U47EF Oliver Archut version

[Murdock]

Where does everyone get these cool green caps from and why are they all the rage in these builds? Not to derail the thread but just a quick interjection.

View attachment 112441

You can sometimes get them on eBay or electronic surplus stores.
ERO was a big manufacturer of capacitors. I think Vishay took over some of their lines so you can also get modern versions of these.

 

[MasterEvan07]

Because people enjoy fetishizing over old things? Especially in the audio world...

There's a catch-22 there. Some old stuff is legendary for a reason and some is very much not - my own firsthand experience has shown me that reality from both ends. The problem is the oftentimes astronomical costs of many of those old worlds items, parts, etc., and that many people never get the opportunity to know for themselves their value of lack thereof, so they go by myth and word-of-mouth.

I'm happy to use my ears and know for myself and I'm also happy to get perspective from others and to test it out.

 

[Roman Beilharz]

You can lower R3 (1G) if you haven’t already. A value of 60meg should make a noticeable difference. But don‘t go too low. I think 30meg is the lowest I would go.
Also between 1uf and 0.33uF are some more values you can try. Maybe 0.5 or 0.68. or a combination of lowering R3 and the output cap.

Or you can lower the cathode cap C4 (47uf) to something between 10uf and 22uf.

Alright, thank you, Murdock. I've just stacked another lemon drop 0.33uF cap in parallel to the one I had and while my bass boost is basically gone then, the resonance notches still have higher/lower amplitudes than with the 1uF ERO cap in place. So 0.66uF doesn't seem to get me where I wanted.

I'll try to lower the 1G resistor next, but I assume I need to stick with 1uF and be happy with the mic as it is.

 

[metalockpick]

When the capacitor value is much too large, the sub bass can saturate the transformer. Then, when you reduce the capacitor size (and eliminate sub bass saturation), you often find more usable bass response (which has happened here).
Transformer saturation tends to garble all frequencies above the ones saturating the core and should be avoided if clean sound reproduction is your goal.
You can try further reduction of the capacitor, try 0.1uF or .22uF.
Some complain about increase of group delay @ very low frequencies when you do this, but a) It's happening at subsonic frequencies (where everything ((especially speakers)) moves slowly), and b) The difference is audible, your ears will tell you what you prefer; go with that.

The resistor value change does something similar; doing both will give you a maximum roll off of -12 dB/Oct if they are both tuned to the same frequency.
You don't need to eliminate (and you won't) eliminate the very low frequencies, you just need to tame them such that they don't cause audible issues.

 

[rock soderstrom]

I was partly aware of that issue, but expected to valuewise be able to find some sweet spot lower than 1uF, where the resonances would not be severe, still giving me at least a minimum of a high pass filter behaviour.

Me too, I am surprised by the amount of resonance. Try smaller C3, like 0.1uf

What type/brand of output transformer do you use? AMI?

 

[rock soderstrom]

Because people enjoy fetishizing over old things? Especially in the audio world...

This happens, no doubt about it. Here the situation is somewhat different, the ERO MKT 1813 are relatively small and flat by design and at least in my area very well available and inexpensive. Soundwise they do nothing wrong so I like to use them. Its s good looking cap, too

 

[terry setter]

Reduce the size of C1 significantly and you'll see classic first-order high-pass filtering. Substantially reducing R3 will change more than just your low frequency response.

 

[rock soderstrom]

Substantially reducing R3 will change more than just your low frequency response.

Terry, can you please elaborate a bit more on this?
I think others here are also interested in this.

 

[Roman Beilharz]

Me too, I am surprised by the amount of resonance. Try smaller C3, like 0.1uf

What type/brand of output transformer do you use? AMI?

I have 0.22uF at hand and could create a 0.165uF using two of my 0.33uF caps. Will give this one last try, but still, if this applies (quoting Transformers for Audio, see link above)

"By adding the capacitor in series with the transformer, a series resonant circuit is created that boosts the output at low frequencies. You can add series resistance to tame the large peak, but a far better solution is to use a much bigger capacitor. There can be no doubt that blocking DC from the transformer will cause far less distortion than any capacitor."

I am quite reluctant to follow the path of further reducing the transformer cap value, afraid of maybe not so obvious side effects (like I can't measure THD properly and would not want to induce too much of it).

This being said, a ginormous Mundorf Supreme Classic 1uF cap has just arrived that I would wanted to try for the hell of it plus I would want to give even larger values a shot as well. While I am at it.

