Piezo-driven unstable carbon contact amplifier

This is my take on the "balance beam amplifier" from H.P. Friedrich's book, "Instruments of Amplification." This is essentially a relay mechanically biased to partway between on and off, with a soft contact  material that varies its resistance with pressure: carbon. Way back in the day, circa 1910, such a device would have been driven electromagnetically. I attempted such a design a few years ago without much luck. This time around I happened upon the very good idea of using a piezoelectric disc in place of a more complicated set of coils and magnets. The efficiency of the disc is very high, and has a very high impedance which matches well to crystal radios made with high count litz wire. The only better possible driver would be a sound powered phone aka balanced armature speaker, and I may try that in the future for even better performance.

The most efficient piezo disc I could find at Radio Shack is the driving element in this design. A 100k resistor is connected to it in parallel, and any weak audio source (such as that from a robust crystal radio) is enough to drive it. I took pieces of the carbon rod from the center of a carbon-zinc ("Heavy Duty") D-cell battery to make the amplifying relay contacts. One piece is secured to the center of the piezo disc and the other is on the end of an adjustable balance arm. Counterweights on the other end of the arm adjust the resting pressure between the contacts.

Two D-cell batteries in series provide power - this runs to one carbon, across the unstable contact to the other carbon, through a 50 ohm winding on an audio transformer, and back to the battery. As the vibrations from the piezo disc vary the resistance between the carbons, the current through the transformer varies and is output from another winding to a sound powered phone.

This thing DEFINITELY amplifies! On a strong local station the output can be uncomfortably loud. Adjustment is very sensitive, and it often will take off in self-oscillation.

Here is an example of the amplification of a small audio signal. The first image is the audio source driving a 1k load, and the second is the amplifier driving a 1k load while fed by the same source.

It's not without added distortion, but the signal is amplified from about 80mVpp to 160mVpp.
Below are some images of the device itself. Not too pretty, but it works well and was made with only hand tools.

 The current path includes the forward portion of the balance arm and would include the pivot itself if I did bridge it with some thin wire. I expect it would add noise and decrease amplification otherwise.

Below are some waveforms from this thing when it is self-oscillating. There was no input when these were taken.

Completed Curve Tracer Circuit

Long overdue, the more or less final functional circuit of the curve tracer. The phase of the AC inputs is important so that the base bias is stable while the device  under test receives power. I used separate transformers for each. R6/R7 is a potentiometer, as is R8/R9. Vvar is supplied by a variable linear regulator with low dropout.

I used TL712CP comparators for U2 and U1, and IIRC a LM833 dual opamp for U3 and U4.

Below is the circuit as thrown together.

It's not pretty, but it works reliably for any small BJT or FET. Outputs to the oscilloscope are posts on the right side - only the ground connection is clearly visible. The rocker switch at the top switches between PNP and NPN modes by converting U3 and U4 from inverting to noninverting unity amplifiers.

The Curve Tracer is Functional

After continuing to ponder why the analog circuit was not functioning, I decided to play with it a bit more. I found one wrong connection and added a schmitt trigger input to the reset comparator.

The results are encouraging. Obviously it needs some refinement. But this curve set is unmistakable as that of a NPN. The ugly drooping of the collector-emitter bias is likely due to the high output impedance and poor voltage regulation of the passive supply. A voltage follower might solve this issue.

I also redesigned the part that supplies collector current, and thought about how to use it for PNP devices, though my first idea is wrong. The schematic below reflects these changes, including the wrong idea.

---

After more fiddling around, and some more circuit changes I started to get nicer plots like this:

It looks like I have base bias stability problem. More tinkering on this is sure to come.

Steps!

So the previous efforts with etching payed off. The circuit works as planned, and while the intervals aren't quite as even as I was hoping, it should be easy to correct with resistor substitution. See below for a scope shot of the ramp output of the summing network and the buffered square waves feeding the decade counter.

Next step is to feed it through an amp with gain variable from 1 to 3 or so, and then a voltage follower, and then I'll really have something.

I think I need a better curve tracer..

While it has been enlightening to look at some basic I-V curves  of two-terminal devices, to ease my experimentation with home made transistors I need something that will step the signal to the third terminal. In the case of cuprous oxide or point-contact transistors this takes the form of a stepped current, and if I do as planned and experiment with crude thin film field effect transistors it would be a stepped voltage.

Even though there are cheaply available commercial solutions, I want to build it myself, of course.

And since I'm really perverse,  I'll be relying on analog techniques.

My first go at a design was through using LTSpice, but I have not yet been able get it to work as it does in the simulations. Big Surprise, right? It's certainly not the first time I've been misled by circuit simulators. 

It works just fine for a variety of different step intervals and step numbers, in simulation.

