dima_k85

A little less than one year ago I wrote my first article about modding the Chinese BM-700 microphone.
The original article is in Russian, but you can use Google Translate to read it.

Since then, I have accumulated some experience in design and decided to make a new revision of CM-1A project.
I used a more powerful software tool for analyzing the original RODE NT1-A circuit – Micro-Cap 12.
It helped to clarify some of the component values and their operation modes.

As a result, I got the following functional model of the microphone amplifier (in the attachment).
« Last Edit: May 04, 2020, 12:32:01 PM by dima_k85 »


dima_k85

Bode diagram and spectral noise density at the output (in the attachment).
Amplifier specifications have become very good.

PCB coming soon…
« Last Edit: April 12, 2020, 01:47:37 PM by dima_k85 »

Khron

Right, so what does this do, that the venerable old Schoeps circuit doesn't (while using fewer components)?
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

RuudNL

Why do something simple, when it also can be done complicated...
There is a solution for every problem!

http://www.vansteenisaudio.nl

dima_k85

the venerable old Schoeps circuit
Do you mean this wretched circuit?

Khron

Exactly. What's so "wretched" about it, though, if i may ask?
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

dima_k85

Exactly. What's so "wretched" about it, though, if i may ask?
Oh, if you did not listen to Doug Ford six years ago, then I can make a model of this "venerable" circuit and tell in much more detail.
https://youtu.be/e5xenXTwAzo?t=452

I will replace the T1 JFET with 2N3819 and T2, T3 transistors  with 2N4403, do you mind?
« Last Edit: April 13, 2020, 05:23:09 PM by dima_k85 »

Khron

I've seen / watched that series of videos several times.  So your argument is (maaaaybe) 3dB more signal level? Poooossibly a bit more headroom on the output?

I could very well be wrong, but most of the time, those might not be issues you run into, with condensers. More often than not, they'll be more sensitive than you need, hence the attenuation switches you often find in the non-bottom-of-the-barrel ones.

If anything, it would be interesting (if possible) to measure some waveforms at various SPLs both at the mic output, as well as, say, right off the JFET source in a Schoeps circuit, and see if and where the signal starts clipping.
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

dima_k85

Schoeps model (in the attachment).
« Last Edit: May 04, 2020, 12:27:11 PM by dima_k85 »

dima_k85

Transient analysis (in the attachment).
Only 446mV * 2 = 892mV out amplitude. And this is with a fully twisted potentiometer P1!


dima_k85

The equality of voltages in the arms of the output transistors can be ensured only at less than 500 mV amplitude (see the attachment).
« Last Edit: April 13, 2020, 05:59:23 PM by dima_k85 »

Khron

"Fully twisted P1" doesn't really mean anything, in itself - you usually want to bias the gate in order to obtain maximum unclipped signal amplitude (regardless of where the potentiometer cursor happens to be).

Not to mention, simulations live and die by the accuracy of the models. I've just run a few LTspice sims on a Schoeps circuit, and with an optimally(?) biased J305 JFET, it's showing me 0.37% THD with a 1.8Vpp input signal, resulting in a 3.1Vpp output signal.

And when was the last time a condenser mic was "too quiet"? Keeping in mind that on the receiving end there's a preamp supplied at 24-30V (or +/-15V), capable of at least 50dB of gain...

By the way - that screenshot does indeed show that the output signal is severely clipped. But is that the fault of the output PNPs, or is the signal getting clipped by the FET itself, earlier in the signal path?
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

dima_k85

Let's look at the Bode diagram and the output noise for Schoeps (in the attachment).
The circled area is not valid for high quality amplifiers.

Khron

That, too, could be a simulation artifact (model parameters, etc). My version, with my models, in LTspice, shows no peaking whatsoever, just flat gain up to 17KHz, where it very gently starts rolling off to a -3dB of 180kHz or so.

I'm sure if i iterated with various JFET / PNP models and biasing conditions, i could make the Bode plot peak somewhere, but... How about testing that with a real circuit? Otherwise this all is just an arguably pointless exercise :)

Sure, "in theory", lots of things may or may not happen, but that's of no use if it doesn't agree with reality, right?

I'm not saying the Schoeps circuit doesn't have limitations - of course it does, all circuits do. I'm only asking, how likely are you to hit those limitations, in real-world (and/or "normal") use? :) Is it infamous for low headroom, by any chance?

Let's look at the Bode diagram and the output noise for Schoeps (in the attachment).
The circled area is not valid for high quality amplifiers.
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

dima_k85

Khron,  your arguments "doesn't really mean anything".
You are trying to justify one of the most miserable circuit, but apart from the "fairy tales" about the SPICE models you have absolutely nothing.
You can see everything for yourself, just make a model, I have given you all the data for this.

Stop spamming my topic! Want to talk about Schoeps circuit - create a topic about it.
« Last Edit: April 13, 2020, 06:28:48 PM by dima_k85 »

Khron

Fine. Here's a screenshot of the AC sweep.

PS: Microphone preamps have relatively low input impedances, usually up to 5Kohm or so, so that 47k in your version might be inaccurate / unrealistic. There's even a chance that might be part of the reason for that HF gain peak in your sim, but that's just speculating on my part.
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

Khron

And here's a transient sim, with the THD result of the Fourier transform. It's a bit higher at 0.43% with 2N4403 PNPs; the previous one was with 2SA970 PNP models.
Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

rogs

How about testing that with a real circuit? Otherwise this all is just an arguably pointless exercise :)

'Real circuits' rarely seem to trouble the super keen 'I've made a sim' brigade....

Describing the Schoeps circuit as ' a wretched circuit' is a comment I've not read before though? .. a very unusual take on a classic circuit!

It takes all sorts of course   :)
« Last Edit: April 14, 2020, 07:37:45 AM by rogs »
www.amx.jp137.com - A DIY RF condenser mic project

RuudNL

My experience is that the Schoeps circuit is a reliable circuit, that has low distortion when the FET is properly biased (on a 'real world' circuit I usually get distortion figures of 0.02%, even at high output levels) and can have very low self noise.
There is a solution for every problem!

http://www.vansteenisaudio.nl

analogguru

Now let's talk about this "wretched" impedance "balanced" CM-1A circuit which is only "balanced" as long as you use it on an low impedance transformer balanced input.  If you use this circuit on a modern electronically balanced microphone preamp (which are the majority today) there will not flow any current through the "balancing" resistor (R15) and therefore the microphone will be in fact unbalanced.  This is ok if you use the microphone with a 2m short cable on your desk, but not if you run the microphone over a 50m long cable in an environment with dimmers and let's see which circuit behaves better.  The "wretched" CM-1A circuit or the Schoeps circuit ?

Oh, you can't find a dimmer pack in your simulation program ?  How sad...... but that's real life....
« Last Edit: April 14, 2020, 08:30:20 AM by analogguru »


 

Related Topics

  Subject / Started by Replies Last post
9 Replies
4944 Views
Last post August 28, 2006, 11:54:07 AM
by Svart
15 Replies
9782 Views
Last post January 23, 2011, 09:43:27 AM
by letterbeacon
2 Replies
3017 Views
Last post January 03, 2013, 07:41:57 PM
by ricardo
Rode nt1 MOD

Started by Sinkia « 1 2 3 » Microphones

44 Replies
4348 Views
Last post May 09, 2019, 06:31:09 PM
by Sinkia