Controlling DDS with Variable Permeability Inductor Tuning

Thread Starter

K1JOS

Joined Aug 17, 2014
12
I have an old radio that uses a permeability variable inductor (~ 1uH - 3-uH) within a vacuum tube VFO (PTO). it offer very smooth, precise tuning over a full 360 rotation that covers 100 kc band spread. I want to replace the tube oscillator with a DDS but retain the permeable inductor to retain the old design as the electromechanical tuning interface. The new DDS/PTO would cover the HF band between 1.8 - 2MHz and that then goes into a mixer stage.

I am looking for a circuit concept that would allow the inductor range of 1 - 3uH to linearly vary a voltage from e.g., 0 -5vdc TTL to be read by an Arduino controlled DDS. The perm inductor is in a shielded, temp compensated enclosure within the radio. Any advise where to start would be much appreciated.

Jerry
 

DickCappels

Joined Aug 21, 2008
7,704
I would start by throwing out the DDS part -it can only make things worse (in my opinion). Just make it a wider range analog VFO.

What is it that you are looking for that you feel a DDS circuit would help you get?
 

Thread Starter

K1JOS

Joined Aug 17, 2014
12
What is it that you are looking for that you feel a DDS circuit would help you get?
The original PTO is electromechanically linked to an old fashion frequency dial read-out whose 1kc markings are quite accurate once mechanically adjusted to an external source like NIST radio stations WWV. The PTO has be be accurate within +/- 0.5kc with a perfect end to end spread of 2 MHz +/- 1kc. The space for the PTO is a shielded area about 2" x2" x2". I am not smart enough to design a new analog VFO that will fit the space AND that can mate well mechanically with the existing variable permeability inductor that is mated to the front dial read-out. I do know that there are DDS's that could easily fit within that small space. Someone gave me a idea that a tiny monostable vibrator could work where the permeability inductor varies the time constant of the PW which is read by an Arduino which controls the VFO output to the mixer stage. Sounds like overkill I know but the radios I am working on are old Collins and they rely on these exquisitely designed PTOs that are becoming impossible to service routinely. You need special test jigs, etc. You can replace these with an external DDS digital readout VFO but then its no longer a beautiful old radio with dials and knobs. I want to design this where every amateur radio hobbyist should be able to remove the analog VFO glob of components and squeeze in 3 small pcb boards: DDS, multivibrator and PIC uP without disrupting the analog tuning dial mechanisms
 

MisterBill2

Joined Jan 23, 2018
9,267
What part is it that fails on that PTO vfo system? That is a big question that I would have. One worthwhile revision would be to replace the tubes with transistors, avoiding FET devices in the oscillator section. Capacitors and resistors do age and fail, but short of abuse the coils and mechanical portions will last a very long time. So it may be difficult to stay authentic but not that hard to stay accurate and stable.
So the second part of the question is what part of the performance needs to be improved? A DDS system may be able to conttrol the frequency more accurately, but not with an aduino, which has no place in an analog radio system.
 

sparky 1

Joined Nov 3, 2018
576
The mechanism might be stepper. First look the adjustment appears to be turning a fine threaded rod back and forth.
If you have a gang of pretuned then it's a gang of solenoids. Some people like to precisely tune over wider range.
That might include 1 of N capacitors that could be selected with older rotary selector rather than encoders.

A few home made variable permeability inductors.
https://nandustips.blogspot.com/2015/04/permeability-tuned-crystal-radio.html
http://www.industrial-electronics.com/measurement-testing-com/rf_design_5.html

Video demonstrates the frequency stability and bandwidth of one design used to adjust a 40M oscillator.

 
Last edited:

MisterBill2

Joined Jan 23, 2018
9,267
Using that variable inductance in the Permeability Tuned Oscillator the resolution is infinite, while a DDS tunes in steps, which may be small, depending entirely on the look-up-table arrangement, which will need to have the appropriate number of bits. So going from a LMO-PTO to a DDS is a lot like traveling from Dallas to Fort Worth by way of Cleveland and Tampa.
And since a DDS is entirely digital in control, it is easy to add a digital frequency display, only needing to convert from binary to BCD. So adding a DDS would be reasonable for the second VFO for working splits, possibly. Use the DDS to set the listen frequency and the PTO to try different offset frequencies.
 

DickCappels

Joined Aug 21, 2008
7,704
I don't understand how you can integrate the two technologies and have it make sense. An inductance tuned oscillator is inherently analog and an DDS is inherently digital, though I could see a case for using a DDS to injection lock an analog oscillator, but this apparently not what you are after.
 

Yaakov

Joined Jan 27, 2019
3,492
I'm wondering whether it makes more sense to try to get a rotary encoder on the knob and calibrate the DDS to the scale.

The goal appears to be to preserve the experience of the tuning, so that might do it.
 

