Err.rh! What makes you think its not "fully impedance matched"?Techie007 said:The most shocking thing for me was to discover that the NT1-A is actually not balanced or even fully impedance matched, contrary to much of the hype surrounding this microphone.
Because when I touch a wire to the two XLR signal lines (one at a time) between the decoupling resistors and capacitors to introduce noise, I hear noise on the active signal lead that I don't hear from the grounded signal lead.ricardo said:Err.rh! What makes you think its not "fully impedance matched"?
Looks like there's nothing magic here. The input FET is indeed a VLN type, with the associated high capacitance, but in an LDC it is not redhibitory. The FET input capacitance is bootstrapped by NFB, and the feedback capacitance is not a big issue since the voltage gain of the FET is quite low in this configuration, so I would say the combination of minimal attenuation and VLN FET explains almost everything.bkdog said:I'm trying to figure out how they get the noise so low...
This node is grounded via a 47u on the minus leg; on the other leg, it goes to the output of the headamp, which impedance is not zero. That may explain the difference, or not...Techie007 said:Because when I touch a wire to the two XLR signal lines (one at a time) between the decoupling resistors and capacitors to introduce noise, I hear noise on the active signal lead that I don't hear from the grounded signal lead.ricardo said:Err.rh! What makes you think its not "fully impedance matched"?
You'd only get that test of yours to pass with a transformer output.Techie007 said:Because when I touch a wire to the two XLR signal lines (one at a time) between the decoupling resistors and capacitors to introduce noise, I hear noise on the active signal lead that I don't hear from the grounded signal lead.ricardo said:Err.rh! What makes you think its not "fully impedance matched"?
Exactly! That's precisely what I meant by "not fully matched". It is nearly matched as the amplifier (when on) almost has an active zero ohm equivalent output impedance, but not quite. It rejects hum quite well, but not high-frequency noise. I think it was one of those "good enough" compromises that helped get this mic's self-noise so low. But it does help explain why I often have faint high-frequency electronic noise in recordings.abbey road d enfer said:This node is grounded via a 47u on the minus leg; on the other leg, it goes to the output of the headamp, which impedance is not zero. That may explain the difference, or not...Techie007 said:Because when I touch a wire to the two XLR signal lines (one at a time) between the decoupling resistors and capacitors to introduce noise, I hear noise on the active signal lead that I don't hear from the grounded signal lead.
Really? Wouldn't a balanced output transistor circuit (like what's in the NT2) exhibit an equal amount of noise when either signal lead is touched? This is what makes the XLR common-mode noise rejection work. Not that noise gets conducted to both leads by a transformer, but that both leads pick up the same amount and spectrum of noise due to fully matched impedance.Matt Nolan said:You'd only get that test of yours to pass with a transformer output.
An impedance balanced output works because interference is picked up in both legs at the same time due to their physical proximity (very low loop area). The impedances need to be tightly matched so that the induced voltages are closely matched so that they cancel well at the balanced input stage at the other end. Your test is a bit off the wall by injecting interference on just one of the legs. That's not how interference pick-up works.Techie007 said:Really? Wouldn't a balanced output transistor circuit (like what's in the NT2) exhibit an equal amount of noise when either signal lead is touched? This is what makes the XLR common-mode noise rejection work. Not that noise gets conducted to both leads by a transformer, but that both leads pick up the same amount and spectrum of noise due to fully matched impedance.Matt Nolan said:You'd only get that test of yours to pass with a transformer output.
That's the limitation of most of the impedance-balanced circuits. But even the reverred Schoeps circuit is not perfectly balanced, since the output transistors are driven by different nodes. But what makes it so good is that the impedance of each leg is very low (less than 10r); as a consequence, the absolute value of the unbalance is also very low. Both these factors contribute to the excellent CMRR. OTOH, the NT1 has 50r resistors in each leg; combined with the increasing HF impedance of the active leg, CMRR is undeniably not as good.Techie007 said:It is nearly matched as the amplifier (when on) almost has an active zero ohm equivalent output impedance, but not quite. It rejects hum quite well, but not high-frequency noise.
