U47EF Oliver Archut version

[rock soderstrom]

Ah ok I get it now

6.28 x 200000000 x 0.000072 = 90432
1/ 90432 = 0.00001106

Wrong

Its 0.000000000072F

fc=1/(6.28x200000000x0.000000000072)

 

[Spencerleehorton]

Story of my life!!!

 

[rock soderstrom]

Story of my life!!!

 

[Khron]

Wrong

Its 0.000000000072F

fc=1/(6.28x200000000x0.000000000072)

Major brainfart there, so very sorry Picofarads are 10^(-12), of course...

 

[rock soderstrom]

Major brainfart there, so very sorry Picofarads are 10^(-12), of course...

You are not alone, I did it wrong at first too....

 

[Roman Beilharz]

Alright, thanks again. Now I am up to speed and truly understood the issue and it's consequences. I hope I am not hijacking this thread and that the OP might find this useful as well: As I right now want to implement a LC/HP-filter into my EF47 mic circuit (I have a problem with a nearby industrial facility creating quite some noise around 25/30 Hz), I am wondering, whether I should

a) lower the value of the capsule resistor even further; say to 25M to get rid of the LF in the first place (e.g. by switching another resistor in parallel to my 100M one to reduce the overall resistance on demand)
b) insert a LC/HP filter right before the signal hits the transformer (still giving me some headache to calculate the I/O impedances correctly to find the necessary capacitor value)
c) do nothing and get rid of the LF-energy with the LC-filter of my preamp

Any opinions on that?

 

[Khron]

You could consider reducing the value of the anode-to-transformer capacitor somewhat, and also use the high-pass on the preamp.

 

[rock soderstrom]

Any opinions on that?

Yes, in my humble opinion, reducing the coupling capacitor from the anode to the output transformer would be a good place to limit the LF range.

You can calculate the lower cut-off frequency using the formula above or the online calculator.

The output impedance of the microphone multiplied by the square of the step down ratio of the transformer is the impedance seen by the tube and the coupling capacitor that forms the high pass.

Example: Haufe T14/1 has a step down ratio of 12:1, thus a multiplier of the impedance of 144. With the output impedance of 200 ohms, this gives you 28.8K ohms.

If you use a 0.2uF capacitor now, the cut-off frequency (lower -3dB point) would be at 28.42Hz.

I hope this is correct and was somewhat understandable.

Edit: this thread could be helpful too:
https://groupdiy.com/threads/ela-m251-c3-anode-to-ot-coupling-cap.84648/post-1099923

 

[rock soderstrom]

a) lower the value of the capsule resistor even further; say to 25M to get rid of the LF in the first place (e.g. by switching another resistor in parallel to my 100M one to reduce the overall resistance on demand)

Folks correct me if I'm wrong, but reducing the Rg is only possible within limits, as otherwise the very weak signal of the capsule is loaded down too much. The capsule prefers to play into a very high impedance.

 

[Roman Beilharz]

Alright, thank you so much, rock. You... well: rock

So with my Haufe BV08 I have a step down ratio of 6,5:1. The square of 6,5 would be 42,25, this value multiplied by the mic output impedance of 200 would then equal 8450 ohms "seen by the tube" and the coupling capacitor. As the latter right now is 1μF, I end up at a current corner frequency of 18.8 Hz. Now choosing a capacitor with - let's say - 0.47μF, I would accordingly end up at a cutoff frequency of about 40 Hz.

But then I was stuck with that very rolloff, there was no going back - no switchable solution. Did I get that right?

 

[rock soderstrom]

But then I was stuck with that very rolloff, there was no going back - no switchable solution. Did I get that right?

First, your maths is right from my point of view.

The general question is, what do you want to record and at what lower cut-off frequency does it become an audible limitation? Even a lower -3dB point of 40Hz is not really a problem from my point of view.

You could actually make it switchable by connecting two 0.47uF in parallel and switching one of them out depending on the situation and requirements.

I wouldn't necessarily do this in operation with the channel fader open, but otherwise I don't see any problems.

 

[Murdock]

I would also suggest to just try a few values.
I myself prefered 0.5uF over 1uF in my EF47 build (EF800, Moby BV08 and 60Meg grid resistor). And I think the resonance bump is at play here as to me it sounded fuller. There isn't much under 40Hz anyway with the stuff I record. So I thought it can't hurt.
Can't back that up with measurements though. But I tried 1uF two times and always came back to 0.5uF.

 

[Roman Beilharz]

Thank cou, Murdock. Excellent point and good to know others have tampered around with those issues as well. The more I think about it, the higher I would want the roll-off to kick in. I am recording solo vocals and stringed instruments 95% of my time, so I am thinking about throwing in a single 0.22μF by default, giving me a cutoff F of around 85Hz and then adding another 0.33μF with a dual switch - (dis-)connecting both ends of the additional capacitor, right? - giving me a total of 0.55μF and a practical lower rolloff of 34Hz.

Would you say these do the trick (160V instead of the 250V ones I currently use)? Or do you have any other recommendations?

 

[Spencerleehorton]

I’ve just realised I have used a different transformer in my first version!!
That’s why the output level is so different, I’ll also change the 200m resistor for 100m by putting another 200m in parallel and the other 200M resistor I can change to a 60M as I have a 50M and a 10M, I’ll also change the anode 1uf cap for a 0.5uf

 

[rock soderstrom]

I’ve just realised I have used a different transformer in my first version!!
That’s why the output level is so different, I’ll also change the 200m resistor for 100m by putting another 200m in parallel and the other 200M resistor I can change to a 60M as I have a 50M and a 10M, I’ll also change the anode 1uf cap for a 0.5uf

Keep us informed about your experiences with the changes.

 

[rock soderstrom]

Would you say these do the trick (160V instead of the 250V ones I currently use)? Or do you have any other recommendations?

That depends on your power supply unit. It may be that after switching on, when the heater of the tube is not yet warm, the 160V is exceeded, because the tube does not yet draw any current. Which PSU are you using? Schematic?

 

[Spencerleehorton]

Keep us informed about your experiences with the changes.

Yes I will, my version 1 sounds great and it’s got a U87 9.5:1 transformer in it, it uses 1G resistors and 1uf caps but I’m happy with how it sounds so I’m kind of loathed to change it.
So sounds so good on the drums, acoustic guitar and vocals.

With version 2 I can use it to experiment, if I can reproduce version 1 first of all then make a few tweaks to suit.

 

[rock soderstrom]

Yes I will, my version 1 sounds great and it’s got a U87 9.5:1 transformer in it, it uses 1G resistors and 1uf caps but I’m happy with how it sounds so I’m kind of loathed to change it.
So sounds so good on the drums, acoustic guitar and vocals.

If you like it, I would leave it as it is, for sure.

You are in a luxurious position, since you can continue to experiment with V2.

Edit: Interesting by the way that the EF800 plays so well with the 9.5:1 transformer. I will try that as well.

 

[Spencerleehorton]

If you can try it and put it waist height about 2 foot away from drums and see how it sounds.
Try it on vocals and acoustic guitar, I preferred it on omni on acoustic guitar cardioid on vocals.

 

[MagnetoSound]

Ah right well makes sense, didn’t think it was making any difference!!
I have four EF80 and they all are quite microphonic, I chose this one as it was the least.
Will change for the EF800 when it arrives.

Did you isolate the tube socket from the mic body? The best thing you can do to help with tube microphonics is to make sure the socket is kept well isolated with rubber bands or similar.

Even so, you’ll still need a decent non-rattly tube in there.