My Vacuum Tube Frequency Synthesizer

[JimB]

Love it!
What a pity that you cannot get plug in breadboards with B9A valveholder bases built in!

JimB

 

[dr pepper]

Doesnt the term breaboard come from a tube prototype system?, one that resembles a breadboard or developed from a common practice using one.
I'm not that well up on tubes, however doesnt a pentode grids capacitance change with anode/cathode/screen voltage, if so you might want to regulate the Ht.
Or maybe use it as a means to tune the xtal, you might even be able to phase lock to a reference if you wanted.

Edit: I see you already mentioned regulation.

 

[JimB]

Doesnt the term breaboard come from a tube prototype system?

Not so much a prototyping "system", but in the early days of radio it was usual to mount the components on a wooden board similar to that used as a base for cutting bread.
Hence the term "breadboard".

JimB

 

[schmitt trigger]

Correct! Don't cheat on the power supply.
Use a tubed power supply. You know 5U4 rectifiers and the like.
If you can find them, include 0A2 or 0B2 type gas regulators.

 

[BobW]

It appears that regulation will not be necessary. When I was debugging the circuit, I had changed out the 5965 tube to see if the flip-flop would work any better with other types. I just discovered that I hadn't put the 5965 back in. It was operating with a 12AX7 (high mu and very high plate resistance). After putting the proper 5965 back in, The flip-flop output voltage swing is much higher and insensitive to the negative grid bias voltage. The bias supply can now be varied from –55 to –130 volts with no effect on operation.

I think the need for regulation tends to be overstated, and is really just an easy way to remove power supply ripple. Regulator use may be legitimate in low voltage high current circuits, such as 5V logic, to avoid using very large value electrolytics and to ensure that the devices safely stay within their narrow voltage range. But, these conditions seldom apply to vacuum tube circuits, where reasonable supply filtering should be adequate.

 

[schmitt trigger]

Got you....
then, in addition to electro caps, use choke filters

 

[atferrari]

If I can get it to work reliably, and with a reasonably small parts count, then the plan is to use tubes in the final build.


Pictures are attached. A couple of scope traces are included. In the first one, the scope shows the 38 kHz crystal oscillator waveform in the lower trace, and the 10.857 (divide by 3.5) blocking oscillator output waveform. Note that dividing by a non-integer number is not recommended, because it requires very fine and critical adjustment of the RC network. Integer division is fairly non-critical.
View attachment 111448

In the next photo, the scope shows the blocking oscillator trace again (upper), and this time the lower trace is the output of the JK flip-flop, which is triggered from the blocking oscillator output, and is running at half the blocking oscillator frequency. If the J and K inputs are tied to logic 1, or left unconnected as they are in this case, then the output alternates state on each trigger pulse.

View attachment 111447
Below are closeups of the breadboarded circuit. The upper breadboard is the crystal oscillator (left side) and blocking oscillator (right side). The 6U8 tube is lying on its side above the breadboard. The lower breadboard is the JK flip-flop, and the rectifier/filter for the -100V bias supply. The bias supply transformer is powered from a variac (grey box on the right) so that I can adjust the bias voltage accurately. The main power supply is the grey box in the upper part of the photo. It supplies both +150V and 6.3VAC for the filaments.
View attachment 111445 View attachment 111446


The mains power here is quite stable. So, I'm hoping that I won't have to add any special circuitry to stabilize the power supplies. But if I do, I'd probably look at a tube type regulator. It would be cheating to use a solid state regulator when everything else is built with tubes.

Thanks for the pictures and comments. My experience with valves is limited to a CW xmtr reformed for AM, which worked in the first try. A second experiment never worked. Next was CMOS / TTL logic and no more valves after that.

I expected something bulky but it is not, really.

A question more than a comment: the output seems rather distorted (and with clipping?). Could that be improved if necessary? I know it does not seem important here.

 

[BobW]

If you're referring to clipping of the 38 kHz signal, that's because I'm taking that trace from the feedback loop after the diode clipper. Taken from plate it's a fairly clean sine wave, though there's a small glitch whenever the blocking oscillator triggers. The glitch appears to be mostly due to inductive pickup in the long parallel tube socket leads. By bending them apart, the glitch got smaller but didn't completely disappear. Some of it may also be coming through the power supply line.

Another thing to note is that in the scope trace of the JK flip-flop, there's a very large leading spike when the state changes. This was because I had the wrong tube in the circuit. After putting in the correct tube, the spike is considerably reduced and the waveform looks much better.

