April 16, 2014:
Revised: v1.1

A Permeability Tuned Oscillator

front view of transceiver
Analog VFOs still have a place in transceivers given their typically low current requirements and relative simplicity. The PTO VFO described here has been built using readily available hardware to provide a conventional VFO tuning mechanism at very low cost. 


Building amateur radio transmitters and receivers often requires the construction of some type of variable frequency oscillator (VFO). This allows the transmitter or receiver to be tuned across the frequency band of interest. Most VFOs tune across a specific band, for example from 3.5 to 4.0 MHz for the 80m amateur band, or from 14.0 – 14.35 MHz for the 20m band.

Most analog VFOs are tuned with a variable capacitor. These days, it's more common to find a digital VFO, such as one using a phase locked loop (PLL) or a direct digital synthesis (DDS) oscillator chip. These often use a microprocessor and a rotary encoder for tuning.

Analog VFOs have the significant advantages of simplicity. They may also have lower phase noise and current consumption. However, analog VFOs can suffer from poor stability caused by temperature changes, supply voltage variations and poor construction.

Analog VFOs also often require the addition of some form of reduction drive for tuning the variable capacitor. This mechanism allows the user to tune across the 100 to 500kHz of the VFO’s tuning range with a tuning rate of 5 to 20 kHz per turn. Since variable capacitors require a 180 degree rotation to go from maximum to minimum capacitance (or vice-versa), a mechanical reduction drive of up to 100:1 can be required. Such mechanical drives are becoming more and more difficult to find, and their size and weight can be a problem.

Permeability Tuned Oscillators

One alternative approach is the permeability tuned oscillator (PTO) where the frequency is adjusted by varying an inductor rather than with the traditional variable capacitor. The inductance in these VFOs is usually changed by rotating a ferrite or brass core into the centre of the inductor’s former.

It’s relatively easy to cover the required frequency range with 20 or more turns of a typical threaded metal bolt. However, the mechanical arrangements of many such homebrew PTOs often leave much to be desired. As these VFOs are tuned, the tuning knob typically simultaneously rotates out from the panel of the equipment. This can result in the tuning knob ending up anywhere from 30 to 50 mm out from the front panel. Not only does this look odd, but any vibration can result in undesirable frequency shifts. It also leaves the tuning control exposed to accidental damage.

Why Make a PTO?

I required a VFO for a 20m transceiver. The transceiver’s IF was about 8.6 MHz  so I wanted the VFO to tune from about 5.4 to 5.8 MHz. An analog VFO would allow me to minimise current drain. However, I did not have a suitable variable capacitor for the VFO nor a reduction drive. That’s when I started to review the work done by others on PTO-type VFOs.

Other PTO designs I discovered seemed to generally fall into two categories – The majority of designs that used the simple “tuning knob on a stick” approach where the tuning control was wound out from the panel, and the second category, with only a few examples documented, where considerable time and effort had been spent on the mechanical design of the PTO in order to provide more conventional tuning of the PTO. In addition, a few of these latter designs also explored improved tuning linearity. However, most examples I found seemed to require the construction of fairly elaborate mechanisms often with specially made parts.
The more I looked at these ideas, the more I believed I could achieve a viable PTO with basic hardware i.e. Parts easily purchased from a hardware store. There were a number of obstacles that I had to overcome, but in the end, I think I managed to overcome all of them.

Anyway, this design described here is the result.

Mechanics of the PTO Tuning Mechanism

The key to this design is the mechanical arrangement which allows the PTO to be tuned in the conventional manner i.e. without the tuning knob rotating into or out from the front panel. You can see the mechanism in the picture at the top of the webpage. It depends on a long threaded shaft which is free to rotate in place thanks to the careful use of some locking nuts, and a parallel fixed shaft, in this case a short length of bamboo (although metal would have been better!).

Rotating the tuning knob rotates the main threaded shaft. It is fixed in position between the front and back panel, but is free to rotate. A set of lock-nuts at the back-panel end of the shaft allows smooth rotation. This threaded shaft also drives a nut which is attached via a solder lug to the tuning slug. This slug enters or exits the main VFO inductor and varies the inductance.

