I was thinking about some of the challenges faced by the builders of the SolderSmoke Direct Conversion Challenge Receiver, and about how similar these challenges are to those described by Wes Hayward W7ZOI and Dick Bingham W7WKR in the November 1968 QST article that launched the direct conversion revolution among radio amateurs. You can read the full article beginning on page 15 in the link above.
Some observations and comparisons:
--The November 1968 QST article said, "This receiver was designed for simplicity and ease of duplication, rather than ultimate performance." Ours too!
-- Wes's receiver has a single tuned circuit in the BP filter. Ours has two LC circuits.
-- Wes's mixer is also a diode ring. He starts out using hot carrier (Shotkey) diodes, but later concludes that ordinary diodes would work just fine. We reached a similar conclusion. But I wonder if the ordinary diodes would work well with a low output level from the single FET VFO (see below).
-- His oscillator uses a single MPF-102 in a Hartley configuration with no voltage regulation, and no buffer. We have two active devices and a Zener diode. This article makes me think we could have made our PTO even simpler.
-- Wes's receiver has a low-pass filter between the mixer and the AF amp. The cutoff is at around 2 kHz. This seems quite low in frequency and may reflect a preference for CW. It features 88 mH coils that are now quite hard to find. The goal of this filter seems to be to prevent signals from beyond the audible frequency range from overloading the AF amp. We ended up using the diplexer from the W7EL optimized QRP rig. I think this diplexer takes care of the problem.
-- The W7ZOI/W7WKR receiver has no AF nor RF gain control. When encountering a strong SSB signals, the article recommends detuning the BP filter. I think our AF gain control, and the mod calling for an RF gain control will give the operator, well, more control and will prevent strong SSB signals from overloading the AF amplifiers. The RF gain pot might also help prevent SW broadcast AM breathrough.
-- One big difference between our receiver and the November 1968 QST receiver: isolation of the VFO. The QST article puts the VFO in an aluminum box above the chassis. We have the PTO without any shielding right alongside the other circuits. Builders might want to experiment with the kind of isolation recommended by the QST article. Would this kind of isolation and shielding improve performance?
-- Wes also obviously contended with oscillation by the AF amplifier, as did many of our builders. The QST article contains a number of recommendations: First test the amplifier to see if you can hear noise. If it oscillates, try increasing the value of the decoupling resistors. (Many of our builders added electolytic caps to the DC power line in the AF amp.) The article recommends trying a .01uF cap across the output. It also recommends keeping the output of the amplifier away from the low pass filter at the input. Wes's design has no transformer as it makes use of high Z headphones.
-- The QST article says that the product detector performed adequetly with an LO injection level of .6 volts peak-to-peak. This seems quite low to me, but perhaps this would work with hot carrier diodes in the diode ring? This might be one good reason to use diodes that have a lower turn-on voltage -- you could get away with using a super-simple VFO even if it provides lwer voltages to the mixer. It might be fun to experiment in this area.
As readers can see, the challenges faced by the builders of the SolderSmoke DC receiver were very similar to those face by the builders of the November 1968 machine. I think all of us should find this very encouraging.
Thanks again Wes and Dick.
Posts
