I wanted to refresh my memory and looked up the 6146B.
Now, some decades ago i used the quite similar QE05/40 in triode.
To my suprice :
RCA 6146B Rg1 fixed bias 30k
Philips QE05/40 Rg1 fixed bias 100k
Now, nobody wants tube run away, it could be that RCA
choosed a bigger safety factor or whatever, doesn matter.
But, what earlier went unnoticed by me, and really puzzles me,
what could be the reason for: ...only FIXED bias is recommended?
Now, some decades ago i used the quite similar QE05/40 in triode.
To my suprice :
RCA 6146B Rg1 fixed bias 30k
Philips QE05/40 Rg1 fixed bias 100k
Now, nobody wants tube run away, it could be that RCA
choosed a bigger safety factor or whatever, doesn matter.
But, what earlier went unnoticed by me, and really puzzles me,
what could be the reason for: ...only FIXED bias is recommended?
It is only a guess as far as when some the grid may get pulled positive by impurities inside the tube.
Going with 30K is just more conservative. Or maybe Phillips had better production tolerances as far as avoiding gasses inside the tube.
You probably want to be sure your tubes get hot enough to activate the getter also.
As far as no warning for cathode bias it may be that when the grid starts running away, there will be more negative bias due to slightly more cathode current.. Not so much with fixed bias.
Going with 30K is just more conservative. Or maybe Phillips had better production tolerances as far as avoiding gasses inside the tube.
You probably want to be sure your tubes get hot enough to activate the getter also.
As far as no warning for cathode bias it may be that when the grid starts running away, there will be more negative bias due to slightly more cathode current.. Not so much with fixed bias.
RCA simply states no value for automatic bias Rg1 and not recommanded.
Philips states Rg1 max 100k, and later, in the pages for CCS operation as modulator or LF amp, again 100k bùt includes also a note, under the recommended operation conditions only fixed bias is recommended.
Worst case for what, what bad could happen with automatic bias that could not happen with fixed bias? Exceeding Vg2 ? The tube can handle 400V on g2 in triode, so unlikely that. Exceeding Ik max during start up ? Possible with a huge Ck, but those where not avaiable at the time. Exceeding Vfk ? Unlikely, Vfk is 135V. What is left ?
Philips states Rg1 max 100k, and later, in the pages for CCS operation as modulator or LF amp, again 100k bùt includes also a note, under the recommended operation conditions only fixed bias is recommended.
Worst case for what, what bad could happen with automatic bias that could not happen with fixed bias? Exceeding Vg2 ? The tube can handle 400V on g2 in triode, so unlikely that. Exceeding Ik max during start up ? Possible with a huge Ck, but those where not avaiable at the time. Exceeding Vfk ? Unlikely, Vfk is 135V. What is left ?
Last edited:
Cathode bias would provide a safety factor with tube runaway, not the other way around
A1, B1, B2, all fixed bias only, but why could you not use automatic bias with A1, especially since the current is almost the same, only 3% difference between no load and full load?
Last edited:
Sorry, I sort of said that backwards. Cathode bias would be safer than fixed bias.
I know someone made an audio amp using the 6146 back in the 60s. Really the 6146 was more expensive than other tubes considering they handled more shock than a 6L6 or 6550. Also did better for RF.
It was sort of a child of the Korean/Vietnam wars as far as a reason to develop the tube. A bit more effort was put into making it work in mobile situations with less worry about heater voltage tolerances.
It was sort of a child of the Korean/Vietnam wars as far as a reason to develop the tube. A bit more effort was put into making it work in mobile situations with less worry about heater voltage tolerances.
Recently heard that Finland is home to the happiest people on the planet. Congratulations on doing things right! America has dropped out of the top 25, but that's because we're going through a painful Whiny Poopy Pants era, and will hopefully recover somewhat within my remaining lifetime.
All good fortune, and stay happy,
Chris
All good fortune, and stay happy,
Chris
Tnx, the affix "f" corresponds to a class B application.
It looks to me now, that for some reason, those datasheet writers felt a need to extra point out that cathodebias is "not recommended" when
operating in B.
Thats propably all that is to it, because inn the same datasheet one can find other types of applications with various cathode resistor values.
From what I understand, this is dangerous to operate two sections with a common cathode.
If one section is weak, the other takes all the current, redplate, etc.
When I used 829/GU29/GMI6 in PP class AB (spud), it was fixed bias, if you measure plate current of each section with a small resistor in series of each plate, you can see quite often plate current is not well matched.
If one section is weak, the other takes all the current, redplate, etc.
When I used 829/GU29/GMI6 in PP class AB (spud), it was fixed bias, if you measure plate current of each section with a small resistor in series of each plate, you can see quite often plate current is not well matched.