Saturday, 4 April 2015

PI and Pixie

A week or so ago I ordered, off ebay and from China, a kit for a 40m Pixie transceiver. This was a tad over £3 including postage! I cant source the parts for that, let alone etch a PCB. This kit came with all the parts and a proper through hole plated, conformal coated and silk screened board! Could it be any good for that price?

Well, heres the kit before assembly

The instructions were no more than a components list, circuit diagram and board layout, but hey, what else do you need? All the connectors were supplied, BNC, 3.5mm jacks, and a DC socket for the power. The crystal is on 7.023MHz, which is a bit unusual for a QRP job, but I can soon change that if needed from stock.

I proceeded with assembly starting as you should with the resistors

The next photo shows the board with all the passives and the diodes fitted. All the parts were there, in fact I had several ceramic capacitors extra!

The completed board, the two transistors being the last parts fitted. The only oddities I noticed were that the sockets all fitted on with a very slight angle, clearly a slight error in laying out the artwork for the boards onto the processing machine, but also that all the tracks were on the 'component' side! This could be some technique for minimising instability, although I suspect someone forgot to invert the artwork!

Once populated, I tagged a PP3 clip onto the board, attached the antenna connector to my QRP power meter (rated 2W fsd) and with a spare jack plug shorted the key contacts

Thats around 600mW by my reconing! Not a bad output for a couple of small signal NPNs. With the Pixie connected to my doublet via the ATU, and a crystal earpiece, several CW transmissions can be heard. Theres no side-tone, but I think this will prove to be a fun little rig.

Ive also found time today to do some more work on the Surf PI metal detector. Ive added all the controls, and started testing. The coil I had produced is, it seems, far too low inductance. The unit just wouldnt behave with it. So, I quickly threw together a new coil of 8 inch or so diameter from enameled copper wire, stuck it with masking tape on the end of a fibreglass rod to get it clear of any metal, and connected it up. Its performance was poor, but at least I could detect a coin with it at about 6 inch. On measuring it showed to be too high an inductance at 410uH. A few turns off and I have a 27t, 310uH 2.4ohm coil for testing.

 The buzzer that comes with the kit is ok, but ive swapped it for a small 1.25 inch loudspeaker, which is a little less harsh in tone. The offset control zeroed fine, and the pulse waveform looks roughly how it should on the oscilloscope. The Delay control doesnt seem to have much effect though, I suspect because the coil isnt properly damped. The next step is to rig an adjustable damping resistance, and properly match the coil and damping resistor.

Ultimately, instead of the damping resistor thats included on board, I will fit the matched resistors into the coils themselves. That makes it possible to change coil for different sizes and purposes, such as a huge coil for really deep searching, or a tiddly little coil for pinpointing.

No comments: