Another Cheap Superhet Kit

[dave_n_t]

There have been a few threads about this sort of kit in the past, so please forgive another one....


I bought this before Christmas, hoping that I'd have some free time to play with it over the holiday period. It cost £3.85, including delivery from China. (I note that the prices seem to have risen somewhat steeply, recently). I finally got around to playing with it last week.



I'll try to write up my experience in some detail, in case anyone else buys one of these. Bottom line: it works very well, and for the price gave me a lot of fun.

The first thing I decided to do (having remembered claims that 'all the components are rejects') was to test, as far as practicable, everything in sight. Here's a picture of the components:





The resistors were all spot on (in fact, within 1% of their nominal values). The ceramic capacitors were OK, but those marked 223 (i.e. 22nF) were uniformly more like 30nF (and no, it wasn't a spurious reading because of leakage). That gave me no worries - they're mostly decouplers. The 100uF electrolytics (all 3) measured 99uF, with acceptable ESRs. The first 4u7 electrolytic was fine; the second was identified by my cheap all-singing component tester as a diode with a 5V forward voltage drop. I suspected the electrolyte hadn't properly reacted with the plates, so I reformed it on 12V for a couple of hours, after which it was fine.

The transistors all had gains of between 200 and 220, indicating close selection at some point (perhaps at manufacturing).

The driver transformer has two separate secondaries of about 50mH each, and a primary of about 0.6H - all hunky-dory. The only thing of note was that the pin-out was a little strange, with one of the secondary windings coming out at a corner and the opposite middle pin.

The ferrite aerial coil was OK, with a main winding of about 0.7mH.
The oscillator coil had the right continuities, and the IFT tuned windings resonated at about 520kHz. (With hindsight, this should have warned me - most mini IFTs like this are tuned to 455-470kHz).

Finally, the little LED was tested (mainly to find out its polarity), and the speaker spoke nicely. In doing this write-up it now occurs to me that I didn't test the polyvaricon; but it turned out fine, in any case.

After all that testing, the result was one dodgy capacitor (which may have reformed in situ in any case) and some caps possibly wrongly marked. Hardly 'reject components'.


(to be continued...)

[AdrianH]

For £3.85 you could not buy the components here, few extra pf across the cans?

Adrian

[dave_n_t]

(part II)

The instructions to the HX207 came as a 2-sided A4 piece of paper; dense Chinese characters on one side, together with a list of components, with European markings; on the other side a circuit diagram, a PCB layout with component overlay, diagrams of transistor lead outs, ferrite aerial leadouts, and resistor and capacitor value marking system, including (usefully) the resistor colour code with Chinese characters representing the colours (I don't need telling that 2 is red, but it turned out to be useful seeing the symbol for 'red').

Assembly was pretty straightforward: headphone socket, resistors, capacitors, tuning cap & switched volume control, semiconductors, assorted inductors, and final wiring up (4 wires: two to the battery compartment, two to the speaker; all wires provided and ready stripped/tinned).

A few notes:

- the volume control had to be mounted as far in towards the centre of the PCB in order not to foul the case.

- the LED was mounted on the print side of the PCB so that it shone out of the hole in the tuning scale decal.

-the ferrite aerial mounting bracket is fitted using the same two holes as fix the tuning cap. This means that the supplied screws are not (as I first suspected) too long. [Fearing fouling the tuning cap vanes, I shortened the bolts before I realised about the aerial mounting bracket. Ho hum. At least they were still long enough to bite!]

- the transistor lead out is opposite to what I was expecting, but the printed sheet, and the component overlay were both right.

- as I said earlier, the driver transformer has an odd lead out. But the overlay again is right

- the PCB had 5 test points for monitoring current at various key points. They're arranged as circular pads with a diameter etched away, to break the track. Full of optimism, I carefully bridged each of them with solder; if things needed to be tested later I could easily unbridge them...

- the circuit diagram helpfully identifies the oscillator can (the one without embedded capacitor), and the 3 IFTs each with a single Chinese character. These characters turned out to be the characters for red, yellow, white, and green (as identified in the resistor colour code). I'm not sure how critical the IFT ordering was, but at least I could do it the way they intended: they were the colours of the ferrite cores.

On initial switch-on I was greeted with a little background white noise, and, after tuning, Radio 5 coming in quite well, but at around 1400kHz (way off the mark). Time for some alignment...


(to be continued)

[dave_n_t]

(Part III)

The fact that Radio 5 was coming in loudly, albeit at the wrong part of the dial, suggested that the IF alignment wasn't too bad (as indeed it had appeared from initial testing of the IFTs). So I tried to adjust the (red) oscillator ferrite slug (more of a 'hat' on these particular coils) in order to get the Radio 5 reception at the right point of the dial. It rotated, but clearly didn't engage any thread, because it stayed at the same height. I unsoldered the can and checked inside. The threads on the ferrite, and on the plastic insert, were both either worn or too shallow. My bodge of the day was to crush the can slightly so that there was a bit more 'bite' on the threads (you can see the slightly deformed can in the photo at the end). Resoldering allowed me to bring the oscillator core to more-or-less the right place (right at the top end of its travel).

Now the oscillator at the low frequency end was OK. To check the position of the coil on the ferrite aerial I (as normal) used a small piece of ferrite rod brought up to the set's own rod to see it more inductance was needed; and a closed loop (of solder!) brought up to see if less inductance was needed. The coil was, by chance, in more-or-less the optimum spot).



Trimming the oscillator (and aerial) trimmers on the variable cap brought about a reasonable coverage at the HF end, too.

I thought I'd tweak the IFTs a little, but they were very stiff. Out with the watch oil - it worked a treat, and they didn't need much tweaking to get the best volume. I checked using a sig gen and they were still well over 470kHz, but I thought I'd leave well alone. I wonder if there had been some testing of the oscillator and IFTs together which entailed setting them to a very high IF in order to try to counteract the osc. coil's non-tuenable core. In any case, performance was now very good (at the low frequency end, a Spanish voice station beyond Radio 5 comes in well; at the HF end, local Radio Gloucester is also good and strong).

The speaker had to be fixed in place with a couple of dabs of hot glue; and the AA cells kept falling out, so I've put a few self-adhesive pads over them. Maybe a judicious piece of foam rubber would do the job, too.



Photos of the end result:


top of circuit board



front of the radio



All-in-all, a satisfying few hours work (play!)

The general design must be over 60 years old now, and still going strong. A record I wonder.