I'll post some more scope traces later when I have some spare time.

Meanwhile, I think I've reached the limitations of the plastic breadboards. I need to start testing how the flip-flop will work sampling high frequency (500 kHz – 2 MHz) signals, and something is severely attenuating these signals.

BTW, I called this a frequency synthesizer project, but that may be a bit of an overstatement. The intention is to build a signal generator that will lock to fixed frequency increments, primarily 10 kHz increments (but possibly other increments too). Thus, it would give fixed frequency steps: 500 kHz, 510 kHz, 520 kHz... etc.
It works on the basis of sub-Nyquist sampling. If the VCO frequency is an exact multiple of the sample frequency, then the output frequency of the flip-flop will be zero. Otherwise, the output frequency will be:
fE=fVCO mod fS
where
fE is flop-flop output frequency (the error frequency);
fVCO is the VCO frequency (the frequency to be controlled);
fS is the sampling frequency.
The mod function gives the remainder of the first argument divided by the second.
So, fE will be proportional to the frequency error and will be integrated with a charge pump to control the VCO.
Alert readers will note that a VCO frequency that is 1 kHz too low will produce the same error frequency as when the VCO is 1 kHz too high. This requires the circuit to be designed in such a way that the VCO will always tend to drift in one direction. The simplest method is to use a resistor to slowly discharge the integrator capacitor.
This is essentially the Huff and Puff stabilizer circuit developed by Klaas Spaargaren PA0KSB, but done with tubes. More info on the method is here: http://hanssummers.com/huffpuff.html

 

[JimB]

This is a test post

 

[BobW]

Thanks for re-opening the thread Jim.

I hate to leave a discussion in limbo. So, after a 2-1/2 year hiatus, I have an update. I continued to work on this for a few days after my last post, trying to get the main VCO and phase comparator to work, but it was an exercise in frustration, and I set it aside. However, I did order some of the parts that I needed.

Also during that time, I would occasionally look over my design calculations, and eventually found some errors in my VCO design.

A few months ago, I built a PCB for the reference oscillator and blocking oscillator divider, since they were working fine, and were unlikely to change. I also made some breadboard friendly tube sockets. This made for a much cleaner experimental setup.

This week, I rebuilt the circuit and had another go. Trying to design the PLL loop filter involved a lot of guessing, because there were a couple of parameters that were virtually impossible for me to measure. After a lot of futile tinkering, I gave up on a pentode based phase detector circuit and replaced it with completely different design using a beam deflection tube. Then, more and more tinkering. Last night, I was just about ready to admit defeat, and rip it all apart. But before giving up, I made a couple more minor changes. Lo and behold, it started to work.

There is still a lot more to be done, but at least I now know that it's not impossible. The capture range is not very good yet. I also need to come up with some kind of indicator that shows when it's locked on frequency. At the moment, I have to rely on the scope and frequency counter.

In the photo, the reference clock is on the small PCB in the upper left. The rest of the circuit is on the breadboard. The frequency counter in the upper right shows the output locked on 1000.000 kHz (bottom line on the counter display). And yes, that's an honest reading.

See the attachment for the current schematic:

The reference oscillator has been discussed previously. I've now got the proper 20kHz crystal installed, and after passing through the blocking oscillator ÷4 frequency divider, It produces a very stable and precise 5kHz reference clock with narrowish (~150ns) pulses.

(The JK flip-flop, although rather fun, was unnecessary, and has been relegated to another forthcoming project.)

The ref clock pulses go to the control grid of the phase comparator tube. The tube is biased so that it's cut off except when it receives a clock pulse. The signal from the VCO is fed to one deflector plate. (It's not necessary to drive both deflectors; the voltage difference between the deflectors is all that matters.) The output at the plates is a pair of differential pulses that correspond to the phase relationship between the ref clock pulse and the VCO signal. These drive the 4 diode charge pump to produce the VCO control voltage. The cathode of the top diode is connected to the cathode of the reactance tube to prevent the control voltage from rising above the cathode voltage, and thus prevents the grid of the reactance from going positive (relative to the cathode). The reactance tube acts as a voltage controlled capacitance in parallel with a parasitic winding on the oscillator coil. Thus, changing the grid bias on the reactance tube changes the frequency of the oscillator. Other than the parasitic winding, the main oscillator is a standard Hartley circuit. The oscillator coil is a universal replacement type oscillator coil used in medium wave tube radios.