I use a Haufe BV08 transformer wound to the amazing Andreas Grosser's specs (RIP).

Also interested in Terry's observations regarding the changes a reduction of R3 might spark off.... could you please shed some more light onto this, Terry?! Thank you very much.

My best

Ro

 

[Roman Beilharz]

Regarding the "huge" resonance: I am not entirely sure about that. I can't see any additional sub bass smearing in the time domain that has not been there already. To me it simply looks and sounds like the opposite of what I wanted to achieve: a first order bass boost instead of a first order low cut.

And don't get distracted by the heavy bumps in the FR of the cardioid measurements above; I guess there could be a mic body resonance going on or group delay phase issues at that given nearfield position of my 2-way speaker in addition to the proximity effect. In omni the 60Hz bump isn't there at all. You just see the bass response coming up to +6dB @ 20Hz (pink line is 0.33uF, blue line is 1uF).




Either way: The corner frequency maths - which seemed theoretically correct and should have given me a -3dB point of about 57Hz - practically clearly didn't apply here. Hence the headaches. Did I get something totally wrong here?!

 

[rock soderstrom]

Either way: The corner frequency maths - which seemed theoretically correct and should have given me a -3dB point of about 57Hz - practically clearly didn't apply here. Hence the headaches. Did I get something totally wrong here?!

It's hard to say, the measurement graph look different now. I am not an expert on this and I am curious about further contributions from other members.

Overall, the thread remains very interesting. Keep us posted with your experiments Roman!

 

[Roman Beilharz]

It's hard to say, the measurement graph look different now. I am not an expert on this and I am curious about further contributions from other members.

Overall, the thread remains very interesting. Keep us posted with your experiments Roman!

Yeah, I will. Well the graph looks different, because the first graph was the comparison measurement in cardioid. This is the same comparison in omni.

 

[rock soderstrom]

Well the graph looks different, because the first graph was the comparison measurement in cardioid. This is the same comparison in omni.

Yeah, I know.There are other areas in the FR that are odd for me. What is relatively consistent is the significant peak around 78Hz.

I am curious to see how the experts interpret this.

 

[Roman Beilharz]

Alright, as 0.22uF (orange) shows about the same bass boost as 0.33uF (pink), so I'd rather stick with the original 1uF cap (blue).

 

[ElSmurf]

Guys, I am really confused now. I've replaced my 1uF-C3 (the capacitor to pass on the signal to the transformer) by one of the 0.33uF foils above

Been following this thread since I'm planning on changing my C2 from 1uF to 0.47uF, and lower R1 and R2 from 1G to 100M. Ordered the parts, should be here in a week. These component positions are from the Nuemann schematic, which Dany's schematic follows.

I'm confused about your statement here, because on your schematic C3 isn't the equivalent of C2 on the Neumann schematic. On yours that is C5.

Also, if you haven't changed the resistor (and using the online calculator) the cutoff frequency would still be in the sub-1Hz region. My suspicion, based on my very limited understanding of electronics, is that your bass boost is resonance above the cutoff frequency that is moving into the audible range because you slightly raised the cutoff frequency. Anyone who is more knowledgeable, feel free to correct me, since I suspect something similar is happening in my D-EF47 using Dany's values for C2, R1 and R2.

 

[Roman Beilharz]

Been following this thread since I'm planning on changing my C2 from 1uF to 0.47uF, and lower R1 and R2 from 1G to 100M. Ordered the parts, should be here in a week. These component positions are from the Nuemann schematic, which Dany's schematic follows.

I'm confused about your statement here, because on your schematic C3 isn't the equivalent of C2 on the Neumann schematic. On yours that is C5.

Also, if you haven't changed the resistor (and using the online calculator) the cutoff frequency would still be in the sub-1Hz region. My suspicion, based on my very limited understanding of electronics, is that your bass boost is resonance above the cutoff frequency that is moving into the audible range because you slightly raised the cutoff frequency. Anyone who is more knowledgeable, feel free to correct me, since I suspect something similar is happening in my D-EF47 using Dany's values for C2, R1 and R2.

Oh, I am really sorry. Yes, I was talking about C5. Will edit my post, thank you for the hint.

Well, following rock soderstrom's calculation here
U47EF Oliver Archut version

I hoped for a corner frequency of 57Hz using a 33uF cap. What's your calculation arriving at sub-1Hz?

