Russian 6N16B (6S6B?) Plate Resistor value?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kingkorg

Well-known member
Joined
Apr 15, 2017
Messages
2,973
Location
Norway
Happy New Year guys!

I made a tube mic in a Behringer B2 body with this tube using attached schematic.
Difference is i have B+ 120vdc.
I noticed i was getting about 40vdc at plate, so i changed plate resistor to 50k and got about 75vdc. 2vdc grid bias.  Is that ok? Did i mess something up? The mic sounds awesome, low distortion, low noise. Only thing i don't like is it's too clean(i used 1Gohm resistors), almost no difference from my FET b2. But that is another subject.

I am not sure as usualy there is 100k plate resistor value in most tube mic schematics. Is it ok to use 50k?
 

Attachments

  • 20180101_135820.png
    20180101_135820.png
    477.5 KB · Views: 171
can't really answer your question but, the electrolyt or tantal 22uF has a big influence on the sound beside the grid to ground resistor.  btw I like something in the middle 150-250M vs 1 giga or 100M.
 
From what i can find on the internet, general rule of thumb seems to be that plate resistor should be about 2x specified plate resistance or more. At the same time voltage at plate should be about 1/2 of B+. These are rough estimations without any serious math involved, i know it's way more complicated.

In my case 6n16 has 17Kohm Rp and my B+ is 120vdc so i am right where i should be. 2k2 as grid resistor gives best thd/output balance so i kept it.

Now, if i am right, why cant i find any schematic/mod with plate resistors lower than 100k? From what i can tell with my very limited knowledge on this matter there is huge potential in altering value of that resistor, along with grid resistor. And since all lower gain tubes i checked have lower plate resistance, why is everyone sticking with 100k?

If this is right, by lowering plate resistor i lower tube output impedance, and i guess i could use lower ratio transformer as well? 

Regarding the value of high value resistors, they impact sound of capsule tremendously, but it seems to be linked to capsule capacitance.  For example i used down to 10 meg elam251 values (ish) and while with k67(797-50pf) it didn't do much, with ck12 capsule i have (rayking 80pf) it rolled off 7db at 16k. That is acoustical measurement, the circuit is still flat. I couldn't measure any difference between 100meg and 1G for example.

Nobody manipulating frequency response with lower value resistors in FET mics? 
 
Maybe check how much dissipation the valve has to do ,if its within the spec it probably should be ok .
If your getting a good result ,I dont see an issue with 50k anode load ,you could always try something in between like 75 k ,and see how that sounds
 
Thanks, i was just puzzled that i couldn't find schematics with lower value than 100k.

If i understand well Ruud had good luck with 6s6b which should be single triode version of 6n16b, would be great to hear how he addressed this matter.
 
I think the u47 only has about 40 volts at the anode ,its works good ,but if you have a real screamer getting up close and personal  you do notice it compressing . I did a modded B-2 with an Ef 86 ,has around 70 volts on the plate, works fine as an overhead on drums ,or even in front of a guitar amp in the 30-50 watt range nicely cranked ,in that case your guitar amp will produce alot more distortion than the mic anyway .
 
There is a neumann Bottle mic that uses a 10k load  I believe ,runs kinda hot ,but a lower load does tend to open out the bandwidth of your circuit too .
 
Awesome, thanks, great info. Same story with m49 plate voltage here. Just found schematic for cmv 563 that uses 50K.
Would that mean that lower resistor drops tube output impedance and that lower ratio transformer could be used?
 

Attachments

  • 123.jpg
    123.jpg
    237.2 KB · Views: 96
Well, incidental capacitances tend to have a larger effect as the value of the anode load is increased , also  gain of the circuit is reduced as the load decreases  .
I have a picture of a graph I'll post in a few minutes once my camera battery charges that illustrates the point well .
 
Here ,its from a basic electronics book called 'Electronics made simple' by Henry Jacobowitz.
 

