[Collins] Collins ham gear redesign

[Dr. Gerald N. Johnson]

>>

>> What would you have designed differently?
>>
>> I think I would searched for a remote cutoff pentode for the 75S IF
>> with more gain than the 6BA6 so the two stages didn't have to be
>> run up at their power dissipation limit to get the desired gain and
>> I would have complicated the power supply to have a 250 or 300 volt
>> output for a more common 6AQ5 output stage that was cathode biased
>> instead of fixed biased from the fragile bias supply to have gotten
>> the desired audio power with less heat from the output tube. I
>> would have never used a selenium rectifier for the bias supply
>> either.
>
>
> There were plenty of better tubes than the 6BA6 by the time the 75A4
> came along. The use of a single triode was a poor choice for a mixer,
> a dual triode such as the 12AT7 would offer more conversion gain and
> a high degree of overload handling followed by improved gain
> distribution..

Ah, the mixers. The 75A receivers used pentagrid mixers that have a few 
known characteristics. Easily overloaded, and very noisy.

The 6U8 and 6AU8 in the S-line were not great choices either. I traced 
the noise sources in my new S-3B (it spent a year on its side out of the 
case after I moved to Texas) and found that the second mixer was the 
primary noise source because the gm of the pentode of a 6AU8 was low, 
and as a mixer it was way lower. That limited the dynamic range of the 
receiver on both ends. I installed an adapter holding either a 6688 or 
7788 from Amperex with 50,000 umhos in place of the 6AU8 with 4000 umhos 
and then was able to drop the signal into the second mixer seriously. 
With a nuvistor 2m converter in front of it, I took the signal from the 
600 watts of W5WXV at about 20 miles from showing about 20 spurs in 200 
kHz down to three and one of those was where he was supposed to be and 
the other two were close to the noise level. I did take the receiver 
gain a little further down than was great on HF but I didn't hurt the 
system NF with the 2m converter out front.

A ham at Rippey Iowa who sold prop pitch rotors and remote antenna 
switches also sold a drop in replacement for the first mixer using a 
double gate MOSFET. He didn't agree that the second mixer was the 
limiting factor, but then if he didn't see it raining he would not 
believe it if I told it was raining. George McKercher was his name. He 
drove a pair of 4-1000A on HF with his S-line. With the plate variac at 
half voltage he only peaked at 1 amp and 3 KV on the meters when chasing 
DX with a visitor present in the shack.
>
>
> OTOH Collins didnt have a decent AGC for any mode on any model until
> the 75S3C. I modified both of my contest grade R4C's using that
> circuit plus added a Medium position.

The 3B and 3C were the same receiver except for the extra crystal board 
until very late production. I never noticed a problem with the AGC in my 
May 1964 3B. I copied that dual time constant filter to the Q-5er.
>
> The 75A family was even poor on AM for ham use, other brands were
> much better; heck my 1941 SX-28 or NC-200 is much better. The HRO-60
> is a dream to use as a battle conditions AM radio. The A4  AGC was a
> joke on CW/SSB and was one of the first things I changed with mine in
> 65.

There is a problem with the 75A-4 and replacement tubes. I have a 
project very slowly in the works to compare RCA 6DC6 to ECG 6DC6. They 
give an A-4 wildly different AGC characteristics with the ECG tubes 
acting more like 6AH6, sharp cutoff at a lower bias voltage than the RCA 
or ordinary 6BA6. Abd that also changes the S-meter calibration 
significantly.
>
>
Lots of more recent ham gear has had hardly working AGC. Icom used
to advertise their amplifier as blowing the receiving operator back
from the operating table, which it did with their receivers (One IC I 
used for FD one time appeared to have no AGC at all. Its owner bought a 
TS-440 after that and it was so different the day he got it he stayed on 
the air for 24 hours.) AGC of the FT-857 is so good that audio is no 
help at pointing 10 GHz dishes. Aiming is much easier with the AGC 
turned off.

