[HBR] Re: HBR -- Part 3

[Peter Bertini]

Are those green sockets Western Electrics?  Your layout looks very
tidy. We're all awaiting more updates!
Peter


On Sun, Oct 12, 2008 at 6:09 AM, Walt Hutchens <waltah at earthlink.net> wrote:
> Heathkit-terminology steps since last time:
>
> Second mounting of parts -- output transformer, volume control, filter
> choke, ON/OFF switch, speaker, and misc.
> First wiring -- filaments and plate power supply
> Second wiring -- 1st & 2nd Audio stages
>
> Only made one mistake wiring the filaments this time! Pins 4 & 5 on 7
> tube sockets; how, you wonder, could I get that wrong? Well ... I
> wonder the same thing.
>
> The audio section worked right away, though I had the interstage
> coupling cap wrong, so had a bit too much hum. With that fixed, it is
> silent at low to mid volume settings, only some hum at full volume.
>
> The full volume hum comes from capacitive coupling between the chassis
> (which is at power line ground voltage) and the neutral power wire
> which is B- and the audio signal ground for all circuits.  This
> voltage is of the order of one volt RMS and usually 'trashy.'  With a
> 750k volume control, just a few mmf of stray capacitance is enough to
> produce an audible signal.
>
> The audio detector has a much lower (than 750k) plate resistance; when
> it's wired it will lower the impedance at this point, eliminating the
> hum.
>
> The audio stages have a high level of negative feedback and thus
> relatively little gain -- I think a voltage gain of 3 from the volume
> control to the voice coil.  But 100 mW of audio for a 4 ohm speaker is
> about 0.65 volts RMS so with voltage gain 3, volume control set for
> 10% of max voltage gain, the detector output will only need to be
> a bit over 2 volts and the 2nd IF output, in the milivolt range.
>
> Sharp cutoff tubes are fine in an IF operating at that low level.  But
> a sharp-cutoff tube in an RF stage is not ideal for ham band service.
> Does anyone remember how unhappy Hallicrafters was when some people
> substituted 6AK5's for the RF tube in one of their receivers?  Maybe
> the HQ-129X?
>
> This is not that bad, because the AGC voltages will be correct for the
> tubes I'm using, but better front-end performance would result from
> using 6EH7's in all three sockets. Then you'd have an 11 tube receiver
> that might possibly be done as a 300 mA series string set ... a
> different design!
>
> 'High fi' audio with frequency response shaped for communications
> purposes makes signals a lot clearer and takes just a couple more
> parts. Use of 1/2 of a 12AU7 for the audio output stage requires a
> high efficiency speaker: the modern midget extended low frequency
> response speakers used for computer and car audio will not deliver
> enough volume.  Speakers from taxi and CB sets are fine.
>
> Most of the circuitry of this set is copied from the earlier band
> imaging set with six tubes, push-pull oscillator and mixer, 80/40
> meters.  So I THINK all the problems with 'transformerless' power
> supply have been resolved.  Yeah, that's what I think ...
>
> I plan to wire the audio and AGC plate detector circuits next. These
> are respectively the triode halves of the 19JN8 1st audio and 2nd IF
> stages.
>
> Following that, I'll wire the local oscillator, so I can start working
> on the oscillator coil while wiring the other stages.
>
> The Eddystone dial that I have is factory-calibrated for six ham
> bands; I think it came as part of a kit of parts for the G3LOK
> receiver design. (Think: HBR with bandswitching, using itty-bitty
> coils ... NOT a great design!) And I'm using the tuning cap that came
> with.
>
> This dial is a nice (silk screen?) printing job on painted metal and
> the scales are very close to linear so it makes sense to try to track
> the oscillator to the dial, rather than recalibrating.  However this
> sort of thing requires a bit of cut-and-try ... probably a week's
> work, on and off, to get a decent result.
>
> This will require a three-point fit. Typically the three parts to be
> adjusted would be the coil inductance, the shunt (trimmer) capacitor,
> and a capacitor in series with the tuning cap, called the padder
> capacitor. These would adjust the low and high ends and the midband
> point, respectively. The W6TC designs use a tap on the coil rather
> than a series padder cap, probably because tapping the coil is cheaper
> than another high quality capacitor. However, moving a tap to get
> midpoint dial tracking right is tough.
>
> On the other hand, changing a capacitor down inside a coil form with
> an APC trimmer in the top may not be simpler ... I'll probably try
> using the tap, first.  I bought 10' of PVC for $2.39 plus tax, and I
> have a bunch of rusty octal tubes ...
>
> Just one more thought: that capacitor tracked that dial with the coils
> that are in that (busted up) front end assembly. Maybe my first effort
> should be to duplicate the in-set coils and copy the oscillator
> circuit parts. Hummm ...
>
> The tube lineup has turned out to be surprisingly tidy:
>
> 19JN8 -- RF & 1/2 mixer
> 19JN8 -- ECO LO & 1/2 mixer
>
> This looks messy, but the RF drives one half of the mixer and the LO
> drives the other, thus each pentode is coupled to the triode in the
> same envelope.  The two triode plates and cathodes are joined; the
> plates go to the 1st IFT.
>
> 19JN8 -- 1st IF & AGC clamp (triode is diode-connected)
> 19JN8 -- 2nd IF & AGC plate detector
> 19JN8 -- 1st audio & audio plate detector
> 12AU7 -- 2nd audio & 3.5 Mcs marker oscillator
> 12AT7 -- BFO & bias rectifier (2nd triode is diode-connected)
>
> 117Z6 -- Plate supply rectifier
>
> There's also a 150 mA 9-pin miniature with a triode and two diodes
> with independent cathodes that could be substituted for the 12AT7 if
> another diode is needed. The 14GT8 was designed for use in combination
> ratio-detector/1st audio service in AC/DC AM/FM receivers.
>
> Another report in another few days ...
>
> Walt
> KJ4KV
>
>
>
> ************************************
> Visit the HBR Receiver Web Site with over 100 pictures of receivers and
> construction notes...... via http://www.qsl.net/k5bcq/
> there is also a mirror (faster response)at http://k5bcq.edebris.com/
>
>
> Retrieve reflector archived data via http://mailman.qth.net/mailman/listinfo/hbr
>
>