This is another example of a cheap FM radio
kit using a super regenerative tuner. The MX801F is one of the Maxitronix
series, and is similar to their MX901F and MX901AF. The notes in this article
are also of relevance to these kits. They were briefly mentioned
in my review of the Science Fair 28-234 AM/FM radio kit.
I had actually ordered the MX901AF in view of it including the AM tuner, but ended up with the MX801F instead. Not to worry, as I was primarily interested in how the super regen tuner in this one worked and whether it was any better than the Science Fair design. I ordered from Terrific Scientific as it was local and wouldn't have to pay so much postage.
What surprised me is how small the chassis was; 12x8cm. Like all these kits containing an FM tuner, the only construction is related to connecting together the audio amplifier components.
Very cute little set once assembled.
I had it assembled in about an hour. It's
the usual spring connector construction on a cardboard chassis. This set
has a plastic surround to make it a bit more rugged. The 9V battery does
not have a holder on the chassis, but merely dangles on the end of its
The instructions were just as uninformative as the Science Fair kit and contained some mistakes.
No technical description at all is given. Following the instructions for assembly, several times I had to remove the tuner from the chassis to proceed as the PCB covers the springs used for the 9V battery connection. Obviously, no one checked the order in which the construction steps were presented. Being of the present day, the earphone is actually a ceramic type. It seems the crystal types originally used in these kits are no longer. While it looks identical from the outside, it's when you look down inside and see the flat diaphragm that you can tell. While this type of transducer is of very high impedance like its crystal counterpart, it does present a much higher capacitance. I measured .035uF. This isn't a problem here as the earphone is driven by a low impedance source, but for something like a crystal radio I would expect problems. What was quite bizzare is the earphone lead is solid core wire! While it is flexible enough, I wonder how often one can flex it before breakage occurs.
Incidentally, I did test the completed set with a crystal earphone and found very little difference.
Low audio output.
Upon powering up I was pleased to see that the sensitivity was what it should be and the sound quality was far better than the Science Fair kit.
Alas, the audio level was very low. It was similar to a weak crystal set. While you could recognise music and speech, it wasn't really of entertainment value, and hopless in a noisy enviroment. Observing the waveform at terminal 8 showed a considerable amount of quench signal but minimal audio. That is a sign of excessive oscillation in the detector.
So, it looked like this one would also require modification.
Circuit of the MX801F. Note the mistake with the 1M bias resistor shown between terminal 4&6. It should connect from terminal 5 to 6. Shown as above, the transistor receives no bias. The circuit on the actual kit is correct. Terminal 3&7 are linked, as are 4&1 and 10&2.
During construction I noted that "MX901AF"
was printed on the tuner PCB, so obviously they use the same tuner in all
kits. However, there was one discrepancy. Instead of the 47K load resistor
shown above (between terminals 1&2), what was on the PCB, and what
is shown in the circuit for the MX901AF, is different. In the other circuit,
the load consists of a 10K and 22K in series with the junction of the two
bypassed to earth via a condenser. This condenser is shown below
as a .05uF, but on my PCB it was .1uF. Yet another difference.
The interesting question is why the change, and which of the two circuits is meant to be the better. Obviously, two resistors and one condenser are more expensive than one resistor, so I'm inclined to think the MX901AF is the correct one. Assuming the MX801F was developed before the MX901F this would also support this theory.
Circuit for the MX901F. Note the 22K and 10K. Interesting to note slight changes to the audio amplifier.
The tuner uses the usual self quenching Colpitts oscillator type of super regenerative circuit. Like other kits, no means is available to adjust the operating conditions, so at best the performance is a compromise.
I went down the same path as with the Science Fair kit to try and improve it. I tried a few different modifications and eventually ended up replacing the 22K and 10K with the 47K as per the original design. This improved things slightly. It was obvious the bias to the detector transistor was way too high, so reduced it by connecting a 47K from base to earth (i.e. across the .01uF). This brought forth a much louder signal but with a very audible quench frequency. The .02uF was changed to .0027uF to fix that one.
To avoid any confusion, this is the circuit as modified.
Unlike the Sceince Fair kit, this one shows the component values of the VHF tuner on the chassis. Note the 47K resistor just to the left of terminal 2.
The performance now was pretty good for
what it was, considering there is no user adjustable regeneration. In fact
the sensitivity is very good with only the 18cm long aerial. Quite a number
of Sydney's low power public radio stations are receivable at my location
in the mid Blue Mountains. However, audio quality appears to be somewhat
dependent on signal strength, with weaker signals being being less distorted
and easier to tune. I did find setting the detector bias level critical,
and the 47K base to earth resistor may need to be a different value with
As I've explained elsewhere, a super regenerative receiver provides highest output and most sensitivity just past the point of oscillation. But, while weak stations are received optimally this way, high power signals cause overload. This is when the detector needs to be adjusted further into the point of oscillation to restore optimum sound quality. Also, the ideal setting will vary from one end of the band to the other.
As I've said I've had to compromise here, and it's better to favour the weak signals.
One other thing is evident, and that's hand capacitance. As the VHF tuning coil is right under the cardboard just above where the knob is, there is a slight variation in tuning when removing your fingers from the knob after tuning in a station. Touching the 9V battery also has a similar effect.
A bit of shielding or a different layout would have been worthwhile here.
Under the chassis. Such an elementary circuit has incredible performance for what it is.
This set is a considerable improvement on the Science Fair model, even without modification. The audio is of quite good quality.
However, to make it a practical receiver to use all the time, it is necessary to modify the detector bias and quench frequency.
Once again, it would have been useful and informative for the intended constructors to explain the operation of the circuit in a simple way at least. The fact that the FM tuner is preassembled doesn't help one in learning anything about it. It appears that the kit designers are incapable of getting super regenerative receiver design correct, so they would be better off providing the tuner board with a TDA7000 type of IC. Given how cheap the Chinese clones of this IC are now it wouldn't increase the cost signifigantly. At least then good results would be guaranteed and the constructor wouldn't be left with an unuseable receiver. While I have the experience to correct the bad design, the typical "For ages 8 and up", or non technical constructor would be at a loss.
Incidentally, by adjusting the relevant trimmer on the back of the variable condenser, the receiver tunes up into the aircraft band with very good results. So, if you are looking for a cheap and simple aircraft band receiver, this kit is definitely worth using.
email me: cablehack at yahoo dot com