Matching FET’s


So the next thing on the building agenda was yet another piece of test gear. I want to start by building the Minima KISS mixer. The focus of so much attention recently as some builders were having trouble with it. And because the LO to RF port isolation was being reported as not being high enough. Having read as much as I can understand about the topic (very little but I’m learning). I’ve become convinced that success with a mixer made from two FET devices probably entails careful “matching” of the FET’s. After all, we match the diodes we use in the simple BFO mixer. So why not the FET’s?

From “Mixer Musings and the KISS Mixer by Chris Trask / N7ZWY:”

“…but that the balance of such a mixer making use of discrete transistors will be poorer than the balance of a diode mixer because of the difficulty of matching the rather complex transistor parameters over the operating range (10).” 10. HF Radio Systems & Circuits, 2nd ed. Noble Publishing Co., Atlanta, Georgie, 1998

Where the transistors mentioned above are VHF FET’s. So the problem would seem well known.

Having decided that matching FET’s is probably important, then how does one go about doing it? A quick scout around the Internet, once again led me back to the “QRP & SWL Homebuilder” website and a section called “FET Matching” (scroll down).

If by some miracle or bizarre twist of fate you are reading this and have have not already discovered “The Popcorn website”, then do yourself a favour and head on over. You’re in for a real treat.

The photo left shows my bridge variant. Setup with shorted links in the sockets where the FET’s are normally placed. This is so you can adjust the bridge balance for exactly zero volts.


So the afternoon finished with using the bridge to match some FET’s.

I was going to match a whole bunch of J310’s. Until I discovered that I didn’t have any J310’s that is! I could have sworn black and blue that I had a whole heap. Apparently not.

So I went and matched a whole bunch of Audio FET’s just for the fun of it and by way of practice. Yep, FET’s even taken from the same cardboard parts strip are all over the place!! Best match from matching 9 samples to one control was 0.010 volts. The photo opposite shows 0.313 volts but most pairings were much higher than this.

Now it should be noted that the general consensus is that for best results in a KISS mixer we should be matching the units for highest IDSS. The bridge I’ve built tells me that both FET’s are the ‘same’ but doesn’t say much about IDSS. So that’s the next experiment. Find one J310 FET with a high IDSS and then find him a partner.

Next up.. More FET matching/selection and building a KISS.

Power Supply Completed.


Well this is how the completed power supply turned out. For those wondering at the over-engineering. Yes, this power supply while very simple is completely over the top for a little Minima. The beauty of the Minima is that it serves as a platform for experimentation.

So I’ve build the supply with both + and -12 volt DC rails. Just in case I might want to play with split power supply audio amplifiers or similar. There is a +9v regulator that feeds from the +12 volt supply. Which in turn feeds to the +5 volt regulator on the logic board. So now the +5 regulator does not get so hot.


The supply can also provide ~40 volts DC. Which may come in handy for getting higher power from a IRF510 linear. Perhaps…



To the left is the Arduino Logic PCB mounted above the completed power supply. With the master power LED on the bleeder resistor winking out at us. “Yes I’m working and don’t poke your fingers in here!”.

To the right is the top view of the logic PCB mounted atop the power supply and the entire module bolted back onto the radio chassis.



Finally another shot of the radio powered up. This time the photo is not as contrived as previous. She is running on her own internal power supply source.

It is interesting to note that this photo was taken with the “Raduino 0.4” sign-on message displayed on the LCD screen permanently. No trick photography nor hung I2C data bus. It just happens from time to time when everything is very stable and the shack has warmed up to a nice cosy constant temperature. As soon as you move the tuning knob it starts displaying the frequency. A quick glance at the source code and I’m convinced that this is normal behaviour. The current Raduino sketch does not display the frequency until there has been some sort of update.

Mostly though when people see the “Raduino 0.4” message forever it means that the I2C data bus is in a hardware fault condition which seems to hang the Arduino Wires Library.

Work continues on the power supply.


Well another weekend of building passes. With such little time as could be devoted to the task between “life” interruptions. You will note that so eager to return to construction was I, that the work bench didn’t really get the tidy-up it urgently needed. Ah well, next time.

In the end I decided just to go ahead and complete the power supply before moving onto the radio proper. Which of course is taking much longer than expected. Here are a couple of photos of the “in progress” build of the power supply. Not yet complete but a good start. The logic board has been removed while the power supply takes shape.


I also went ahead and re-wired the ribbon cable to the logic board. So that it is now long enough to reach the new board position above the power supply. I’ve been making up ribbon cable with IDC connectors for years. And they never give me any real trouble. Until this morning that is. First attempt and no LCD display. Cable was found to have many intermittent wire faults. Any attempt to “repair” the faulty cable just seemed to make matters worse. I can normally save an IDC connector for re-use but not this time. I had to scrap the connector and start over. Second time worked a treat thank heavens. Still it wasted a good hour or two.


This last image is the front of the radio again. This time with the LCD powered up and displaying information under the control of the logic board. However the photo is very much contrived. As the logic board is simply dropped behind the front panel on a piece of insulating paper. Just to see what it would look like.

Next time. Hopefully the completion of the power supply. And maybe the start of the KISS mixer and some FET matching.