24 152 Left 2
SYMATTR Value2 AC 1
SYMATTR SpiceLine Rser=75
SYMATTR InstName V1
SYMATTR Value SINE(0 100u 100meg)
SYMBOL Misc\\battery -656 512 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V2
SYMATTR Value 9
SYMBOL npn 336 112 R0
SYMATTR InstName Q2
SYMATTR Value BFR193
SYMBOL cap 352 384 R0
SYMATTR InstName C3
SYMATTR Value 10n
SYMBOL res 400 368 R0
SYMATTR InstName R5
SYMATTR Value 17k
SYMBOL ind2 384 208 R0
SYMATTR InstName L1
SYMATTR Value 1µ
SYMATTR Type ind
SYMBOL ind2 544 320 R180
WINDOW 0 36 80 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName L2
SYMATTR Value 10µ
SYMATTR Type ind
SYMBOL res 608 208 R0
SYMATTR InstName R6
SYMATTR Value 1k
TEXT -688 832 Left 2 !.tran 0 5u 4.5u
TEXT -680 864 Left 2 !K1 L1 L2 1
***************************************
The code is the text between the ******.
The file extension is .ASC
Blogger will not let me post .ASC files?

Sunday, February 17, 2019

100 MHz amp - comparing sim to calculated

This is the circuit
Ic(Q2)
Vbe
I bias
V emitter
Q1 collector current
Input and output
Time base adjusted to show waveform

The math and the model follow each other fairly well. The math was using ß = 93. I think my model uses ß = 125. I expected a little variation because of this.

I assumed a 1k load and used the 4T:14T FT37-61 transformer. The transformer could be adjusted for the desired Zout.

complete receiver - less tuner

Edit: I will have to do some testing but the build doesn't function. The AF is great but the RF amp failed. I may have a bad hookup. It will require some testing.I injected a signal into the detector and it looks good.

EDIT 1: I had R5 connected to +Vcc. Connecting it to -Vcc should fix the problem. I'll try that next shop time.

EDIT 2: I moved R5 from +Vcc to -Vcc and it was like flipping a switch. Looking good now it just needs a tuner. That will be my project for next week.

Saturday, February 16, 2019

The frontend - RF amp and detector

I used the P416 Germanium transistor for the Rf amp.
Looks good now to add a detector.

The green is the generator output.
The blue is the detector output.



It has 30 uv pk-pk input and 150 uv pk-pk output.
Here is the circuit.
Time for a build.

Transistor datasheets Russian Germanium

I checked the datasheets to make sure I was right in my last post. The transistors I've been (MP40) using are rated 1MHz. The P416 is rated 40 MHz. They said a TRF with 1 RF and 2 AF stages was as good as a regenerative with 2 AF stages. Maybe a P416 RF stage is in order?

How many stages should a receiver have?

I posted a link to the book in April 2017.

Russian Data Sheets

If you can't read them here you can follow the link and download the book.

Bypass the power supply to prevent cross feed

Using the old Germanium transistors provide some RF immunity. They have 1 MHz high frequency limit. This might be another reason to build a TRF receiver. A crystal set can perform quite well. With an AF amp even more so. I think this is part of the advantage the old timers had. The limits of their components were in their favor. In the final build I may add an RF amp which will help with sensitivity AND boost noise which could have been below the detector threshold. It is give and take. We get the bad with the good. Then a band pass filter could be used to help reduce out of band noise.
Anyway I like the little amp and it could be used in my next build. I don't have any variable capacitors yet. Peebles Originals has some in route. When they arrive I will add the front end and see how it works out.
Anyway this is the final schematic for the AF amp. I added some bypass caps (in the red box).
For the input to reference AC ground I need to bypass the battery. Without the bypass the signal is passing through the battery to ground. This could produce cross feed and cause instability.


Friday, February 15, 2019

amp with mosfet output

Here is the circuit as designed.
I printed the schematic and taped it to a board. I thought I would drive a pin at each dot in the schematic. That cherry board as to hard so I drilled a hole at each dot and screwed a lug to it. Then it was just connect the dots with the parts.
It is very stable. No motor boating or whistling. I connect my fluke meter to the output on it lowest range and read .02 millivolts. I connected my AF generator and set the output to 5 millivolts then checked the input and it was less than a millivolt on the input. I set the meter to read db amd adjusted the input to -50db the output read about -20db. The reading were at different resistances but the meter would not read the signals at the low levels used in the test. (the reading are relative to the voltage drop at each point)  I connected a speaker from a telephone hand set and it was room filling volume. I think this one is the best so far. Maybe it will be put on a metal ground plane with a front panel for the input and output jacks, switch and volume control.