The above circuit seemed like a reasonable way to go about things. I use a voltage doubler to create a current pulse once per cycle timed to coincide with the middle of a flat portion of the stepped ramp applied to the base of the device under test - DUT, or Q4 in the schematic. The ramp is generated by the interesting configuration of Q2 and Q1, which are fed by square waves from the comparator U2, buffered by Q3. The basic idea is that one cycle of the square wave, via C2, charges C1 through Q1. This turns on Q2 which reinforces the original voltage on the base of Q1. In theory it should remain steady until another pulse arrives through C2, stepping the voltage higher until the comparator U3 triggers, turning on M2 and shorting the base of Q1 to ground, discharging both C1 and C4, resetting the circuit for another accumulation. In practice the circuit proved very unstable, either ramping up and not discharging, or else discharging and not ramping up. There was a critical setting of R6/R7 that allowed repeated ramping, but there was a varying dead time between each ramp, and it is neccesary to be able to control both the peak ramp voltage and step interval over a wide range.

I've tried to duplicate the problem behavior in simulation, but have been unable so far. The circuit as built, minus the 5V regulator:

One way I prototype circuits that shouldn't require a ground plane is this - the copper clad is sawed into little islands, with long strips glued down with cyanoacrylate for ground and supply voltages. The 16-DIP socket also has its own "surf board."

After browsing for more ideas on how to create the stepped ramp (that don't include using a microcontroller), I decided to try out something with basic digital methods. Looking around in my collection of parts I found I had several each of surface mount CD4013 and CD4017, a dual D flip-flop and a decade counter. I decided to use the flip-flop to generate a square wave from the 60Hz input by pulling the clock, data, and reset pins high, and feeding the AC signal to the set pin via a large resistance and/or small capacitance. The Q pin should then have a relatively clean square wave. This gets buffered by a CD4049, which I only had in DIP version, and fed to the CD4017. The output pins of the CD4017 have resistors tying them all together and buffered/amplified by an op-amp. As I want the voltage step value to be relatively constant, the resistors are nonlinearly spread in value from 10k to 200k. Output 0 has 200k, and output 9 has 10k. The circuit is not yet finished, but it did present an excellent reason to try out hand-drawing and etching a PCB.

This was a test of two methods new for me: hand drawing with a sharpie marker, and using 1:1 31% HCL and 3% H2O2, or muriatic acid from a hardware store and hydrogen peroxide from a drugstore.

I decided a little breakout board for a 16 pin surface mount (SOIC) chip would make a fine test of my ability to create usable single sided PCBs at home. The results were excellent! This is way easier than messing with heat transfers of toner and ferric chloride. Plus, the etchant takes on a lovely emerald green color. Supposedly this mix will not have to be discarded; it can be rejuvenated by re-oxygenation by bubbling air through it or adding more H2O2. A capful of muriatic acid may eventually be needed, says the internet, but only if that pretty green color can't be achieved through oxygenation.

Freshly etched bare board.

half populated next attempt at stepped ramp generator for curve tracer

I suppose even the most extreme radical fundamentalist homebrew electronics devotees such as myself can get dragged into the future. All the traces came out fine on the above board, which is my first ever home-built circuit to use surface mount chips. As you can see, I don't currently have a stock of surface mount resistors, but that will change soon as this first attempt went so well for the SOICs. That 5V regulator missing from the previous circuit can be seen attached to this one. Also seen is the oddity of a DIP being attached as if it were a SOIC. Too think of how much more time it takes me to assemble circuits dead-bug style, or Manhattan style, or ugly style, or on perf board, when compared to how fast I made this PCB and how easily these chips when on..    ..and I was already not feeling very well today. If I were to drill a few strategic vias before etching I could even do this for 2 sided boards without much added difficulty.

Quick and Dirty Spot Welder

So a while ago I was gearing up for my first attempts at building a crude vacuum tube - a triode - in the manner described in "Instruments of Amplification," by H.P. Friedrichs. The electrodes need some form of welding for assembly - soft solder won't cut it. Rather than getting a small welding torch and teaching myself to silver solder or straight-up weld very tiny pieces of metal, I decided to build a spot welder out of parts from my Junque Pile.

Any means of passing sufficiently high current through a joint of metal such that it briefly melts and fuses will work as a spot welder. I have several power transformers from microwave ovens laying about, and so one of them seemed a likely candidate for cannibalism.

I took a hacksaw to the high voltage secondary winding while being careful to not damage the primary winding. After cutting off one side of the secondary flush with the core, the rest can be hammered out using a small piece of wood as a punch. I paralleled up four lengths of 10 gauge stranded wire and wrapped as many turns of this as would fit through the core windows previously occupied by the secondary winding. This ended up only being four turns. 

The output is fed to the work piece via any number of methods, though my favorite is a modified hot wire insulation stripper. I exchanged the original nichrome elements for thick copper and rewired it to put the copper bits in series with one another when the gap between them is bridged by something I wish to spot weld.

This image shows the incredible simplicity, and a clamp mounting for larger work pieces.

