My first tube mic schematic

[pasarski]

I finally have the schematic for my tube mic. I'm still short of many of the components, but thought it's a good time to send this schemo here, if anyone's interested to look at it for obvious mistakes (this is my first build)? I started from the ELA M 251E circuit, took some values from the Royer circuit, and apparently ended up to what is basically a G7 circuit modded to 5840 tube. I tried to change values in the PSU for lower plate voltage, don't know if I calculated right.  5840 needs about 7,5 mA of current (I think). Don't know if it's getting enough. Here's how it looks. Could this thing possibly work?

http://www.flickr.com/photos/47781016@N06/4390421363/

It would be also nice to hear if there's something that's not needed. I'd like to keep it simple.

EDIT: Ooops, there was one quite big mistake, no voltage to the plate. Changed the picture.

 

[Gus]

First learn ohms law.  current = voltage / resistance.  Lets make up some numbers if you have a 100K plate the

Voltage to the "top" of the of the plate resistor is 100VDC

Voltage at the plate is say 50VDC

The voltage across the 100K plate is 50VDC

50VDC/100k =.5ma this is the plate current

Now read up on operating points of tube circuits and tube curves.

Now go to the tube curves with the tube set in triode mode if a pentode

Look up the plate current at and find the needed bias voltage

Problem is you might need to trace the curves yourself because you might not find a triode curve of a pentode

If you look at something like the Royer 2001 tube circuit say the supply voltage is 100VDC and you have a 100K plate the MAX plate current you could have would be less than 100VDC/100K = 1ma

I wish people would stop calling circuit that are not exacty the circuit the circuit name just call it a cathode biased or fixed bias triode plate out tube microphone gain stage or what ever it is.

The detail are what counts However simple tube microphone circuits will often work and pass signal even not designed correctly.

 

[pasarski]

Thanks. Sorry for my noobinesh, I just looked at the operating points in pentode mode for a class A amp (that's how ignorant I am at this point. And calculated it all wrong anyway. So I have max plate current of 140VDC/(5,6K+100K) = 1,3 mA? If my B+ really is 140.

I wish people would stop calling circuit that are not exacty the circuit the circuit name just call it a cathode biased or fixed bias triode plate out tube microphone gain stage or what ever it is.

The detail are what counts However simple tube microphone circuits will often work and pass signal even not designed correctly.

What comes to this, I didn't know what's the policy, sorry. Thought that it would be taking the credit from people who I have copied if I don't mention the sources. Should have known better (EDIT: Whoa, I see now how badly I named the topic, changed it). Well, lot to learn, and I sure want to get in the details some day, after I know the basic theory better.

 

[Gus]

If you read all the threads you can find at this site about microphones you might find one with the part values and voltages for a 140VDC powered 5840 circuit I posted.

 

[pasarski]

Ok. I thought I have read them all I try to search more, maybe that one have just slipped my mind.

Thanks for answering, i have lot's of respect for the work you and others on this site have done, and amount of information available here is amazing!

 

[Gus]

http://www.groupdiy.com/index.php?topic=35464.20

 

[PRR]

As a matter of "style", I like to see inputs on left, outputs on right, B+ at top and gnd/B- on bottom, and heater lines not drawn.

With the tube symbol you used, input (capsule) left of tube would avoid a lot of runaround. Less work for your pencil, and less eye-crossing for anybody who might bother to look at your drawing.

750K seems an unlikely cathode resistor.

Is that really 10uF across the capsule?

 

[pasarski]

I copied some values from Gus and followed PRR's suggestions about the style.

http://www.flickr.com/photos/47781016@N06/4396273492/

Looks better now, right? (And 750K and 10uF were mistakes)

Couple of things I'm still struggling with. What determines when you can make a circuit without a grid coupling cap? Can I omit it in this design?

And the top resistor in the PSU voltage divider. I can't determine it's exact value without knowing the voltage at the plate can't I? And I can't know the voltage without building the circuit first, right?

 

[PRR]

> Looks better now, right?

Get the tube out of the cellar:

> And 750K and 10uF were mistakes

So are the rectifier connections. Re-plagiarize those.

I don't want to discourage, but you should be working on something simpler.

 

[pasarski]

I don't want to discourage, but you should be working on something simpler.

I know. But I have plenty of time for this and a friend to help me with the soldering, and I'm only interested in mics, not stomp boxes, so no, you're not discouraging me, I'm gonna pull this off somehow some day.

And thanks to you're help (I really appreciate it!) it's at least starting to look simpler

http://www.flickr.com/photos/47781016@N06/4396917495/sizes/o/

EDIT: I noticed the voltage divider in the mic was connected to polarizing voltage line instead of ground, all by my self  , so I think I'm making progress.

