Thoughts on these two FET output circuits?

[k brown]

Scans of my handwritten copies of two Audio Technica mic power modules; the actual schems are not online anymore and I don't have AT spec sheets that contain them. There may be minor errors, but I was just curious to get general overall impressions of comparisons between the two.

One is the circuit inside the AT871R (a very similar one is used in the AT2035), the other is an old external power module, the AT8533, that was used for many models of AT mics. Both were designed for the 2SK660 FET.

The transistors marked 'Q' I was not able to identify (the 871 is all SMD) so don't know the pinouts.

The 871 has what seems to me (my understanding of electronics theory is quite limited) a rather unusual method of biasing the FET, as it needs no DC blocking cap; the AT8533's biasing is quite similar to that in the Shure SM90/91 power modules, but nothing else I'm aware of - FET source has no resistor to ground.

Just wondering if the 871's more complex/complimentary circuit is mostly about increased output current, increased linearity, more sophisticated biasing of the FET, better conditioning of the incoming DC, or some combination thereof.

Thanks for having a look.

 

[Ricardus]

Try to do double blind testing many times, and then draw your conclusions.

 

[k brown]

. . . meaning you see no reason one would sound better than the other?

 

[Ricardus]

I can't say, but bias often makes our brains believe things that aren't true.

 

[k brown]

What's that to do with any of the questions posed in the OP?

 

[Ricardus]

Because it's not even worth discussing if you can't tell them apart double blind. So try it.

 

[k brown]

Because it's not even worth discussing if you can't tell them apart double blind. So try it.

For the simple reason that not one of the questions posed in the OP would BE answered by a listening test.

My questions were technical ones; why would an EE design the more complex circuit; what is it designed to achieve that the simpler one can't? What are the functions of the additional parts? Seriously doubt AT just wanted to spend more money on addtn'l parts and assy time - circuit obviously has a purpose - what is it? These are valid questions, since they are designed to do the same basic thing - serve as microphone biasing/output circuits utilizing the exact same FET; so why the difference?

Never heard of techinical curiosity; desire to increase one's understanding? I thought that was part of what went on here.

 

[Jen]

The harmonic distribution of SE circuits is different than PP. The SE circuit has very different behavor when adding current to the load, as compared to when it is restricting it. the valve effect is assymeteric. Push Pull circuits were invented to make symmetrical current control, theoretically reducing assymeteric distortions. That being said, some people prefer the sound of that distortion. Careful listening to each of these may result in selection of a personal flavor preference, not nevessarily which is actually more accurate.

 

[k brown]

Thanks for that; makes sense. At least I know now the reason for designing such a circuit.

 

[JohnRoberts]

I don't like the sound of either one listening with my eyeballs...

The linearity of each depend on actual operating conditions, bias, signal level, etc.

That's why we melt solder and see how stuff sounds in use.

JR

 

[k brown]

Thanks.

Not sure what you mean by 'bias' - aren't both of these providing bias to the 2SK660?

Operating conditions are obvious: The inputs to both are from the specified FET in a condenser microphone, so mic level in. Output designed to be connected directly to a typical microphone preamp. What's the mystery?

 

[k brown]

Over on another forum I got this response:


"The 871 circuit is balanced input, dc coupled, the other is unbalanced capacitor coupled.

The 871 is the better circuit as it would have greater noise immunity on the FET side."

 

[Mr. Moscode]

Something doesn't make sense. In the 8533 the capsule has to run on 1 diode drop? Is the diode a led or zener? After the hi Z constant current diode everything is bootstrapped to the output. Are these accurate drawings. And what are the circles with the Q inside?

 

[k brown]

Something doesn't make sense. In the 8533 the capsule has to run on 1 diode drop? Is the diode a led or zener? After the hi Z constant current diode everything is bootstrapped to the output. Are these accurate drawings. And what are the circles with the Q inside?

The diode is not an LED or a zener; that much I remember from the published AT schems that I used to have printed copies of. The Shure SM90/91 boundary mic power modules had a similar FET-biasing scheme, but had a resistor (R115) in position of that diode. If memory serves, it was actually a double diode in a single package (like seen on the 871 schem).

As I said in the first post, the circles with 'Q' inside them are SMD transistors I couldn't identify.

I also said in the first post that the schems may have errors, because they're based on hand-drawn copies I made long ago of the AT printed schems that I no longer have and are not available online.

I also found an original thru-hole version of the 871; output parts are slightly different. My drawn schem shows the two push-pull Qs as 2SC2545 and 2SC1084. Q5 (input volt reg) and Q2 (feeding voltage to the FET) are also 2SC1084. The two summing Rs on the Q3 and Q4 are 10 ohm, instead of 200 ohm as on the SMD version.