 

[ElSmurf]

Oh, I am really sorry. Yes, I was talking about C5. Will edit my post, thank you for the hint.

Well, following rock soderstrom's calculation here
U47EF Oliver Archut version

I hoped for a corner frequency of 57Hz using a 33uF cap. What's your calculation arriving at sub-1Hz?

Looks like the C2/C3/C5 thing got me confused as well - that 57Hz should be correct.

Have you swapped out the 1G resistors yet?

 

[Roman Beilharz]

Looks like the C2/C3/C5 thing got me confused as well - that 57Hz should be correct.

Have you swapped out the 1G resistors yet?

No, and I am not at all keen to do so, because then I would have to solder out the relay first. I think I am done here for now.

I have put in that expensive Mundorf/Germany 1uF cap and hence I am pretty much back at the bass response I had before, but the trebles sound a lot nicer to me now and so I'll simply use my preamp's LowCut and be a happy man.

Just compare these audio demos (all 3 cardioid ones will do for this) in your DAW, skip directly to the very last section with the Djembe recording and look out for presence and fast transients. Night and day, at least when listened to through very good speakers or quick (magneto/electrostatic?) headphones.

Then compare the female vocal (first section after the sine) and maybe the Viola (second section after the sine wave) of all 3 omni recordings. Also pretty impressive, me thinks.

AUDIO 01
This is where my journey had started: Mic&Mod stock configuration with Replica Microphones capsule and transformer plus JJ EF86 tube.
Mic&Mod stock AUDIO cardioid
Mic&Mod stock AUDIO omni

AUDIO 02
Now the same circuit with a Thiersch M7 capsule, Haufe BV08 transformer and Telefunken NOS EF86 tube.
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 AUDIO cardioid
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 AUDIO omni

AUDIO 03
This is the current state: Thiersch M7 capsule, Haufe BV08 transformer, Telefunken NOS EF86 tube, Mundorf 1uF cap and I am using a homemade cable resembling the VOVOX scheme for signal (all voltage wires well shielded in a SommerCable core, double higher gauge ground, +/- signal twisted massive copper lines on top, unshielded, wrapped in techflex, massive ferrite beads on both ends for RF/HF protection).
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 MundorfCap RVX-cable AUDIO cardioid
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 MundorfCap RVX-cable AUDIO omni

 

[MasterEvan07]

No, and I am not at all keen to do so, because then I would have to solder out the relay first. I think I am done here for now.

I have put in that expensive Mundorf/Germany 1uF cap and hence I am pretty much back at the bass response I had before, but the trebles sound a lot nicer to me now and so I'll simply use my preamp's LowCut and be a happy man.

Just compare these audio demos (all 3 cardioid ones will do for this) in your DAW, skip directly to the very last passage with the Djembe recording and look out for presence and fast transients. Night and day, at least when listened to through very good speakers or quick (magneto/electrostatic?) headphones.

Then compare the female vocal (first passage after the sine) and maybe the Viola (second passage after the sine wave) of all 3 omni recordings. Also pretty impressive, me thinks.

AUDIO 01
This is where my journey had started: Mic&Mod stock configuration with Replica Microphones capsule and transformer plus JJ EF86 tube.
Mic&Mod stock AUDIO cardioid
Mic&Mod stock AUDIO omni

AUDIO 02
Now the same circuit with a Thiersch M7 capsule, Haufe BV08 transformer and Telefunken NOS EF86 tube.
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 AUDIO cardioid
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 AUDIO omni

AUDIO 03
This is the current state: Thiersch M7 capsule, Haufe BV08 transformer, Telefunken NOS EF86 tube, Mundorf 1uF cap and I am using a homemade cable resembling the VOVOX scheme for signal (all voltage wires well shielded in a SommerCable core, double higher gauge ground, +/- signal twisted massive copper lines on top, unshielded, wrapped in techflex, massive ferrite beads on both ends for RF/HF protection).
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 MundorfCap RVX-cable AUDIO cardioid
Thiersch M7+Haufe BV08+TelefunkenNOS-EF86 MundorfCap RVX-cable AUDIO omni

View attachment 112475

Is the Mic&Mod platform/U47 essentially the Archut-designed circuit?

 

[Roman Beilharz]

Is the Mic&Mod platform/U47 essentially the Archut-designed circuit?

Well, I'd say the circuit is inspired by it.