 

[rodabod]

I'd recommend taking the G7 and copying it directly, and maybe compensate the bias a little by changing the cathode resistor if necessary. As Gus pointed out, you'd need to consult the valve's characteristic data (curves specifically).

 

[pasarski]

I'd recommend taking the G7 and copying it directly, and maybe compensate the bias a little by changing the cathode resistor if necessary. As Gus pointed out, you'd need to consult the valve's characteristic data (curves specifically).

Actually that's how this started, I wanted to build G7. Then I ran into bunch of 5840 valves. Thought about building the Royer MXL 2001 circuit, but wanted to make it multi pattern (I'd like to buy a Tim Cambell CT12 when I get the thing working). Then read here about all the modifications that could make the Royer circuit even better, like a cathode bypass cap and the grid resistor. At one point I freaked and wanted to have ELA M 251E style pattern control for better cardioid operation, but have now put that aside. So I decided to build G7 PSU, modified for lower B+ (cause I don't need/can have so much for the plate, and am going to use a cheap chinese capsule of the body donor mic before I buy the CT12, and don't want the membranes sucked to backplate, I thought 70 V is good compromise). One thing is that the mic I'm thinking to be a body donor won't fit the EF86 and a Lundahl in it.


Then Gus linked me to info that he have used 140 V B+ (same as mine), 100K plate resistor, ank 1K cathode resistor with a bypass cap (Royer used 750 but with out the cap) with 5840. So that's where I'm standing at the moment. I haven't been able to find 5840 triode transfer characteristics curves  (I have found plate caharcteristics curves), but I have now values that Gus suggested. Does the different polarizing voltage affect these values?

So I really don't want to go a month back in time and just build the G7, which would probably be a very fine mic, but wouldn't give same amount of pleasure as trying to figure things out step by step and understand what I'm building instead just build.

So is there still drastic mistakes in the last schemo I posted, or could I just try it and mingle with some values if needed?

EDIT: I gladly take "yes" as an answer, and try to figure it out on my own.

 

[rodabod]

Triode curves:  http://scottbecker.net/tube/sheets/137/5/5840.pdf

I'd eyeball about for 100K plate @140V B+: -2.5V bias @0.75mA idle.

So, that works out about 3K for your cathode resistor. You can select6 this value with a pot if you prefer. Or you may prefer the sound with it biased a little hotter.

 

[pasarski]

Triode curves:  http://scottbecker.net/tube/sheets/137/5/5840.pdf

I'd eyeball about for 100K plate @140V B+: -2.5V bias @0.75mA idle.

So, that works out about 3K for your cathode resistor. You can select6 this value with a pot if you prefer. Or you may prefer the sound with it biased a little hotter.

Thank you! I can finally use plate characteristic curves!

(Don't read the following, I'm just writing down my realization, and I know I should have known these things before I start to mess around)

Yes, 140V/100Kohm=1.4mA , my max current, I draw the load line (hard to be accurate when this low on the scale), and yes, 0.75mA is probably good for me (have read that 1mA or little less is good). So -2.5 V bias. 2.5/0.75=~3K, yes, I got it!

EDIT: removed some nonsense.

I think I've learned something. I have much more I'd like to ask but don't want to be too annoying and bring it all up in a single thread Thank you all for help, you guy's are magnificent!

 

[pasarski]

And rodabod (and PRR, I think I misunderstood you  at first). You convinced me to go for a ready made design (G7 with PSU modded for 140V B+ and 5.7V heater supply) cause my experience is what it is. (Maybe modding the body donors head basket to one layer mesh to suit G7 circuit better?) Sounds solid?

 

[rodabod]

Changing the mesh could improve things. Maybe not. It just depends on the combination of the different parts.

Make a G7 with a 5840 and adjust the parts as you go along.

 

[pasarski]

Make a G7 with a 5840 and adjust the parts as you go along.

That's the plan. And gonna use trimmers in some spots to test different values.

 

[rodabod]

That's the plan. And gonna use trimmers in some spots to test different values.

Cool. Check what sort of headroom you get with different bias levels (vary Rk) and also note the change in sound. You might find that colder bias results in a change in distortion characetristic.

 

[pasarski]

Ok. And just a noobie terminology check: hotter bias = more current flow and smaller negative grid voltage( or vice versa :)?

Bidding on the body donor in evilbay and ordering the missing components in few days. Fingers itching.

 

[rodabod]

Ok. And just a noobie terminology check: hotter bias = more current flow and smaller negative grid voltage( or vice versa :)?

Correct. Cold bias tightens up that tap.