[HBR] Another Receiver Project -- HBR-4, Part 11

[waltah at earthlink.net]

Finally got around to sensitivity checks with the increased mixer 
drive level.   Sensitivity is signal for 6db S+N/N.

80M (3800) S9 = 60 uV   Sens = 0.8 uV 

40M (7300) S9 = 15 uV   Sens = 0.25 uV

There's certainly no need for an RF stage.   Of course it remains to 
be seen how serious the problem of in-band spurious signals will 
be -- there's at least one big one in each of these two bands but I 
have given that problem no attention yet.

The 30 and 20 meter bands are now working.  The only excitement 
there was the discovery that on 20 you don't need much coupling 
at all between the input tuned circuits -- anything over 1 mmf led to 
the classical double peak of overcoupled circuits.   There's 
substantial coupling through stray capacitance between switch 
decks and the trimmer wiring.

With WWV (10 Mcs) available I found that the 100 kcs calibrator 
needs a larger fixed cap to come on frequency.

I have a somewhat long wire (~3 wavelengths?) for 20M pointed 
toward Europe; there are Europeans all over the dial.  

I said:
> Those 50 mmf trimmers tune so sharply on the high bands that 
> you really have to hunt for the crystal frequency.  

Jim:
> I just yanked plates off them...

These are ceramic rotaries, 7 mm in diameter.   I used the crystal 
and RF trimmer assemblies from the FT-101 and mounted seven of 
those little trimmers on a scrap of GFRP for the premixer plate.   
It's not easy even to extract these trimmers from the boards -- 
there are always one or more bad ones to change and it is one of 
my least favorite jobs.   

I'm pretty sure the FT-101E used some smaller ones; if the tuning 
rate is a serious enough problem I probably have some spares I 
can liberate.

> G3PDM (or was it G3DPM) has a design in one of the RSGB handbooks
> of that era. Dualtuner straight into a 7360, xtal filter at 1680
> kHz, etc. His solution to the LO problem was interesting: 

> The actual LO is a dualtriode balanced oscillator, directly
> driving the 7360 deflectors. It is varactor tuned. 

G3PDM was the builder and it's a very interesting approach indeed; 
I'm glad you didn't mention it before I got well into this project.   The 
essential idea is that you build a bandswitching LO tuned within 
each band by a varactor and lock that to a synthesized signal.   
That keeps the impurities in the synthesized signal out of the 
received signal path and since the synthesized signal doesn't have 
to be large amplitude, they're going to be easier to contain, 
anyway.  With a good quality oscillator, phase noise should be 
negligible.   It takes maybe three or four more tubes -- he did it with 
solid state devices though.  

Actually I think I'd have had the sense to stay away from it for now. 
Direct synthesis is simpler and there's plenty that's brand new to 
me, even as the project is.

The hardest radio project I ever tackled was putting an ARC-38A 
aircraft transceiver back on the air.   The ARC-38 was a Collins AM 
set, frequency locked on a whole bunch of channels 2-24 Mcs as I 
recall.   RCA got the contract to convert a flock of them to SSB; 
thus was born the phase-locked ARC-38A.   Fifty-some tubes, 
mostly subminiature.   It took a few weeks of troubleshooting but I 
got it working, +/- problems caused by newly dead crystals.   And 
the nastiest problems were in the PLL circuit -- most of them being 
(as in all RCA military sets of my experience) due to lousy RCA 
quality control.   

The HBR-4 is a lot more fun.   Tomorrow I need to return to the 
balance of the mixer stage; I'm certain I don't have that right.   

Walt
KJ4KV