Attachments

  • P1010125.JPG
    P1010125.JPG
    287.6 KB · Views: 110
I varied the value of the cathode resistor, to set the correct anode voltage.
Because of the tolerance of the tubes, the anode voltage will always vary (a bit) when using a fixed cathode resistor value.
I also checked the distortion of the tube stage, as a function of the anode voltage.
And indeed, in most historic tube microphones you usually see an anode resistor of 100K.
 
Consider Ohm's Law. It is the variation in tube current, generated by the input signal, passing through the plate resistor that provides the voltage swing at the output.

The plate resistor gives this current something to fight against. For a higher value of resistance there will be greater voltage gain, up to the point at which the plate voltage collapses under load and distortion occurs. At lower values of plate resistance, there is less voltage gain and bias current becomes more critical.

The CATHODE resistor is adjusted to set bias current at idle. For a given value of plate resistor, the cathode current fixes the plate voltage at a useful operating point. For most small signal audio tubes, 100k (or thereabouts) tends to provide a sweet spot between good voltage gain, linearity, and bias point.

Naturally, circuits vary by application, tube type, and compromise beween all of the points described above.

Hope this helps.

 
Hi,
Would that mean that lower resistor drops tube output impedance and that lower ratio transformer could be used?

When you lower plate resistance (for a given b+ voltage) you change the current running through the tube (the lower Ra the more current).

As rp (internal plate resistance of tube), mu (gain) and  gm (transconductance: ability to vary current in response to a voltage) are linked if you change one of the parameters you'll change the others and when you change current through the tube you change the parameters.
So with more current through the tube transconductance increase and  rp decrease.

Here is a nice article about all that and the 'optimal' bias point of triode grounded cathode:

http://www.tubecad.com/articles_2003/Grounded_Cathode_Amplifier/Grounded_Cathode_Amplifier.pdf

As already pointed by others 'optimal' bias point will depend of your goal for the circuit: best voltage swing, best power transfer, etc,etc...

Other points must be kept in mind though, for example noise  (grid current) could be an issue as you increase current. It could depend of tube type some are better than others with this.

To come back to your circuit, in using 50k anode resistance you end with 0.9ma through your circuit. This is almost 1.5x current through classical circuit (m49, u47) but it is 'only' 2/3 of a more modern Sony c800g for example (1.5ma).   

All this circuit works but have a different set of trade off (lower headroom for the m49/u47 than c800, higher grid resistance for c800g to keep noise lower from the capsule part of the circuit).
 
Wow, thanks.
I will pop a simple switch ( and possibly keep it there) to switch between 100 and 50k and test real life difference and how it sounds on different sources/spl. It comes to that after all.

I am sucker for that kind of mods, switchable stuff. I actually had a question recently how i could achieve this exatct thing you guys are explaining now. I didn't get any real answers.
 
Use two potentiometers (linear taper) - for example 200k on the plate and 4.7k on the cathode.
Try adjust circuit in both ways. Then you will hear real difference.
Beside that, important is transformer ratio. For 10:1 or higher rather i would go for 100K or more.
Try this tube in parallel ;) (also higher plate resistance).


 
Sorry to revive this but I just stumbled upon this thread when I was researching the effect of higher plate resistors.
Because I found this schematic of an early M49.
Notice the 200k plate resistor? The early IRT version also had this.
Maybe worth to try...

index.php

 
Murdock said:
Sorry to revive this but I just stumbled upon this thread when I was researching the effect of higher plate resistors.
Because I found this schematic of an early M49.
Notice the 200k plate resistor? The early IRT version also had this.
Maybe worth to try...

index.php


Virtually all of the early AC701 circuits utilized a 200 k anode resistor and fixed biasing of the grid off the DC heater voltage, the ratio of the dividing resistors set to produce a grid voltage 1.6 v lower than the cathode, which is tied to the plus side of the DC heater voltage.  The polarizing voltage of the capsule was determined the same way, determined from the two 1 meg resistors in series across HT, and the polarizing voltage (60 v) taken from the junction of the two resistors.  The fixed bias is nice because this makes the polarizing and grid bias voltages independent of tube aging.
 
Back
Top