[ARC5] Re: [Milsurplus] SCR-274-N Transmitter Dial Accuracy

[David Stinson]

----- Original Message ----- 
From: "Michael Tauson" <wh7hg.hi at gmail.com>
Subject: [Milsurplus] SCR-274-N Transmitter Dial Accuracy


> The military version Type K transmitter was specified as having a
> frequency precision of .03% of the dial reading......
> Western Electric apparently was unable to make tuning components
> consistant with the design requirements.  
> As a result, A.R.C. requested and got a variance for WE 
> that allowed the .04% radios so long as they were
> tagged as such.

You might be surprised how specifications like ".03%" come about.
I worked on a years-long project that belonged to one of the 
national labs.  It involved large DC drive motors,
maintaining a torque within 4% of a specification.  
We struggled and struggled for months, spending millions on this.  
Finally, during a big meeting, someone asked:
"Where did this 4% specifications come from?"
After much hemming and hawwing, it finally came out 
that the design engineer had, in effect, 
pulled the number out of the air because it "sounded right."   
After some experimentation, we discovered 
that 10-15% worked just fine 
and the project  finally moved forward.
I wouldn't doubt that someone pulled .03% "out of the air,"
with no real data to show why that kind of spec was needed,
just because it "sounded right." 
Given the intentionally wide selectivity of the receivers,
.03% is silly.  At 4.4 MC, the Army Airways freqs,
.03 % is about  1.3 KC.  
If you relaxed that spec to .05%,  it's 2.2 KC.
With maximum error on both ends of the circuit,
you're 4.4 KCs away from each other- 
still within the IF of a BC-454.  
I think that, like our lab engineer, some junior-grade 
Sig Corp officer made that spec up to sound impressive
and everyone got stuck trying to meet it.
That kind of "design" is very common, 
especially in government projects.

73 Dave S.

As an aside:  This wasn't the first collaboration between A.R.C.
and Western Electric.  They had a similar deal  
that covered the last of the GF/RU sets.  
Correct me if I'm wrong,  Michael:
>From the 1920s through the war, while A.R.C. did all the
design and engineering,  they actually built at their factory
only small lots like the RAT and VHF AN/ARC-5, prototypes, 
inital "spec" runs like the first sets of 1470-NY-41
and accessories like shock mounts and control boxes,
farming out the big radio unit orders to contractors, 
which they then rebranded.  Stromberg actually built almost
all the "RU" receivers branded for A.R.C.  I haven't found
data saying who built the transmitters.
I don't know if A.R.C. built the A.R.C.-branded 
AN/ARC-5 sets, but it wasn't Stromberg;  
S.C. sets have obvious differences.