Thread Starter

K1JOS

Joined Aug 17, 2014
12
I hope the attached pictures can help you understand the objective. The first is the PTO which basically when opened has a glob of passive components, I think sealed up for thermal stability. I also think back in 1959 it was easier to just remove and replace the entire PTO. The second picture is the front main tuning knob of the radio which is hard-coupled to the PTO which is bolted down behind the front panel onto the top of the chassis. As you can see the permeability inductor shaft is not centered on the face of the PTO enclosure, it is low, off centered to allow it to mechanically couple with the front panel knob. The front panel knob ALSO is hard coupled to a gear mechanism that turns the frequency dials. Note I said plural, dials, because in fact there are two plastic face-dials each with a different gear ratio so that the first 360 rotation shows as going from 0 kc to 100kc and continuing into a 2nd rotation, the dial shows 100kc - 200kc. So keeping this original feature that tunes from 0 - 200kc in exactly 2 rotations of the permeability inductor is critical. Part of the calibration is adjusting 2 counter-rotating mechanical stops on the dial shaft (hard coupled to the inductor shaft). I hope you can see this is not dealing with some Zenith tabletop home radio.
 

Attachments

Thread Starter

K1JOS

Joined Aug 17, 2014
12
To cover an HF range from 1.8MHz - 30MHz, there is a crystal oscillator with 10 precision crystals to cover the amateur radio band segments and the XO is mixed with the PTO which provides a range of 0-200kc (actual range is 2.500 - 2.700 MHz) that then goes into 2 IF amp stages. I am no thinking maybe easier to forget the DDS/PIC idea and just go with the original oscillator schematic and build it with SMT Components on a small PCB and tie in the existing inductor. The tube can stay on top for appearance.
 

Thread Starter

K1JOS

Joined Aug 17, 2014
12
What angle does the PTO input shaft rotate through?
720 degrees, its done by double end stops on the PTO shaft. On the first complete 360 the first endstop is caught and is rotated along with the shaft through the 2nd 360 rotation and then it gets stopped. The mechanical design of these radios is very ingenious and rather precise for the late 1950's when designed.
 

Attachments

Thread Starter

K1JOS

Joined Aug 17, 2014
12
The display dial actually consists of two disks each having gear teeth around the circumference but each disk with a slightly different gear ratio. This allows the display of the front disk with line indexes to move more slowly than the rear disk with the kc numbers. I tried attaching a video to show this with my finger moving the rear disk only but the website doesn't allow mov files so a picture will have to do. You have to also consider that these dials are directly hard-coupled to the PTO shaft. Its a very intricate mechanical setup and so I need to retain the inductor mechanical characteristics in any new electronic design.
 

Attachments

Alec_t

Joined Sep 17, 2013
12,201
Instead of the variable inductor, if you could get a 2-turn good quality potentiometer you could couple its shaft to your fancy tuning knob and derive a signal for tuning a solid-state oscillator.
 

dendad

Joined Feb 20, 2016
3,909
I have converted a number of old crystal locked radios to tunable using Arduinos and Si5351 boards.
http://www.sadarc.org/xenforo/upload/index.php?threads/basic-arduino-vfo-for-crystal-replacement.25/
But, the latest version has an animated display similar to your dial. It is driven by an ESP32.
http://www.sadarc.org/xenforo/uploa...th-adaptive-step-rate-and-vernier-display.11/
I really like the look and operation of this one. The code is not mine, but I've modified it to suit my application.

But to try an answer to your question, have you read about the LVDT sensors?
https://www.sensortips.com/featured/lvdt-electronics-part-1-excitation-and-demodulation/
https://www.sensortips.com/featured/lvdt-electronics-part-2-interface-circuitry/
Maybe you could add a couple of windings to your coil to make one?

Or, just ignore the coil and use a multi turn pot attached to the tuning drive as @Alec_t mentions above.

EDIT: here is the ESP32 VFO code if you are interested...
And the forum post talking about this version...
http://www.sadarc.org/xenforo/upload/index.php?threads/bills-vfo.77/
Feel free to modify to your heart's content.
 

Attachments

Last edited:

MisterBill2

Joined Jan 23, 2018
9,267
The reason for replacing a very stable very durable and very reliable PTO system has not been made clear. The availability of identical parts for repairs only matters to those who are wanting to keep the whole system as originally built. The only parts in the PTO that are subject to normal electronic components are the resistors, the capacitors, and the tubes. Replacing the tubes with either bipolar or FET transistors, or even a high frequency amplifier IC would be an approach that would be far simpler, and much more likely to deliver good results on the first attempt without a whole lot of mechanical destruction that could not be undone.
To have a very accurate display of the receiver tuned frequency I have seen digital mixers applied to the VFO , the BFO, and the crystal oscillator, and the resultant frequency fed to an accurate counter. That counter can be as accurate as the user can afford, although resolution and accuracy. to less than 1 Hz is useless in the real world.
 

MisterBill2

Joined Jan 23, 2018
9,267
Certainly the resolution of that counter package is adequate but I saw no hint about accuracy. And certainly the added effort to include digital mixers is well worthwhile in that it compensates for any deviation in any of the oscillators, with the only offset required being compensation for the BFO offset from zero beat.
And I would also point out that the technical support from such on line offerings is usually limited to what is seen in the posting. Did it even mention the power requirements??? Or the amplitude of the input signal? Or how to set the offset value?
 
Top