MXL603 tube/trafo mod

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

bezen4uk

Well-known member
Joined
Jul 27, 2009
Messages
134
Hi guys,
This week I finished modding my MXL 603.
Now it's 603T :).
http://www.flickr.com/photos/fuzzard/5511207555/
Guts: russian tube 6S6B, custom made transformer with 8:1 ratio, point-to-point montage, few caps and resistors. Very simple and nice sounding mike!
Just tried at the local studio. We've made some tests side by side with AKG 451, Neumann 87 and Oktava MK 012. Yeah, there is some lack of bass response (I think it's capsule or maybe transformer). Mids and highs ... I never heard such bright  response (in positive sense). Just subjective impressions.
Will post circuit asap.
The questions how to force the iron in this mike are welcome :)
 
Hey, that looks great - really nice work to get it all fitted in there so neatly!

Would be nice to see the circuit if you can share it.
 
Circuit, as promised:
http://fuzzard-audio.com/MXL603_tubemod1.pdf

Plate current is 0.39mA; voltage drop at R6 is 0.85V; amplification of tube is about 18; output impendance (measured at the secondary winding of transformer) is 230 Ohm.

So, the tube is biased like VF14 in U47 mike.

The heater is powered by current source regulator.
 
wow! very nice! And it looks like an effortless build sure it wasn't.... where did you get that transformer?
too bad i sold my set of 603s a few years ago, since then the price has gone up at least 150 euro compared to what i payed then..

greetings,

Thomas
 
re: bass drop, have you tried a larger output cap (if it even fits)?  
 
transformer:
I got laminations from old russian transformer (it's EI laminations with 0.224" stack width). It was impossible to fit standart bobbin, so I made bobbin with less height. Please look at attached picture: at the left part you see how the laminations assembled in standart core and right part - how I assembled them in this transformer (without I laminations).
The primary is wounded with 3600 turns (0.05mm wire) and inductance is 120H (at f=300Hz and U~=100mV).  


Final assembly is really easy, the most pain is transformer and PSU.
 

Attachments

  • pic.jpg
    pic.jpg
    97.2 KB · Views: 23
Looks like the transformer inductance should be sufficient for decent bass response.

I don't know this tube, but maybe its output Z is a bit high for a 8:1 transformer. Maybe a lower plate resistor would help, 200k seems rather high, no?
 
Unweighted noise at the output of transformer is 4uV rms when input is shorted (by a 1uF capacitor) and 12uV rms for open input.
I will post more accurate data after measuring with PC.

See in attach 6S6B characteristics (I've measured a bunch of these tubes to know what they are). Grid voltage increment 0.5V, X-axis is plate voltage in V, Y-axis is plate current in mA. Actually, very linear tube.
 

Attachments

  • 6S6B.png
    6S6B.png
    11.2 KB · Views: 15
Hi Rossi,
Thank you for your thoughts.
The tube has about 12kOhm in this circuit. Actually, I used same plate resistor and voltage as in Neumann circuit with AC701 tube. 200kOhm resistor is used to get more DC current and to spend less AC current, ie to get more efficiency. I think I'll rewind transformer to 9:1 turn ratio (it will decrease output impendance and AC load at tube) and increase primary turns to 4000 (to increase L).  

Hi Stewart,
There is measuring circuit in attach. Microphone is turned on. At first you measure the voltage of unloaded generator (100mV~ for example). Then you measure the voltage with switch turned on. The voltage will drop. Output impendance can be calculated:

  Rout=R*k/(1-k),
R - sum of measuring resistor and output impendance of generator;
  k=U2*/U1,
    U1 - voltage when switch is off, U2 - voltage when switch is on.
 
 

Attachments

  • Rout.png
    Rout.png
    1.5 KB · Views: 43
Thanks Bezen,

I use something similar (well, an impedance bridge & scope) for dynamics and ribbon mics, but haven't measured any 'active' microphones with that approach.

I was wondering if you were putting a signal into the mic amp and then measuring the output somehow.

Nice work on the transformer.

Stewart
 
Hi Stewart,
When I measured output impendance, I shorted the input of headamp with 35pF capacitor (dummy capsule) and shielded with foil. I didn't tried to measure with signal applied to the input of headamp - it will change "impendance" for sure. I tried this method at different frequencies and levels: there is impendance  increaseing at 30Hz (caused by output capacitor and inductance of transformer) and at 100kHz (resonance of leakage inductance and capacitance of transformer). Level dependance: impendance increases at very high level applied to the circuit (I think it's due to tube cut-off).
 
THD specs (output is loaded by 1.2 kOhm, input signal is fed through 180pF capacitor, f=1kHz):
0.15% (25mV rms input);
0.5% (78mV rms input);
1.6% (245mV rms input).
It's mainly second harmonic, third harmonic is 20dB lower.
Gain is about 3dB.
Approximate headroom calculation:
If the capsule's sensitivity is 8mV/Pa then we have THD=0.5% at SPL=114dB and THD=1.6% at SPL=124dB.
Just compared with Neumann KM54 specs, looks comparable.
I think I will decrease sensitivity at 3-5dB with NFB (now it has too high output).
 
Back
Top