A Mixture of Things

So many projects, such scattered documentation, such great effects of ADHD type behavior. Polyprojecting, if you will, seems an apt description.

Since when does blogger.com only allow me to add files that have already been uploaded to google docs? Ah, only if I allow a potentially unsafe cross-scripting as described by NoScript.

Should I write about an attempt at an antenna-powered transistor radio? 

The circuit attempts to use all the unwanted RF energy incoming on the antenna to power a single transistor amplifier - which is fed the wanted signal from a high-Q LC tank. It worked about as good as any passive radio I've built, though sometimes heterodyne whistles were produced, indicative of actual amplification. The schematic is somewhere in my piles of notes if anyone if interested.

---

Perhaps I should write about my attempts at making a simple vacuum triode from scratch?

Or about the promising attempts at building a regenerative radio with said homemade triode?

 The circuit was patterned after some of the earliest regenerative radio circuits - in hopes that the marginally marginal performance would still be enough to coax out some real RF amplification and/or a solid RF oscillation. I can say with great confidence that the device worked extremely well as a detector, and that it would at times oscillate, but I could not get stable heterodyning, which suggests the oscillations may have been the result of nonlinear negative resistance from such a  gassy tube rather than gain as one would expect from a proper vacuum triode.

The filaments were taken from a 12V tail light bulb, the grid and plate were made from bits of scrap, the structure was assembled with a homemade spot welder, the feedthroughs were sealed with epoxy, and the envelope itself with non corrosive RTV. The vacuum was provided by a brand new dual stage roughing pump supposedly capable of 10 microns, though a check with small hand held tesla coil/violet wand indicated pressures more like 100-200 microns.

---

Maybe this post should be about the crystal radio for the FM band that I built, along with the three element yagi antenna I made for it. Yes - this is a passive radio that receives FM, and demodulates it with a phase discriminator rather than just slope detection. This allows remarkable clarity and frugality of signal use. I can't claim that this is an original idea - I mostly just followed Robert Weaver's work: http://electronbunker.ca/FMCrystalSet.html

It is tuned by fringing capacitance - a small square plate of copper is moved by a long leadscrew, which tunes higher if brought away from the resonator, and lower if brought closer. 

To get a sufficiently high Q for this thing to work, a helical resonator was chosen over an LC tank, as the latter simple can't give sufficient performance in the 100MHz range, where the helical resonator promises a Q of 2000 or more, in theory. 

The inside of the case is lined with copper, forming the outer surface of the resonator. With standard 3-element yagi with a folded dipole active element, without audio amplification I can listen to 1 or 2 stations comfortable. Add in some audio amplification, and the number jumps to about 6 or 7 stations, some of which come in with hi-fi quality.

Strange IV curves from Galena, interesting ones from other things.

I recently repaired the video board on my 54600A oscilloscope, which unlike my other scope, has an XY display mode and so can be used as a curve tracer. See below for a very simple front-end circuit needed to make this happen.You'll have to invert channel 1 to get the plot in the right quadrant(s).

I decided to take a look at some various cat's whisker radio detector minerals when excited with either AC or pulsed DC. I was able to observe negative differential resistance in several materials. This includes flame-treated galvanized steel, iron pyrite, and even galena! Any of these could be put to use as an amplifier or oscillator. There were also various different curves produced that had more ordinary diode characteristics, some bi directional, some not.

This is a healthy response from a piece of flame-treated galvanized steel probed with a sharp point of copper. Copying Nyle Steiner, I heat a piece of the metal one one side till it's red hot and throwing off sparks from the other side, and look for active spots on the side not hit with the flame. The above curve is displaying at 1V/division horizontal, and 10mA/division vertical.

 One of my home made detector stands. Right now it is loaded with a hunk of iron pyrite, and the whisker is made from phosphor bronze. The next two IV curves were captured from this setup.

 This one is quite obviously useful for amplification. Applying AC made it much harder to find a stable example of this characteristic. With DC excitation only, it was fairly easy. 500mV/div horizontal, 10mA/div vertical

This is a more typical response for the iron pyrite, bidirectional diodes with a forward voltage of about 2.5V

  This is one of the curves from phosphor bronze on galena. There appears to be a backward leaning curve that might be useful for sustained oscillations. 100mV/div horizontal, 1mA/div vertical

Again, an interesting apparent negative slope region from galena probed with phosphor bronze. 50mV/div horizontal, 1mA/div vertical

I could also generate very weird curves like this one, which while not showing a clear negative slope region certainly implies that one must exist, even if it is a very abrupt transition. Galena is a more interesting material than I suspected. 50mV/div horizontal, 1mA/div vertical

While once again researching the topic of homemade transistors, I have come across references to some people getting continuous wave oscillations out of galena, either in a normal cat's whisker detector configuration or with two point contacts, like the earliest transistors. Perhaps I have a suitable sample of the mineral to achieve this!