On 7/7/2014 2:53 PM, Carl wrote:
>
>> One of the first mods I made to my new S=3B was to replace the BFO
>> adjustment pot that was switched at the CCW end of rotation with a
>> pot with a push pull switch. So when I set the BFO frequency for my
>> likes (or for narrow shift RTTY) I could switch to SSB and back
>> without having to go through the setting process again. I don't
>> know why it wasn't built that way from the beginning, unless the
>> engineers never used a modern TV set with a push pull switch. I had
>> no trouble at all buying the better switch and pot assembly in a TV
>> parts store.
>>
>> The filters of those days had a lousy time response turning
>> lighting into extended crashes and power line noise into continuous
>> noise (as seen on a scope that I connected to the jack I added to
>> the end of the receiver IF strip). A few years later I used a
>> vintage Q-5er with converter and wired up my receiver/TX patch
>> panel so I could use either the Q-5er or the S-3B or both with my
>> transmitters. Often on 75 meters I could copy through thunderstorms
>> with the Q-5er and couldn't with the S-3B because of that filter
>> ringing. The Q-5er selectivity came from loosely coupled ferrite
>> core IF transformers at 85 kHz and had a more Gaussian amplitude
>> response with a fine time response. Lightning came through as
>> clicks, not crashes. The modern Collins mechanical filters offered
>> as options in the Yaesu FT817, 857, and 897 don't ring nearly so
>> much.
>
> Thats why I like to use a NC-300 or HRO-60 under those conditions.
> One with a 80 KHz IF and the other brute forcing it with a load of
> 455kc IF circuits. The Collins R-390 is another decent storm radio
> but not the 390A. The USN RBB and RBC by RCA is perhaps the peak of
> electrical storm radios but I havent used a R-390 at sea. Yeah, I
> have those also (-;

390 and 51J-3 should be better than 390A and 51J-4 because of the lack 
of vintage mechanical filters. I have a 390A at the barn that the 
previous MARS member never got working, mice and moisture have probably 
not improved it in the last 15 or 20 years. I'm no long in MARS but MARS 
didn't take back any of the equipment though I asked them to, especially 
the 6' rack TMC independent sideband receiver with 100 tubes. A squabble 
over a repeated signature for it is why I left MARS though I was Iowa 
newletter publisher and technical coordinator. Its still with me but 
filled with mouse nests, I don't think I want to apply power.

I have a 51J-3 with product detector and better AGC added by an ISU 
professor, W0PFP who is now a SK. I'm not sure I've turned it on since I 
moved. It was working good except its RF stage will oscillate on the 3 
MHz band if peaked. It has a white panel that was tan with pipe smoke 
when I got it.
>
> Another USN radio is the mid 30's RAK and RAL regens, the peak of
> regen development. It uses an audio peak limiter as a form of AVC
> that is extremely effective and easy on the ears. I used a RAL often
> at sea for hamming and have the pair here also. The RAK (it tunes
> 15-600kc) is better on CW only and feeding a 455kc IF or building a
> crystal controlled down converter into it would be an interesting
> experiment.

The RAL and RAK were used on liberty ships through WW2 based on the 
design philosophy of nothing new, only proven designs and if the 
operator was careful there was no leaked LO for enemy subs to detect. If 
the operator let it oscillate it had two Rf stages to cut down that 
radiation but probably wasn't perfect.
>
> Carl KM1H
>
I don't know if the modern fancy SDR have come to it yet, but an SDR needs
two AGC loops. One to protect the D/A converters from overload and 
another to control the audio level. The front end AGC is seeing wide 
band signals, how wide depends on the specific design, so if that is the 
only AGC, a strong signal a distant frequency can be pumping the audio 
level. There needs to be a post detection AGC, like the limiter of those 
regens that works on only the audio to the user. I suggested that to one 
of the SDR gurus in 2006 and he hadn't though about that at all.

That audio level pumping can come from using a narrow audio filter after 
the RF AGC in most any receiver. Makes the audio or DSP filter 
relatively useless on a crowded band. Yes, AGC ought to be derived after 
ALL the selectivity but I sometimes use a passive low pass filter on the 
speaker wire to cut down on post filtering wide band receiver noise, a 
problem when the filter is followed by two mixers and 7 stages of wide 
IF that contribute almost as much noise as the front end passes through 
the filter. Plus the noise artifacts of the DSP audio filter.

I have a 2m transceiver that has two filters on the IF strip, one at the 
mixer (its single conversion with a VCXO for the LO), and one right 
before the product detector. Its called a Hohentweil.

If it had a good tuning range the Hohentweil would be an interesting VHF 
transceiver and IF for microwave. Not all the Minnesota area tranverters 
are rock stabile and sometimes there's a want to shift 2 or 2.5 MHz to 
allow more than one direction of contacts from the central hill top 
location at the same time. And as I work on higher bands a 432 or higher 
first IF would be beneficial in image rejection without lossy complex 
millimeter wave filters.

73, Jerry, K0CQ, Technical Adviser to the Collins Radio Association

>