Sunday, September 20, 2020

Final thoughts on the PCT

 Millivolt in millivolt out not very impressive?

Look at the current! Microamps in and milliamps out. There has been many who deny the transistor being a current amp. They say it is a voltage amp. Others say it is a power amp. Ohm's law relates voltage, current and resistance. It is a matter for others to argue. Anywho, I added R2 so you can see Q1 base current is microamps.

The circuit is a common base amp direct coupled to a common collector amp.

I redrew it to show the 2 stages. 

In the days of hollow state amplifiers we used "those amplifying transformers". The interstage transformers were step up. With the high input Z amps they could boost the signal with the transformer. Notice T100 is 5:1. This boost the current. Since the PCT is a current driven device this is a good thing. Microamps in milliamps out.

Another amp I had on file.

The three junction transistor or remote base transistor compared to the PCT.

Here we have the SCR structure. Interesting? In another time we also had the SCS which is the same structure with two gates. By adding an anode gate we could have a PUT which is another interesting device.

And now the SCR equivalent with discretes. Look familiar? My PCT discrete is the inverse. PNPN rather than NPNP. The pinout is different too. Can you see it?

It would appear grounding the gate and driving the cathode with proper battery polarity should produce an amplifier? The semiconductor doping would be different but it could be a fun experiment.

EDIT: The simple two stage receiver is such low part count it could be a weekend project to wind a couple transformers and build it. Just two PCT transistors or two pair of Ge BJTs. A couple of AF output transformer could work?
What would the turns ratio be for an 800:32 ohm earbud transformer?

Friday, September 18, 2020

discrete build of PCT - proof of concept


This is my test board and the transistor I will use to do the test. It is a simple layout. The ohmmeter is across the emitter and base. I apply power to the collector and base threw the 1k resistor. With this jig a junction transistor will read a low value when forward biased. When the collector voltage is applied it will go down about one half. A PCT will read a negative resistance when the collector voltage is applied. The test jig could be powered with a 6 volt battery. I like using the supply so I can adjust the voltage and see the reading with different values. 

I connected an emitter to each end terminal and clipped a test lead to a base - collector pair then connected the far end of one pair to the base connection on the test jig.


The meter reads the base - emitter junction resistance.

I set the supply to 6 volts and the meter displays a negative resistance. I have PCT transistor action. There you go proof of concept.

revisting the point contact with discrete equivalent

We can make a Lambda Diode equivalent with two JFETs and it function much like a Tunnel Diode.

We can make a SCR equivalent with two transistors and it will function much like a SCR.

We can use complementary transistors and make a remote base transistor which acts much like a Point Contact Transistor.

The PCT collector is formed after fabrication. This forming causes doping which produces the PN hook. A PCT will not function without the PN hook. If we simply make a point contact diode with two points we have to parallel diodes but no transistor action. forming one point creates the hook and transistor action. Clear as mud? Lets look at the PCT structure.

The two diodes are just two PN junction sharing a common wafer until the collector is formed. The phosphorous in the wire is diffused into the P cathode forming an N collector. Fig 309 is the equivalent circuit made by junction rather than points. The center P is the remote base. in a SCS it would have a pin connected to it. The SCS is basically a SCR with a cathode and anode gate. The two gate connections allow the SCS to become a SCR or PUT. Fig 309 was a common device in the early days of semiconductors. Anywho the secret of the PCT is the junction formed when the collector is pulsed with current and the remote base transistor hidden in the device. Once you understand that you can use complementary transistors to make the equivalent circuit and you should be able to understand there is a great difference between the PCT and the junction transistor. Anywho lets take a look at the equivalent. I show the voltage and current in the analysis.

The forever dispute about transistors is do they amplify current, voltage, power etc.

The PCT has an alpha greater than 1. It amplifies current. Yes?

Micro amps in mill amps out. The current flowing threw the load resistor drops voltage. The current being higher the load voltage is higher. Careful don't start that back and forth again! 

You can work it out in your own mind. It could be Q2 is a common collector amp and what I marked Emitter is actually  the remote base?

Wednesday, September 16, 2020

comparing the Ge to the Si transistor amp

The same circuit using Ge and Si transistors. The only real difference is the 1.5 volt and 45 volt supplies. Q3 has just over 20 volt Vcc it could use a little adjusting. The Si circuit could function with a higher voltage. Anywho it is the same circuit both would fail if the supplies were switched.

Tuesday, September 15, 2020

Loop stick antenna NOS and replica

 The antique that I used as a pattern.

My replica. I will add a slug and screw. Inserting the slug it varies from 100uh to 300uh.

Friday, September 11, 2020

Adjustable antenna coil from scrapsI wound a coil on the pen barrel and inserted the core.

This post gives instructions on winding an atenna coil.

This is good for what it is but what if you want this instead?
It is a bit to expensive for most bottom feeders.

This one was about half price but still out of my reach.Besides it is usually more fun to make your own when you can.

This one offers a starting point.

 We need a screw, nut (insert) and tube (pen barrel).

A core , screw and nut
I drilled a hole in the screw and inserted the core lead. A little solder should keep it in place.

I would a coil on the pen barrel and assembled them.

