[R-390] R-390A s/n improvement 6HA5 mixer - CONCLUSION

[Larry H]

Hi Norman,  That's an excellent question and one I should have addressed.  I thought about it during my tests, but forgot to mention it in my post.  Thanks for asking.
I let all the tubes warm up for 5 minutes or more during the rating.  I know that's not much, but after rating each tube and using them in an appropriate location, my S/N did not change any measurable amount after 12 hours and 24 hours.  I do not know if the tubes I did not use would improve if baked in.   There is certainly that chance.  In Roger's write ups, he did mention that SOMETIMES a used tube would have a better 'merit' than a new one, but that just might have been that way even when new.
As I was rating the noise merit of each tube, I was also rating the gain of each one.  I measured the diode load voltage with 1 uv input.  As you would expect, I found differences there as well and recorded the merit and gain on each tube.  I then went through them and selected the lowest noise with best gain, always weighting the noise as the most important.  As Roger stated: low noise and low gain is not of any benefit.  The tubes must have reasonable gain.  I drew the line at about 90% of maximum measured with this method.Regards, Larry

      From: Norman Ryan <nnryann at yahoo.com>
 To: Larry H <dinlarh at att.net>; R-390 Forum <r-390 at mailman.qth.net> 
 Sent: Tuesday, August 9, 2016 8:33 AM
 Subject: Re: [R-390] R-390A s/n improvement 6HA5 mixer - CONCLUSION
   
Hi, Larry,
I'm intrigued by this tube noise issue and have contemplated swapping tubes for obtaining optimal s/n as you have done.  I wonder, does the tube have to be thoroughly warmed up?  Would its noise characteristic change for the better (or worse) if the tube were allowed to burn in overnight in the receiver?
Taking a bunch of tubes (32 in your case) and waiting for each to warm up to fully operational condition before measuring s/n would take an inordinate amount of time.  To be sure, a fair assessment of each tube for comparison purposes can be obtained by subjecting each candidate tube to an identical brief (say, five minutes) warmup.
I still wonder if performance changes measurably when the receiver is left on overnight so as to let the tube under consideration fully optimize.
The differences you note among all those tubes are striking -- I had no idea the s/n of good testing tubes could vary so much.
Many thanks for all your work and for sharing your findings with the group.
Vy 73Norman

      From: Larry H <dinlarh at att.net>
 To: R-390 Forum <r-390 at mailman.qth.net> 
 Sent: Tuesday, August 9, 2016 12:32 AM
 Subject: Re: [R-390] R-390A s/n improvement 6HA5 mixer - CONCLUSION
  
While testing and measuring the results of this mixer test, I wanted to find a more precise way to adjust the IF gain for optimum performance, as it affects the S/N more than I suspected.  More on this in a later post.


I spent a few days reading through the Pearls and Y2KR3 and found a couple posts by Roger Ruszkowski on how to improve the S/N by rating the noise level (merit) of tubes.  I have known for a long time that tubes could be noisy and that one should swap some to find the best.  Well, what I was doing was too subjective and really did not accomplish much. After I read Roger's write up on how to actually measure the noise 'merit' of a tube, and putting it into practice on 1 of my 390As, I was able to improve the S/N on it USING 6C4 MIXERS from an average of 1.9 uv to an average of about .5 uv.

In order to accomplish this, I had to use a lot of 6C4's (14), 5 6DC6's, and 13 5749's. 
I started by getting the IF and AF to meet Roger's 30 db modulation on/off test described in the Y2KR3 manual.  I was not able to get to 30, but did get to 29.  I then went to work on the overall S/N.  I started with the spare tubes I had, but that did not work well enough (I only got it down to 1.1 uv).  I had to buy more and pull the ones in use in my other 390A and even my SX-111A.  I had the most trouble finding quiet 6C4's, but finally did.  Here's my S/N results:

    MH: 1.9, 3.9, 7.9, 9.9, 15.9, 21.9
    UV: .36, .42, .35, .42,  .54,  .68

Because I was able to acheive a good S/N level, I wanted to see if using a 6HA5 as a mixer would improve these numbers any.  I know that with this level of S/N, improving it does not improve weak signal reception, but just wanted to see.  I put 1 in the 2nd mixer and realigned.  Well, it did improve the S/N, but only very little, about 14%.  I put the 6C4 (my best 1) back in.

At this point, I thought I'd see if it would meet Roger's end-to-end 30 db modulation on/off test described in the Y2KR3 manual.  Well, it does not quite.  I only got it to 26 db.  But, the performance is good enough for me, so I'm not going to pursue it.

When I'm listening to SSB, I prefer not to hear tube noise between sentences.  On my 390A, it needs to have a low S/N (about .5 uv) and sensitivity of 1 uv for -7 v on diode load in order to accomplish that.  Now it does and my SSB reception is good (without using 6HA5s).  And, when listening to AM that fades almost all the way out, I no longer get that annoying hissing noise during the fade.  This low S/N is real nice in these 2 areas.

As a further test, I transplanted some of the tubes from this 390A into my other 1 and got similar results.  The S/N on it went from average of 1.8 uv to .51 uv.  Now I know it's S/N problem is also tubes.  

In conclusion, I'm glad I looked at using the 6HA5 as a mixer.  It forced me to determine the real problem with its poor S/N.  Most, but not all of it was due to the 6C4's.  Now my big problem is what to do about my other 390A and my sx-111a, since I took the quiet tubes out of them.  I think I'll put 6HA5 mixers in my other 390A until I can find some more low noise 6C4s.

Regards, Larry