Note: A 2N7002 MOSFET should work. The other transistors are general purpose Germanium AF transistors. You could use silicon transistor but would need to adjust the bias.

Wednesday, February 13, 2019

TRF radio from a 50's magazine

This is from a 50's magazine. It has a tuner and detector feeding an AF amp. I made a test circuit and tried a couple of my in stock transistors to see how they match the sim. It appears the AC128 is a good match. I only changed the transistors and bias resistors in the series of test.

1 mV will drive my earbud to a good level so it look good. You can use a high Z headphone connected where L5 is.

Adjusting the capacitor values could improve the low response but it is good enough for my ears.

I reversed the transistors and adjusted the bias. The shots are forward and reversed transistors. 

The circuit works both ways but the "proper" hookup has about twice the output. When I rotated the transistors I did need to change the bias. This will be my AF amp for the 50's receiver. I need to put a tuner ahead of it.


Tuesday, February 12, 2019

Transistor Theory and Practice by Rufus Turner

 This book was published in 1954. It gives the theory and construction techniques in an age that the devices were still being developed. A very good read for anyone desiring to understand the devices. Mr Turner made transistors from high back Germanium diodes. His articles were published in several magazines in the 50's.

Transistor Theory and Practice

Monday, February 11, 2019

Audion Oscillastor in sim

The Audion Oscillator is a very straight forward circuit which should perform well. The common problem with an audio oscillator is oscillations so they took an audio amp and added feed back to make it oscillate.
It looks like one of the early AF amps with a tank in the output where the head phone would be. R1 is the output tap and C4 is the feedback. The CB amp is non-inverting so the feedback is in phase with the input.
The green is across R1 and the blue is the tank. If this circuit was build with a plug for the coil it could be used in a multi band set.
I adjusted the sweep to get a better look at the signals. Looks like it is time for a build.

Sunday, February 10, 2019

effect of reversing an alloy junction transistor in the circuit

Making an amp is fairly simple at first glance. Just put some forward bias on the transistor and apply a signal. The circuit can misbehave is several different ways. A transistor has three regions it can be biased to.
1. saturation
2. cutoff
3. active
With no bias it will be at cutoff.
With to much bias it will saturate.
When it is biased in the active region it will amplify.
It can be biased too much and then be over driven and distort too.
So the question is why a circuit will not operate when built "properly" but will work when the transistor is rotated.
I took two transistors and put them in my transistor hfe testers.
The one on the left is in the socket properly. The one on the right is reversed. They both show hfe but the one connected properly is much higher.
I reversed them both. The one on the right is correctly inserted now. So what does this tell us about the circuit in question?
hfe is the forward transfer or ß. The collector current will be Ib * ß. If the circuit is biased into saturation with the transistor inserted properly and you reverse it ß is reduced so it is no longer saturated. In order to give more help I would have to see the circuit in question but the simple answer is to reduce the bias.

Note: The transistors being tested are germanium micro alloy.

Wednesday, February 6, 2019

building a 2 stage PNP amp - with design data considerations

I want a low Z input and a low Z output. For max power transfer you need Z match, To reduce power loss you want Zgen << Zload. If the generator Z is zero ALL power is sent to the load. If the generator Z is infinite no power is consumed. When Zgen = Zload half the power is consumed in the generator. For this amp I want 50 ohm in and out. So what choices have to be made, what options are there?

So from the chart CB 100k - 500k Rout and CC 150K - 300K Rin. CB 30-1k Rin and CC 1K - 20K Rout.
So a CB feeding a CC would be:
30 ohm in - 100k out feeding 150K in - 1K out.
Loading the previous stage is to be avoided so I listed the low values for each parameter. I want 50 ohm in and out . The 1K out is the only problem. I found some 1300 : 8 ohm transformers for $.70 each so I use one and it all works out.
The circuit is very simple. 2 transistors, 4 resistors, 4 capacitors and 1 transformer.
100uv input will drive my earbud.
The frequency response is flat.
The as built. I took an old battery apart and nailed the connector to the breadboard. Using finish washers and screws allows changing the transistor. The MP40 works well.
With the battery plugged in it is ready to try. I have some V-Caps on order from Peebles Originals. When I get them it will be time for a build.
For the math minded a chart from TI.
EDIT:
I put a speaker on the output and it was easy listening level. I put a high Z headphone across the transformer primary and it was good volume too.

EDIT:
I put a tuner and antenna on it and tried some other transistors. It worked with all I tried but I think the GSS109 output and P416B was the winners. I tuned a station and changed parts. The signal was fading a little so my test could show different results if I tried it again. Some transistors made little difference while others seemed to give a boost.