The tuning slug has a second solder lug on it which slides along a second parallel bamboo rod. Bamboo was used because I happened to find that these were just the right diameter to fit the hole in the solder lug, and, of course, since it was cut down from a large packet of kitchen skewers used for barbequed meat in the kitchen, it was cheap and readily available. Since this rod is mounted parallel to the tuning shaft, it prevents the tuning slug from rotating and ensures it slides smoothly in and out of the main inductor.
Ideally, there should be an additional spring fitted between the solder lug on the bamboo rod and the front panel to maintain a little tension on the tuning mechanism. This would improve the feel of the tuning mechanism as well as probably reducing any frequency shift due to vibration. However, I’ve yet to find a suitable spring.

The PTO VFO “chassis” is made from three pieces of blank PCB, one forming the base plate and the other two forming the back and front panels. A pair of thick brass washers are soldered onto the front and rear panels. These have the same diameter as the hole for the tuning shaft. They provide additional support for the tuning shaft.

Because I couldn’t find a threaded shaft long enough in shops where I live, the tuning shaft was built from one 50mm long 6mm diameter bolt and a matching 30mm long 6mm diameter bolt. Both bolts were “cheesehead” slotted zinc-plated bolts which were readily available from several of the local hardware stores. 6mm diameter nuts and bolts were used because the diameter of the unthreaded portion of these bolts perfectly fitted into the shaft hole on the tuning knob I used.

The first 50mm long bolt is inserted into the PTO chassis through the back panel and two nuts are threaded on to form the rear mount. The first nut is threaded on such that the tuning shaft can just (and only just) rotate. The second nut then locks the first nut in place. It also acts as a limit for the tuning slug, preventing it from coming completely out of the PTO inductor. I could have added a further nut onto this shaft if the second (lock) nut had not been in a suitable location to provide this function by itself.

The slotted head of the second 30mm bolt was then cut off with a hacksaw. This allows the unthreaded shaft to fit into the tuning knob. A nut was partially threaded onto the end of the first 50mm long bolt and then the cut-down 30mm bolt was inserted through the front panel hole and threaded into the other half of the end-nut. This had to be attempted perhaps five or six times until the two shafts were absolutely axially aligned and tight.
 An additional nut could have first been wound onto the 50mm bolt to provide a further mechanical tuning limiter, to limit the extent to which the slug enters the inductor, but again, this was unnecessary in my case. I allowed the slug to fully enter the inductor until limited by the (pink plastic) inductor former.

I would have liked to have added a pair of nuts on the inside of the front panel like those of the back panel, but the hardware I had available did not allow this. As a result, I could not completely remove the free play in the tuning shaft although it is quite acceptable as it is.

The inductor former was mounted using epoxy glue onto the front panel. The tuning slug was fully wound into the coil while the glue was drying and the former was carefully aligned and clamped to ensure it was parallel with the tuning shaft and bamboo support shaft.

PTO Coil Former

Ideally, I would have liked to use a temperature stable inductor former, such as a ceramic, PTFE or even phenolic type. However, none of these were available so I just tried a plastic tube made by cutting down a felt-tip pen just to see if the mechanical arrangement would work. As it turned out, the temperature performance of the felt-tip plastic turned out to be fairly good, so the bright red former ended up in the final PTO.

Permeability VFO Inductor Tuning Slug

Some initial tests with all sorts of metal hardware quickly showed that the variable inductor used in the main PTO tuned circuit had to be tuned using a brass core. Tests using other materials failed to give satisfactory results. Unfortunately, I was unable to find any brass bolts (I would have preferred to use a 6mm diameter brass bolt with a matching nut) so I was forced to improvise. Searching around the hardware stores, I discovered some 240V mains plugs rated at 15A (They look like they are good for more like 30A!). A similar plug is shown in the photo below.

230VAC plug

Figure 1 : Heavy duty 240VAC mains plug

The “pins” on the connector were beautifully finished solid brass slugs. These were each mounted in the plug with a 3mm bolt.

The plug was surprisingly cheap, and very easy to disassemble. One of the thinner (6mm?) diameter pins was found to be ideal for the task. The 3mm bolt is used in the plug to hold it in place, threaded into an axial hole in the end of the pin, and in this PTO, this is used to hold the two solder lugs in place.

Winding the PTO Inductor

Some testing with a simple close-wound coil and the brass slug on my bench resulted in a non-linear tuning rate for the PTO. I found that the tuning rate was too slow at one end of the tuning range but much too fast at the other.