I soldered the first one.soldered and glued glued the second. I think the glue helped.

The first adjust from 30uh - 100 uh.

The second adjusts from 40uh - 200uh.

The core is 3/4" and the screw is 40 tpi so the coil requires 30 turns for a full range adjustment.

If your core doesn't fit a tube you have available you can make one from craft paper or paper tape.

Check the junk box and think about it.

Enjoy the hobby.

The connector is a 
strip of sheet metal about 1" long. I wrapped it around a rod and crimped it with needle nose pliers to form a loop. Then holding the loop split the ends to form mounting ears. They had a drop of glue applied then masking tape to hold them in place. a more traditional treatment would be to cut a strip of cardboard such a s a cereal box and split it to allow the tabs to be inserted. then gluing it in place. I think I may make a couple more using that idea but for now the proof of concept is looking good. Well? Not looking good but functional. The store bought ones are prettier but the price is right on these.

Tuesday, September 8, 2020

Earbud amp using transistors from GE computer boards - 2

I had a little shop time and made a box for the amp. I just used a piece of scrap flashing. It measured 3" X 3-1/2" that should be big enough for the amp and battery. When I can get back in the shop I'll do the build.

It will need a volume pot, input jack, output jack and battery holder.

I changed the transistors and changed some bias in the hopes to match the transistors better and reduce excess power consumption. After I do the build I will measure the current drain. 

Earbud amp using transistors from GE computer boards

I plucked some GE and TI 1960's era germanium transistors from an old computer board. This project will provide a use for 3 of them. I laid it out in a modular form so I can make adjustment to biasing by simply clipping in different resistors. If time permits this build will begin later in the day. First step is locate or build a box for it.

Basically I will lay out the transformers and transistors then add the bias resistors. Simple enough? Time will tell.

I will be using transistors manufactured for IBM by TI using the IBM assembly line. The data in the sim is based on "similar" transistors. Time for the build after I do my outside chores.

Sunday, September 6, 2020

Using the RC filter to form a differentiator

Take a look at the amp response from the last few post and consider this circuit.

The designer is in control. You can establish the gain and the frequency response by inserting a RC circuit on the input, feedback or output of the amp.
Something to think about the next time you see a circuit with all those extra unless parts.

Saturday, September 5, 2020

putting a filter on an amp

The FB amp as built before gives a gain of 100 which is established by R3 and R4. Rc and Rl are equal so we see 1/2 on the output. If we make Rl 10 times Rc we will get a gain of 100 if the transistor is capable (Hfe or Beta are at play here also) This could be the RF amp on a BCB receiver and serve quite well but assume we have a short wave station or flame thrower of a local at the high end of the band what can we do? Consider the filter we used in the last post.

I placed the filter across the FB resistor and ran a frequency response. As you can see it will virtually kill any short wave signal. The filter I used earlier was set to 1600Khz. You can adjust the values and pass a higher or lower frequency.
This series was in response to a question about the LM386 output filter. I think a review of the datasheet may shed more light on that. I have hear more than once the "cheap" LM386 amps are trash and prone to oscillations. I have also heard the "expert" on a radio forum say the filter is not needed. This makes me wonder if they left the did not use the filter and created the stability issues? The number one issue we have as home builders is keeping the output from the input especially when higher frequencies are concerned.
Consider a short wave build operating at 4 Mhz with an AF amp that will amplify 4 Mhz. Now let the AF amp oscillate. You just "built a regenerative receiver".
Use a ground plane.
Keep lead short and run them close to the ground plane.
Separate (shield) input and output.
Consider using filters to control the band pass for your amps.
Most of all enjoy the hobby!

Friday, September 4, 2020

Filter the noise out

 Examine this series of shots with the idea that the bypass is filtering certain frequencies. Pay attention to the generator frequency, generator Z and load Z.

From the series of shots you should be able to see some relationship between the generator Z and the bypass resistor value.

In order to function we would set R bypass less than Z gen. The filter is less effective when Z gen is 50 ohms.

Z bypass also needs to be less than Z load. Simply stated the bypass must be lower than the source and load.

You should be able to see a reaction to the input frequency.

That reaction is defined as 1 / ( 2 PI * F * C).

The calculated value for the cutoff or bypass frequency is 1.6Khz.
Final screen shot shows the response to 1600hz.
The last amp I built was very stable with the input shorted but would pick up strays when the input was open. Putting a low value resistor across the input helped but it would then load the tuner so a filter was called for. I used a simple L filter and it worked but this filter is used on LM386 amps so it could be a good alternate.

Wednesday, September 2, 2020

low power speaker driver

With less than 10ma drain on the battery driving a speaker is not to bad.

I'll have to take time to build this one. I filed it here so I don't lose it.

I built it and it sings. With my generator on minimum it is loud. With the attenuator on 20db it is just barely there. I can not crank it up much it hurts my ears.
With a 6 volt supply it draws 4.5ma.
with a 9 volt supply it draws 7.25ma.
It passes the finger buzz test too.

Tuesday, September 1, 2020

TUP - TUN complimentary amp

I used the universal design for the preamp and added 2 diodes to keep bias in working range on the output the 2 4Ohm resistors help prevent cross over distortion.