[Elecraft] Up-conversion.

[N2EY at aol.com]

In a message dated 8/12/06 8:46:40 PM Eastern Daylight Time, 
jack.smith at cliftonlaboratories.com writes:


> 
> It's interesting that Collins added the mechanical filter to the R390 
> (thus producing the R390A) as part of a 1955 cost reduction study. The 
> study report is floating around the Internet and I took a look at it 
> again tonight to refresh my recollection. The mechanical filter 
> substitution is discussed almost completely in terms of improved skirt 
> selectivity with little mention of cost savings. 

I think there were other changes, too. IIRC, the 390A is in some ways simpler 
than the 390.

It's also interesting 
> 
> to note that the mechanical filters made the receiver unusable for 
> direction finding as there was too much phase shift change with 
> frequency when compared with the R390's tuned transformer design.
> 

The story goes like this:

Back during the Cold War, No Such Agency had the idea for a secure comm 
system which eventually became known as F9C in FCC mode labeling of the time.
They wanted to use R-390As for the system, but the phase shift through the 
mechanical filters was a problem. So they had a version of the R-390A made with 
R-390 type LC IF strips. This variant was designated R-725.

To cover their tracks about the reason for the variant, a cover story about 
an electronically-steered RDF system was invented. The cover story was so 
convincing that the "other folks" apparently never became aware of, nor penetrated, 
the F9C system. Eventually the F9C system was replaced by newer technology, 
which is why we know the story now.

It is my understanding that, long after the F9C system was in operation, an 
electronically steered RDF system using the receivers *was* developed! 

All ancient history now. 

> Drake's first R4 used 50 KHz IF for selectivity but later receivers went 
> to crystal filters.
> 

Actually the R4, R4A, and R4B all used LC filtering at 50 kHz for selectivity 
and passband tuning. Only the R4C used all-crystal filtering. 

> I imagine the move to crystal filters was driven by the desire to reduce 
> receiver size and to remove the need for skilled technicians to align 
> the IF stages. 

I think the main reason was the amateur HF SSB transceiver, which began 
replacing 'separates' in the late 1950s and early 1960s. While the first ones 
(Cosmophone 35, KWM-1 & -2) used 455 kc mechanical filters, other manufacturers 
went to crystal filters in the HF range, probably to both save money and avoid 
being dependent on Collins. 

IMHO, the amateur HF SSB transceiver is what really made SSB popular with 
hams. By the early 1960s, a ham could buy a 100-watt-class SSB xcvr for about the 
price of a really good receiver. No critical zero beating, and the whole 
works was much smaller and lighter than the common desk-crusher AM rigs.  


In the VHF and UHF world, Motorola stuck with potted LC 
> 
> IF filters at 455 KHz (the famous "Permakay") well past the 1970's, 
> although GE went to crystal filters with the Master Pro (and maybe 
> before that; I'm a bit hazy on GE's gear between the ProgLine and 
> MasterPro.)

Could it be that obtaining a suitable FM-bandwidth filter was easier/less 
expensive (for Motorola, anyway) with LC circuits? 

> > I'm not sure of the exact dates when it was first done, but by the mid 
> 1960s 
> > the use of computers to do circuit simulation and calculation was 
> mainstream 
> > in electronic design. Such design tools probably had an effect in that 
> many 
> > "paper designs" could be tried out in a short time, particularly for 
> things like 
> > filters.. 
> >
> >   
> I received my EE undergrad degree in 1968 and I recall using a 
> transmission line simulation program. Batch mode, submit your card deck 
> and get the answer back the next day. Zverev's classic book on filter 
> design was published in 1967 and its data tables are derived from 
> computer programs, but the book itself is silent on using computers to 
> design filters.
> 

Well, I graduated EE school in 1976, and by then programs like SPICE were 
part of the curriculum. Still batch processing, though, but the whole idea was 
neither new nor revolutionary then. 
> 
> > Agreed. The main filter in the K2 uses selected microprocessor crystals 
> and 
> > some varicaps, yet gives very good CW performance and multiple bandwidths. 
> By 
> > comparison, one can easily spend half the cost of a basic K2 (or more) on 
> a 
> > couple of packaged CW filters for an HF IF. For VHF? I don't want to go the
> re.  
> >   
> I've been quite pleased with the filters as narrow as 200 Hz that I've 
> built using 8 MHz micro processor crystals. 

My point exactly. 

I have to grade them for 
> 
> minimum Q and frequency matching, but the resulting filters are textbook 
> matches to the Gaussian prototype designs.
> 

It is my understanding that the crystals in Elecraft products are graded in 
similar fashion. I think they are selected/matched for a number of different 
parameters.

Sometimes I wonder how many "microprocessor crystals" in the HF range 
actually wind up working with microprocessors...

> One of the two articles I've submitted  to the ARRL for consideration 
> for QEX covers some of my crystal filter work. Have yet to hear if it's 
> accepted, however.
> 
> 

I hope they are accepted. 

73 es tnx de Jim, N2EY



> Jack
> 
> _______________________________________________
> Elecraft mailing list
> Post to: Elecraft at mailman.qth.net
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft    
> 
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
>