Don't forget that CMRR involves as much the receiver than the source. You may have an issue with your mic pre.But it does help explain why I often have faint high-frequency electronic noise in recordings.
With its higher open-loop gain, the AKG probably has a lower output impedance.moamps said:The similar design (at output) is used in AKG 414b-tl2. I used these mics several times without any problems.
http://cloud.akg.com/8694/c414_b_uls_tlii_service.pdf
Thank you. I'm pretty sure I don't have the center terminated capsule version. It's possible that my 10 megohm multimeter lowered the polarization voltage slightly during measurement.ln76d said:Great work!
I moded few of these - for center terminated capsule version - there's pretty good option with jumpering one diode in DC converter.
Voltage goes down around 10V and microphone sounds much better. For edge terminated - depends on revision.
BTW i always measured little bit more than 75V of polarisation voltage.
Also there's something different on schematic, but currently i can't check it.
There was no double 47nF capacitor and definately there was one 33nF and also 100pF.
Have you measured this?Techie007 said:The frequency response of this microphone does bother me. I've seen it described as a "bright microphone", but I certainly would not call it that. Perhaps it sounds that way to people missing their HF (>10kHz) hearing. What I'm seeing is 3–6 dB peaks around 4–7 kHz (which makes it sound harsh), with severe rolloff as the frequency goes up to 20kHz (which makes it sound dull or lack crispness even though it's harsh). It seems to vary from recording to recording.
This example is not valid IMO, because a distant recording is bound to be "duller" than a close one; it's just a matter of direct-to-reverberated sound ratio.Techie007 said:If you need an example of what I'm talking about, see here.
Techie007 said:With that recollection, my guess is that the capacitor on Q1 would be the 100pF one. I read several places of people improving certain microphones by replacing the surface mount capacitors with quality though-hole ones. Perhaps I should look into this? An added bonus, I'd find out the real sizes when removing the old SMD ones.
The frequency response of this microphone does bother me. I've seen it described as a "bright microphone", but I certainly would not call it that. Perhaps it sounds that way to people missing their HF (>10kHz) hearing. What I'm seeing is 3–6 dB peaks and valleys in the 2–8 kHz area (which makes it sound harsh), with severe rolloff as the frequency goes up to 20kHz (which makes it sound dull or lack crispness even though it's harsh). It seems to vary from recording to recording.
Won't shorting the last diode (in D3) just raise the bias voltage by 0.4V? That diode is really there to allow the last capacitor to filter the choppy booster output from power GND while allowing another capacitor to filter the now quiet output from audio GND without causing a ground loop between the two grounds. If boosting the bias voltage would help, I could try a slightly higher voltage zener diode. Replacing the 15V zener with a 17V one should raise the bias voltage from 70 to 80V. Can't really go higher than that since the 40106's absolute maximum voltage is 18V.
Based on several reports of significantly improved frequency response and "openness", I did try removing the thin wire netting inside the headbasket. Unfortunately, all this seems to have accomplished is making the microphone pick up a moderate 60 Hz hum. I was able to remove the netting without removing the headbasket, which remains fully attached and grounded as originally designed. The hum can be eliminated by placing the old netting over the back, sides and top of the headbasket, but it looks tacky. Speaking of which, the headbasket still looks like a serious acoustic limiter all the way around its lower half. I am wondering where I could buy a better headbasket inexpensively and solve the openness and new noise problem.
Changing the output (47uF) capacitors seriously makes a huge difference? What size/type do you like and how does it improve the sound from the original? I'm all ears about ways to improve this microphone short of spending $$$ (would rather put the money toward new mics and keep these). For instance, I've been thinking about Jfet options:
Anything that will improve its HF/transient response while also improving (or not seriously worsening) its sensitivity and noise floor would be great. Any thoughts or other Jfets out there I should be aware of?
- NJ450L - NT1-A original for reference; 0.9nV, 35pF
- 2SK170 - Very similar, 0.95nV, 30pF
- LSK170 - Possibly good alternative; 0.9nV, 20pF
- LSK189 - If gains from real low capacitance will overcome higher noise; 1.8nV, 4pF
- MX16 - Perhaps best alternative; 1.1nV, 4pF
Enter your email address to join: