https://canadianvintageradio.com/

Radio Restoration, Victoria, BC (2026)

06/03/2026

After my recent sojourn into the wonderful world of vintage solid-state electronics, its about time some 'hollow-state' goodness graced my workbench... so, the Geloso G209 receiver has now made the short 'leap' from the side table into the limelight.
With some luck, I may find some time to give it a 'once-over' tomorrow.

06/03/2026

I spent some time earlier today checking a few different NOS transistors and diodes from my junk box and comparing the resultant curves using either the bipolar transistor curve tracer or the 'Octopus' as appropriate. Attached to this post are some examples, including some of the older germanium and silicon parts.

06/03/2026

I am not sure what happened, but my earlier post from today has been wiped (text, photos, captions, links and comments!). So I have recreated it below - luckily I had retained a copy of the text...
---------------------------------------
Busy weekend, but I managed to find some time to wire-up the curve tracer, test it, debug it, and complete the labelling.
As I started the wiring, I decided to make a mod to the transistor socket fixture - I had made the original one from a small IC socket, but found the spacing between the two rows of sockets a little cramped. Also, I decided to add a switch to allow switching between the two rows of sockets, thus allowing rapid comparison of curves between transistors mounted in the sockets for matching purposes. That complicated the wiring (of course) and more drilling of the box. With the bipolar curve tracer wired-up, I gave it an initial test with the circuit boards mounted in the box. It worked, but I found the traces to be noisy and distorted towards the higher voltage ends of the traces. I found that there were two issues - yes, caused by the more complex wiring (of course!). The first was noise being picked up by the lead running from the 'Base' socket to the circuit board. This was mitigated by screening this wire. The distortion was caused by the lead dress of the 'Emitter' and 'Collector' leads - re-dressing these closer to the circuit board mitigated the distortion. Sounds simple, but troubleshooting these effects took a couple of hours.
That done, I completed the wiring of the 'Octopus' curve tracer circuit. I had mounted the six preset pots on a tag strip, and I wired this up fist before installing it into the box. That done and with the wiring completed, I tested this circuit - it did not work(!). Oh boy, what had I done... ?
Well, to cut a long story short, because I had both the bipolar curve tracer and the Octopus wired to the same transformer secondary (the Octopus wired across one half of the centre-tapped secondary), this was effectively shorting out one of the Octopus test leads. At first I though switching the ground connection off to the bipolar curve tracer would fix this, but when I though some more, I figured this could result in damage to components on the bipolar curve tracer circuit board. I then though that perhaps switching the +/-15vDC supply from that board would resolve the problem, but, again, no, as it was the ground that was causing the issue.
This meant that all connections from the power supply would need to be switched to power only either the bipolar curve tracer or the Octopus. This would require a five pole, two-way switch, ie. two poles to switch the 'X' and 'Y' BNC 'scope output connections (already done by a two-pole, two-way toggle switch), and three to switch the +/-15vDC power and the ground connections.
There was insufficient room on the front of the box for any of the rotary switches I had in my junk box, and the maximum number of poles on a miniature toggle switch I had in stock was four. I considered wiring in a multi-pole relay to do the job and replacing the toggle switch with a three-pole, two-way part (the additional pole to switch the relay), but in the end I went for the slightly less elegant/complex solution of using a second toggle switch (three-pole, two-way) to switch the power supply, and leave the original two-pole, two-way switch in place, switching the 'scope outputs. This means that the two switches must be thrown in the same direction to switch between the bipolar curve tracer and the Octopus - not a big deal. Once that was done, the two circuits operated properly.
Next, I set-up the six voltage/current pre-set pots in the Octopus circuit: three set the output voltage to the probes, here 1vAC, 5vAC and 10vAC, and three set the current limit, all set for 1mA. The article I mentioned in a previous post describes in detail how this is done.
I would note that I think there are a couple of errors on the schematic/parts list in the article: I found that I could not adjust the current to 1mA on the 1vAC setting - the limiting resistor (pre-set pot) is specified as 100ohm for this, whereas a resistance of around 1Kohm is needed for 1mA with a 1vAC output(!). I replaced the 100ohm pre-set with a 10Kohm pre-set and was then able to easily set the current to 1mA. Also, the series resistor for the LED power-on indicator is specified as 1Mohm(!) - even with a sensitive blue LED this would result in an extremely dim LD (or it would not illuminated at all!). After some experimentation, I used a 3mm blue LED fitted into a chrome bezel, and a 1.2Kohm series resistor, fed from the transformer secondary via the 0.47uF capacitor as specified. This arrangement worked well.
With that, I checked operation of the Octopus on a diode, a Zener, a capacitor and a resistor - all seemed to be giving the correct response on the 'scope. I will provide some examples in my next post.
There is a very good video by 'X-Ray TonyB' that describes how the Thaikits bipolar curve tracer works and how he built his, plus some mods he did )and why). I adopted his higher 1ohm resistor recommendation and his suggestion of allowing switching between two transistors to compare curves for matching. I may eventually adopt the x10 amplifier mod as well, but not at the moment... The video can he found here: https://youtu.be/lZUSjWBBq4Y (there are several, others as well worth watching).

06/03/2026

I don't know what happened, but the post I made earlier today has been wiped from Facebook! - text, comments, photos, captions and all! Luckily I had retained a copy of the text, but I had to regenerate all the photo captions - what a pain!
Here is the missing post....
--------------------------------
Busy weekend, but I managed to find some time to wire-up the curve tracer, test it, debug it, and complete the labelling.
As I started the wiring, I decided to make a mod to the transistor socket fixture - I had made the original one from a small IC socket, but found the spacing between the two rows of sockets a little cramped. Also, I decided to add a switch to allow switching between the two rows of sockets, thus allowing rapid comparison of curves between transistors mounted in the sockets for matching purposes. That complicated the wiring (of course) and more drilling of the box. With the bipolar curve tracer wired-up, I gave it an initial test with the circuit boards mounted in the box. It worked, but I found the traces to be noisy and distorted towards the higher voltage ends of the traces. I found that there were two issues - yes, caused by the more complex wiring (of course!). The first was noise being picked up by the lead running from the 'Base' socket to the circuit board. This was mitigated by screening this wire. The distortion was caused by the lead dress of the 'Emitter' and 'Collector' leads - re-dressing these closer to the circuit board mitigated the distortion. Sounds simple, but troubleshooting these effects took a couple of hours.
That done, I completed the wiring of the 'Octopus' curve tracer circuit. I had mounted the six preset pots on a tag strip, and I wired this up fist before installing it into the box. That done and with the wiring completed, I tested this circuit - it did not work(!). Oh boy, what had I done... ?
Well, to cut a long story short, because I had both the bipolar curve tracer and the Octopus wired to the same transformer secondary (the Octopus wired across one half of the centre-tapped secondary), this was effectively shorting out one of the Octopus test leads. At first I though switching the ground connection off to the bipolar curve tracer would fix this, but when I though some more, I figured this could result in damage to components on the bipolar curve tracer circuit board. I then though that perhaps switching the +/-15vDC supply from that board would resolve the problem, but, again, no, as it was the ground that was causing the issue.
This meant that all connections from the power supply would need to be switched to power only either the bipolar curve tracer or the Octopus. This would require a five pole, two-way switch, ie. two poles to switch the 'X' and 'Y' BNC 'scope output connections (already done by a two-pole, two-way toggle switch), and three to switch the +/-15vDC power and the ground connections.
There was insufficient room on the front of the box for any of the rotary switches I had in my junk box, and the maximum number of poles on a miniature toggle switch I had in stock was four. I considered wiring in a multi-pole relay to do the job and replacing the toggle switch with a three-pole, two-way part (the additional pole to switch the relay), but in the end I went for the slightly less elegant/complex solution of using a second toggle switch (three-pole, two-way) to switch the power supply, and leave the original two-pole, two-way switch in place, switching the 'scope outputs. This means that the two switches must be thrown in the same direction to switch between the bipolar curve tracer and the Octopus - not a big deal. Once that was done, the two circuits operated properly.
Next, I set-up the six voltage/current pre-set pots in the Octopus circuit: three set the output voltage to the probes, here 1vAC, 5vAC and 10vAC, and three set the current limit, all set for 1mA. The article I mentioned in a previous post describes in detail how this is done.
I would note that I think there are a couple of errors on the schematic/parts list in the article https://www.arrl.org/files/file/QEX_Next_Issue/May-June2017/Ferreora.pdf : I found that I could not adjust the current to 1mA on the 1vAC setting - the limiting resistor (pre-set pot) is specified as 100ohm for this, whereas a resistance of around 1Kohm is needed for 1mA with a 1vAC output(!). I replaced the 100ohm pre-set with a 10Kohm pre-set and was then able to easily set the current to 1mA. Also, the series resistor for the LED power-on indicator is specified as 1Mohm(!) - even with a sensitive blue LED this would result in an extremely dim LD (or it would not illuminated at all!). After some experimentation, I used a 3mm blue LED fitted into a chrome bezel, and a 1.2Kohm series resistor, fed from the transformer secondary via the 0.47uF capacitor as specified. This arrangement worked well.
With that, I checked operation of the Octopus on a diode, a Zener, a capacitor and a resistor - all seemed to be giving the correct response on the 'scope. I will provide some examples in my next post.
There is a very good video by 'X-Ray TonyB' that describes how the Thaikits bipolar curve tracer works and how he built his, plus some mods he did )and why). I adopted his higher 1ohm resistor recommendation and his suggestion of allowing switching between two transistors to compare curves for matching. I may eventually adopt the x10 amplifier mod as well, but not at the moment... The video can he found here: https://youtu.be/lZUSjWBBq4Y (there are several, others as well worth watching).

06/02/2026

Busy weekend, but I managed to find some time to wire-up the curve tracer, test it, debug it, and complete the labelling.
As I started the wiring, I decided to make a mod to the transistor socket fixture - I had made the original one from a small IC socket, but found the spacing between the two rows of sockets a little cramped. Also, I decided to add a switch to allow switching between the two rows of sockets (see info in the last paragraph below), thus allowing rapid comparison of curves between transistors mounted in the sockets for matching purposes.
Doing this complicated the wiring (of course) and entailed more drilling of the box.
With the bipolar curve tracer wired-up, I gave it an initial test with the circuit boards mounted in the box. It worked, but I found the traces to be noisy and distorted towards the higher voltage ends of the traces. I found that there were two issues - yes, caused by the more complex wiring (of course!). The first was noise being picked up by the lead running from the 'Base' socket to the circuit board. This was mitigated by screening this wire. The distortion was caused by the lead dress of the 'Emitter' and 'Collector' leads - re-dressing these closer to the circuit board mitigated the distortion. Sounds simple, but troubleshooting these effects and implementing the solutions took a couple of hours.
That done, I completed the wiring of the 'Octopus' curve tracer circuit. I had mounted the six preset pots on a tag strip, and I wired this up fist before installing it into the box. That done and with the wiring completed, I tested this circuit, and - it did not work(!). Oh boy, what had I done... ?
Well, to cut a long story short, because I had both the bipolar curve tracer and the Octopus wired to the same transformer secondary (the Octopus wired across one half of the centre-tapped secondary), this was effectively shorting out one of the Octopus test leads. At first I though switching the ground connection off to the bipolar curve tracer would fix this, but when I though some more, I figured this could result in damage to components on the bipolar curve tracer circuit board. I then though that perhaps switching the +/-15vDC supply from that board would resolve the problem, but, again, no, as it was the ground that was causing the issue(!).
This meant that all connections from the power supply would need to be switched to power only either the bipolar curve tracer or the Octopus. This would require a five pole, two-way switch, ie. two poles to switch the 'X' and 'Y' BNC 'scope output connections (already done by a two-pole, two-way toggle switch), and three to switch the +/-15vDC power and the ground connections.
There was insufficient room on the front of the box for any of the rotary switches I had in my junk box, and the maximum number of poles on a miniature toggle switch I had in stock was four. I considered wiring in a multi-pole relay to do the job and replacing the toggle switch with a three-pole, two-way part (the additional pole to switch the relay), but in the end I went for the slightly less elegant/complex solution of using a second toggle switch (three-pole, two-way) to switch the power supply, and leave the original two-pole, two-way switch in place, switching the 'scope outputs. This means that the two switches must be thrown in the same direction to switch between the bipolar curve tracer and the Octopus - not a big deal. Once that was done, the two circuits operated properly.
Next, I set-up the six voltage/current pre-set pots in the Octopus circuit: three set the output voltage to the probes, here 1vAC, 5vAC and 10vAC, and three set the current limit, all set for 1mA. The article I mentioned in a previous post describes in detail how this is done.
I would note that I think there are a couple of errors on the schematic/parts list in the 'improved Octopus' article https://www.arrl.org/files/file/QEX_Next_Issue/May-June2017/Ferreora.pdf ): I found that I could not adjust the current to 1mA on the 1vAC setting - the limiting resistor (pre-set pot) is specified as 100ohm for this, whereas a resistance of around 1Kohm is needed for 1mA with a 1vAC output(!). I replaced the 100ohm pre-set with a 10Kohm pre-set and was then able to easily set the current to 1mA. Also, the series resistor for the LED power-on indicator is specified as 1Mohm(!) - even with a sensitive blue LED this would result in an extremely dim LED (or it would not illuminated at all!).
After some experimentation, I used a 3mm blue LED fitted into a chrome bezel, and a 1.2Kohm series resistor, fed from the transformer secondary via the 0.47uF capacitor as specified. This arrangement worked well.
With that, I checked operation of the Octopus on a couple of diodes, a Zener, a capacitor and a resistor - all seemed to be giving the correct response on the 'scope. I will provide some more examples in my next post.
There is a very good video by 'X-Ray TonyB' that describes how the Thaikits bipolar curve tracer works and how he built his, plus some mods he did )and why). I adopted his higher 1ohm resistor recommendation and his suggestion of allowing switching between two transistors to compare curves for matching. I may eventually adopt the x10 amplifier mod as well, but not at the moment... The video can he found here: https://youtu.be/lZUSjWBBq4Y (there are several, others as well worth watching).

