2010-05-12
More Sports Ears Related Musings
A Cheaper Compatible Receiver
Checking eBay the other day I found a 3-band consumer radio that covers the Sports Ears frequency for $5.48 AUD. Naturally enough I ordered it to check out. The shipping was about $5 on top, so < $11 landed. Apparently you can also order it direct from the Suntek store and get free shipping. The unit is "Kaide" brand, model designator "BC-R30". It is a little physically larger than the AFL Sports Ears product, with 2 VHF FM bands and MW. Turn-the-dial tuning, covering 64-86 MHz, 86-108 MHz and 530-1600 kHz.
It operates from two AAA batteries, pulls about 17 mA quiescent on all bands and features a red LED power indicator. It has a small telescopic antenna, a front-panel speaker and a 3.5 mm stereo headphone jack. The FM detector is only monophonic, but like the Sports Ears unit (and unlike so many cheap radios) the actual socket is wired to support stereo headphones not the mono magnetic ear piece of years gone by. Unlike the AFL Sports Ears unit its dial indicator is easy to read, and even at full (ear-splitting) volume the audio amplifier has very little distortion. The unit does not come with earphones, one must provide their own if they desire private listening.
The front panel speaker offers flexibility that none of the Sports Ears range have - perhaps by design. One could say so as not to annoy near-by fans in the stadium, or perhaps because of its limited practicality in the noisy stadium environment. Less charitably, to minimise costs and encourage more sales.
Inside, the BC-R30 is based on the CD9088CB (a TDA7088 clone). This particular chip is vaguely similar to the old TDA7000 low-IF (70 kHz) with FLL detector architecture and is aimed at the very cheapest end of the FM broadcast receiver market. It provides acceptable quality FM detection and features scan/reset varactor or turn-the-dial polyvaricon tuning with or without AFC. Being low-IF with FLL recovery it has rather weaker selectivity than the ceramic filtered quadrature detector system used in the genuine Sports Ears unit. As supplied I found the mute circuit annoying, so I changed R6 to 10 kΩ to disable the mute. I also personally find the capturing effect of the FLL demodulator annoying, and prefer the smoother tuning of other architecture FM discriminators. That said, the unit is sensitive and more than adequate for the task, even with the mute unmodified.
The MW feature of the receiver is based on the CC7642 AM TRF 3-pin TO-92 chip (a clone of the TA7641 which is a clone of the old ZN414 or MK484). The arrangement is even simpler than that suggested in the TDA7088 datasheet for an AM/FM solution which utilises the CC7642 at a 455 kHz IF using the TDA7088 as an RF amplifier and mixer. Rather the CC7642 works directly at the RF frequency as a TRF receiver. The drawback is absolutely pathetic selectivity, especially with the fairly poor Q of the loop-stick used. One can hear 2BL 702 kHz across most of the band for example, even under some of the weaker stations towards the top of the band.
Indeed, this receiver is built for a price and it shows it. There is little else you could remove and still have a functional receiver. The genuine Sports Ears unit has a technically much more advanced architecture, especially on MW where it is a single-conversion superhet and uses a ceramic filter for much better performance. The BC-R30 has no IFT cans or ceramic filters in it at all, in fact only three inductors, the two VHF LO tanks for the two FM bands and the MW loop-stick antenna, there is no front-end track-tuning tank on VHF! This is a "feature" of such minimal AM/FM receivers which minimises the production costs and simplifies the factory alignment requirements.
The lack of a track-tuned front-end means the FM detector will also harmonically detect out-of-band transmissions. For example, on the "SCH" band, I can easily hear TV channel 7 audio (187.75 MHz) near the bottom of the dial "64". this is a 3rd harmonic response: LO = (187.75 - 0.07) / 3 = 62.56 MHz. (I confirmed this by listening to the LO using my VR-500, the LO being locked to the RF signal by the FLL is WBFM modulated by the signal being received.) This harmonic mixing is either a feature or a major limitation depending on what you want to listen to. For the 70.2 MHz Sports Ears channel the harmonic spur and image responses are in sparely populated regions of the spectrum and will likely cause no problems.
Still, none of these limitations of the BC-R30 make the Sports Ears receiver look like a great deal. The Sports Ears receiver costs 8 times more than BC-R30. At least with Sports Ears you are getting a real superhet with fairly good selectivity. It is arguable that fixed "UMPS" tuning is also preferable as you do not need to retune, and can flick between your favourite FM station and UMPS easily. The Sport Ears receiver also pulls a little less current, so its batteries will last somewhat longer. That said, the BC-R30 is quite usable if you don't want to buy the genuine product.
The big feature of the BC-R30 is the TDA2822 audio amplifier. Even at only a 3 volt supply it can deliver several hundred milliwatts of audio at very low distortion. That combined with the front-panel speaker and better treble response makes the BC-R30 a pretty nice radio for casual use, especially as it covers part of the VHF TV band, and for the price is fairly competitive to the Digitor/Tecsun receiver available from Dick Smith I suggested as another alternative.