Looking at another design (See reference 2), I noticed it used an inductor which had an initial close-wound section, and then a series of wider spaced turns with slightly increasing spacing until the final few turns of the coil which were once more close-wound. This can be seen in the photo opposite.

Commercial example of PTO coilI began to understand the very clever solution this designer had identified and which I’ve subsequently come across again in several other PTO designs. The initial close-wound section of the PTO coil sets the (rough) minimum inductance of the tuning coil. The next section of the coil then uses a variable pitch to give linear tuning across a specific frequency band. The final few close-wound turns are likely to provide a good mechanical finish for the inductor.

I decided to try something similar. The result, after several attempts, can be seen in the photo below.

Side view of PTO VFO showing coil turns
Figure 2 : The PTO oscillator inductor uses variable spaced turns

The tuning rate was much improved over the original close wound coil, and gave an acceptable result for my 20m transceiver.

The PTO Oscillator

As noted earlier, this PTO had to cover a relatively wide range, from 5.4 to 5.8 MHz. Since I wanted to maximize the range from the variable inductor, I wanted an oscillator design that avoided, as far as possible, using any additional capacitors in the tuned circuit. This ruled out the Colpitts and Clapp oscillators. Similarly, for simplicity, I wanted to avoid any tapped inductor. That ruled out the Hartley oscillator.

A configuration that I’ve used before is the Franklin oscillator which is an approach I've used before and documented here on my website. This design relies nearly completely on just the tuned circuit inductor and capacitor. I adapted an existing design from PY2DYW which allows the PTO inductor to be grounded, useful in the mechanical arrangement I used, and which also included a buffer amplifier. This isolates the oscillator from the mixer load and provides an output level of about 2Vpp for use with diode balanced mixers.

The oscillator schematic is shown below.
PTO schematic
Figure 3 : The prototype PTO oscillator used low cost NPN transistors

A close look at the prototype photos will show that I didn’t bother to fit the trimmer capacitor, C5, which is shown in the schematic. I found I did not need to add this because my coil and capacitor combination gave almost exactly the tuning range I was wanting.


The simple chassis was temporarily covered with a tin-plate shield. This will be replaced later with a more robust cover made from blank PCB. The PTO output was terminated on a low cost RCS socket mounted on the rear panel. DC connections were provided via a 1000pF feed-through capacitor (for +8VDC) and a grounded solder lug.

The tuning coil was not wound particularly accurately, as you can see from the photo above, but the tuning rate was none-the-less quite acceptable. In total, 22 turns of the tuning knob were required to cover 500kHz, from 5.35 to 5.85 MHz. At the lower (CW) end of the PTO tuning range, the tuning rate was around 10 kHz per turn. Across the subsequent SSB section of the 20m band, the tuning rate steadily increased to 20 – 30 kHz/turn. This tuning rate felt very comfortable when using the PTO with the receiver in the 20m transceiver.

The PTO oscillator draws about 20mA from the regulated 5V rail.

Stability was reasonable, but not adequate for long term use. I tried two capacitors for C4 (150pF) in the VFO. The first was a ceramic type, the second a polystyrene type. I measured the VFO frequency drift for about 90 minutes in each case.

Drift of the simple PTO VFO
Figure 4 : The prototype PTO oscillator was tested using ceramic and polystyrene fixed capacitors to determine the basic drift characteristics for the PTO VFO

After about 45 minutes following power-up for each capacitor type, and in the reasonably stable workshop environment with the VFO covered with a thin tin-plate cover and without any other insulation, both versions became quite stable. Neither drift curve is great, but these plots show that, with the right combination of capacitors, say a 100pF poly and a 47pF ceramic, the different drift characteristics of all the components would probably roughly cancel, and the resulting VFO would be quite stable.

I may add a little “huff and puff” stability circuit to improve this still further at a later stage. Space has been provided for this small additional board in the PTO chassis.


Here is a list of some previously published PTO designs some of which provided valuable insights for this PTO design.

1.    Walter Horn, I1MK , “A High-Precision Permeability-Tuned VFO”, QST July 1964

2.    A commercial PTO
which was used in the ITT Marine R700M HF receiver is described here

3.    KD1JV has published a number of designs using simple PTOs on his website. One simple example is described here for his MMR-40 transceiver

4.    WA6OTP describes his very well known PTO design here

5.    A low cost 40m receiver with an equally simple PTO using a “drinking straw” coil former is described here

Want to go back to the main page? Click here to return directly.