05/29/2026

I decide to use an 'improved' (more versatile) circuit for the 'Octopus' part of the curve tracer (article by Paulo Ferreira, PY3PR in the May/June 2017 issue of QEX - downloadable from the internet).
I spent some time this morning marking-up a paper template for the project box that will be housing the combo curve tracer/Octopus circuits. The project box is a NOS one (from a set of three nested boxes), that are made from Bakelite(!), with an aluminum lid - these must be quite old!
I decided to mount the controls and connectors on one side of the box rather than on the aluminum lid - partly because drilling the Bakelite is a bit easier than the aluminum, but also that this avoids 'flying leads' from the lid to the inside of the box, where the circuit boards, power transformer and a tag-strip are located.
That said, care should be taken when working with Bakelite, as the the type of filler(s) used in the mix ranged from fine sawdust to asbestos powder - the latter obviously posing a health hazard if released as fine dust when drilled/sanded, etc. I did all the drilling in the garage with a fume extractor and wore a half-face respirator.
I ran out of time, so did not manage to start any wiring - job for later or over the weekend...

05/28/2026

I powered-up the bipolar transistor curve tracer this morning and..... yeah! - it works. So far I have only tried a small NPN device in it (2N3904) and measured its hfe, but will be testing PNP and power transistors later.
Of course, my PEAK atlas DCA55 can measure hfe, but seeing the curves allows for better matching of transistors if I need to do that - and, besides, its always more interesting to look at curves than numbers(!).
I would really like a 'proper' curve tracer to play with (Tektronix made several models, and there is a Chinese one that is still being made), but these fetch a premium price these days, especially for ones that are quoted as fully working (even the Chinese one costs around $1500, and one video on YouTube shows its innards - quite shoddily built! - worked though). For the use I would have for one, I cannot justify paying so much, but will keep an eye out for a 'bargain' or 'fixer-upper' one...
Next job is to build it into the project box and add the 'Octopus' circuitry.

05/28/2026

So, what is the next vintage radio project? - its the Italian-made ham bands only Geloso G209 that has been sitting on a bench in the garage for a couple of months. So far, it has only made it as far as the side table next to the bench as I am still working on the curve tracer project, but hey, its one step closer to being checked-out on the bench. I recall from my initial 'once-over' in the garage that it looks to be in good shape overall, so maybe not much work will be required? (famous last words before a nightmare!).
Also, the project after the Geloso arrived on my doorstep yesterday (literally!), in a huge box, weighing in at around 54lbs. I watched the delivery man struggling and cussing on the porch camera (I waited until he was gone before I ventured outside to retrieve it with a dolly...).

05/28/2026

I have had some correspondence with folks about my narrowband FM/SSB 10GHz kit. One question was asking about the 'snap' (step recovery diode) multiplier design and which diode was used. I think I have now found the article from which I based my design on - attached to this post.
I also found the design of the harmonic filter I built for the 384MHz signal source (also attached to this post).

05/28/2026

My next vintage radio project has now moved from the 'waiting room' to the 'shop. However, while I was moving stuff around in the 'shop, I found a kit I had bought a short while ago that said "build me!", so with the bench being clear, I spent a couple of hours putting it together last night, and then testing part of it out (the power supply). The power supply works ok, but I ran out of time before I could couple-up the main circuit board.
The kit is a simple bipolar transistor curve tracer that is used with a 'scope ('X' - 'Y' inputs), so nothing 'special', but something I thought would be interesting to play with. I intend to build it into a project box with some simple additional circuitry and switching to allow dynamic testing of some other component types, eg. diodes, Zeners, capacitors and resistors), ie. the ubiquitous 'Octopus' circuit.

Radio Restoration

06/03/2026

06/03/2026

06/03/2026

06/03/2026

06/02/2026

05/29/2026

05/28/2026

05/28/2026

05/28/2026

05/28/2026

Address

Website

Alerts

Contact The Business

Shortcuts

Share