Converting a Common AM/FM Radio
For comparison purposes, I picked up a $12 AUD "Sansai" AM/FM radio at Hot Dollar in Cronulla. It is a fairly bulky unit, has a monophonic earphone socket, and a really pathetic dial scale pointer that sometimes sticks. It tunes the usual 88-108 MHz and 530-1600 kHz bands. It has a belt-clip in addition to the lanyard shared by the other receivers and runs from two AA batteries making it a bit cheaper to feed although it pulls about 20 mA quiescent despite having no power indicator LED. There are no screws holding the shell together, just rather fragile moulded clips.
Inside that extra $5 shows, the radio is arguably better engineered than the BC-R30. It is based on the SA2003 (a TA8164 clone). This chip implements a full superhet AM/FM receiver with AM AGC and a quadrature detector for FM. The implementation utilises track-tuned circuits for both bands and ceramic filters/quad network. It just happens to use the TDA2822 audio amplifier. Why it only has a mono earphone socket despite a dual-channel amplifier is a bit of a puzzle.
The MW selectivity is provided by the 455 kHz ceramic filter and is fairly similar to the Sports Ears receiver. Needless to say, a vast improvement over the BC-R30's MW receiver.
For VHF WBFM the ceramic quadrature network based detector is actually a little touchy to tune compared to the LC quadrature network detector in the Sports Ears receiver. I am unsure if the ceramic IF filter pass-band does not exactly correspond to the ceramic quad network or if the filter itself is a little too narrow, but distortion on deviation peaks is noticeable unless tuned exactly on frequency with a strong signal. Despite the locking effect of the FLL in the BC-R30 I think I prefer its detector for overall quality of recovered modulation with less than precise tuning.
Anyway, conversion was a simple affair. Although I had hoped there would be enough adjustment range in the existing coils, this was not the case, I was forced to replace the oscillator and front-end tanks. I replaced the LO coil with a 5 turn one (5 turns on 5 mm ID, 700 um tin-plated Cu wire) and replaced the pair of bandset caps (an 20 pF and 30 pF) with a single 56 pF capacitor. For the front-end tuned circuit I replaced the coil with a similar 5-turn coil and needed about 10 pF extra capacitance across the existing gang + bandset cap to place 70.2 MHz mid-dial. The turn spacing was tweaked with the tracking trimmers at mid-swing. The resulting receiver works just fine, but doesn't quite track-tune properly. This is of no consequence for use with just Sports Ears, and I did not believe it was worth the effort to unsolder and measure the polyvaricon properties to compute a L & C solution for near-perfect tracking.
When I said "trivial" affair in my previous article, I may be exagerating a little. Without test equipment, especially a signal generator and frequency counter this task would have been pretty daunting, especially if you want to preserve proper track-tuning or add band-switching. I found my wavemeter and tone dipper especially helpful for this realignment, more so when I got completely lost and couldn't sniff enough LO energy to drive the counter properly.
A Signal Generator For Testing
In order to locally test receivers and converters for the Sports Ears frequency I built a simple test source. It is carefully shielded with a fairly weak output power and internal PSU so the signal can be attenuated to very feeble levels for sensitivity testing. When terminated in a dummy load the signal is undetectable near unit. Despite no special precautions being taken with the AF connector line to prevent RF leakage, very little escapes there. Originally an internal AF oscillator was planned, but was found to be unnecessary and the socket added to allow a range of audio sources to be used for testing.
The circuit delivers about -13 dBm into 50 Ohms and has sufficient stability for the purpose. It tunes about 52-90 MHz with the values indicated. The LED based transistor base bias is to allow operation from a wide range of supply voltages (near constant emitter current). This was more useful during testing than in the final circuit with its 3 volt battery supply, you may fix the base bias with resistors if you prefer.
The oscillator is very conventional, grounded base Colpitts topology. I used a T37-6 core for the tank coil and it seems to offer sufficient stability, even without embedding it in wax. A trimmer allows tuning to the frequency of interest (in this case 70.2 MHz), and it will only drift 50 kHz or so over an hour once set and drifts much less once completely warmed up. The base bias is modulated by the audio signal to provide quite reasonable frequency modulation, with only a small amount of parasitic AM. Deviation achievable far exceeds that required for FM broadcast band standards (75 kHz).
Incidentally I also built a quick hack that delivers several hundred milliwatts which I'll only provide a photo of. Not so much because I am concerned about what people may do with such a device, but rather than the design is embarrassingly bad!
As you can see, it has no output low-pass filter and matching. It is rather inefficient, pulling about 1 Watt DC. The multiplier stage is very coarsely tuned and the output is not very pure, but it works for higher power testing with external clean-up gear.
9 comments.
Attachments
| title | type | size |
|---|---|---|
| VHF WBFM Signal Generator Circuit Diagram Source | application/postscript | 15.001 kbytes |