Insight into "U47-inspired" with EF12 tube?

[soapfoot]

beautiful.  Thanks so much.  I'm working on a schematic right now.

 

[0dbfs]

Please post your schematic when you are ready!

So in your opinion, for my project-- with a regulated 6V supply for the EF12 heater, I can assume a very small load for the 105V supply for the rest of the mic?

Should be small compared to the heater current but will need to be close to compute the voltage drop across the CRC in the PSU.

I believe the plate-current is largely dependent on the cathode/grid operating points and if you know plate current and voltage-drop you can then figure out the load. This gets into analyzing the tube data sheets (which there may be none for an ef12 in triode mode).

How does one estimate the effective plate-to-cathode resistance (other than measuring a range of operating points / V&I)?

Will you be using fixed bias with the cathode seeing a fixed voltage between heater return and 0V or will you use an auto-bias scheme maybe with bypass cap?

If the cathode bias is coming from the heater line a regulator may introduce undesirable artifacts into the grid circuit.

These tube mic circuits sure do look pretty simple. I learned a little more today so thanks.

Best,
jonathan

 

[soapfoot]

Ah, I see.  This gets more complicated than I thought!

 

[soapfoot]

Please post your schematic when you are ready!

will do!

Should be small compared to the heater current but will need to be close to compute the voltage drop across the CRC in the PSU.

I believe the plate-current is largely dependent on the cathode/grid operating points and if you know plate current and voltage-drop you can then figure out the load. This gets into analyzing the tube data sheets (which there may be none for an ef12 in triode mode).

Still trying to understand what you're saying here.  I do have data sheets for EF12.  Don't know if they include any provisions for triode connection.

Will you be using fixed bias with the cathode seeing a fixed voltage between heater return and 0V or will you use an auto-bias scheme maybe with bypass cap?

Here's what I've designed--

If the cathode bias is coming from the heater line a regulator may introduce undesirable artifacts into the grid circuit.

Would love to hear more about this.  What type of undesirable artifacts are you talking about?

Thanks again!

 

[soapfoot]

data sheets for EF12--

Telefunken
http://www.mif.pg.gda.pl/homepages/frank/sheets/076/e/EF12.pdf

Philips (English)
http://www.scottbecker.net/tube/sheets/030/e/EF12.pdf

 

[soapfoot]

been doing some searching about artifacts from linear regulators in audio circuits. Seems that they allow some ripple through and add some noise of their own.  Perhaps I should avoid this and do a more standard DC supply with a choke.

 

[0dbfs]

Ah, I see.  This gets more complicated than I thought!

I don't know that for sure (whether it get's more complicated or not) so I wouldn't take my word on it but this is what I have observed in several instances.

Looks like the curves are for triode connected so that makes it easier. I don't remember exactly right now how to decipher/determine your operating points from the curves but there are threads around here about that too.

Would love to hear more about this.  What type of undesirable artifacts are you talking about?

Hum. It's because the heater line has to have absolutely no ripple if the cathode is referenced to it or it interacts with the grid operating point WRT the cathode. And if the grid voltage moves... That turns into audio.




Best,
j

 

[soapfoot]

Cool.  What I should probably do then is forego the regulation and create a standard DC filament supply with an LC and some RC pi filters and a voltage divider/rheostat allowing me to fine-tune heater voltage within a range of, say, 5-6.3v

 

[rodabod]

Just a quick point; your curves change as you underheat. Any bias voltage assumptions may at this point be thrown out the window if you so wish!

 

[soapfoot]

I'm revisiting lots of things right now.  Judging from this curve, if I'm reading it correctly:

If I had a B+ of, say, about 40-50 volts or less, and a bias of about 1.7 volts, then (assuming no underheating), current draw would be somewhere in the half a mA to 1mA range.

Right?

This is very little.  Does it seam unrealistically small?

 

[MagnetoSound]

Yes, now compare to (underheated) VF14. 

 

[soapfoot]

Yes it's unrealistically small, or yes I'm reading the graph right? 

 

[MagnetoSound]

It's not unrealistic.

Do your Ohm's Law on the U47 schem.

 

[soapfoot]

The U47 circuit has 105V input, 130k plate resistance, and 34V plate voltage.  This is a voltage drop of 69v.  Dividing this by the 130k resistance gives me .53mA.

Is my logic sound, and would this be correct?

 

[MagnetoSound]

Your logic is sound, but your answer is incorrect!

 

[soapfoot]

that's because of a simple mathematical error, I think:

voltage drop is 71V, not 69 (doh).

71 / 130 = 0.55mA.

Correct now?

 

[soapfoot]

Actually, I know it's correct, because 0.55mA is printed right on the schematic! 

 

[MagnetoSound]

... and relax

 

[soapfoot]

You've been super helpful.  I can't begin to thank you enough.  I'm learning a lot and that is always good.

 

[soapfoot]

Please post your schematic when you are ready!

Here is my first draft.  Can anyone see any glaring problems?  

Most of what I did is copied from other schematics and adapted a bit.  You will see obvious nods to the mk7 power supply.  I added voltage dividers/rheostats at the end to allow fine-tuning.  I added fixed resistors before the rheostats so that the voltage could never fall below a certain threshold, so that the entire range of the pot would be 'close.'

Please let me know where (if anywhere) I went wrong.

Thanks!