Help with diode ID?

[k brown]

I'd like to make some clones of the Audio Technica AT8533 power modules (sans transformers), but the schematics I had long ago (AT has successfully scrubbed the net of all schematics) don't give a part no. for the diode feeding voltage to the transistor's emitter, and the mic's FET.

Same biasing circuit was used in the ATM35's PM (AT8532), and the AT863R boundary mic.

These are all power modules for mics using the 2SK660 FET.

I was able to ID that part in the AT863R module (same circuit, but with line level output added); it's an SMD: 1SS226, described as an ultra fast switching diode. The 8533 diode is a glass through-hole part. Would a 4148 be fine here; does it need to be an 'ultra fast' part? I'm really not equipped to deal with SMD parts.

Circuit voltage to that diode is ~4v.

Thanks.

Here's my hand-written transcribe of the 8533:

 

[Khron]

The 4148 is "ultrafast" already. Plus that circuit is DC (or "worst case" audio range), so i wouldn't worry.

 

[k brown]

Thank so much - that was my hunch as well.

Have you ever seen a FET biased this way?

 

[kingkorg]

Not sure if it applies, but look at first minutes of the video. He modded the mic with a diode here.

 

[k brown]

Interesting possibility; can a 4148 function as a constant current source?

 

[abbey road d enfer]

Have you ever seen a FET biased this way?

What FET? Do you mean the capsule? Nothing unusual there. Power to the top, signal from the middle, bottom to ground.
Or do you mean the bipolar transistor?
I agree that the role of the diode is somewhat intriguing at first view. Looks like the constant-current hypothesis could be right.

 

[k brown]

This is the circuit of the AT8533 power module for their mics using the 2SK660 FET; the FET connects directly with the TB3M connector shown at the left of the schem. The drain is on the top pin, and source is on the bottom.

AT had and has a huge range of mics where the only electronics in the mic head are the K660 and a couple of RF caps between source and ground, and drain and ground.

So the FET is being biased by the 1k resistor and diode between the source and the drain - near as I can tell . . .

It's the only circuit like this that I've seen, and it sounds much better than their other power modules which simply bias the FET with a source resistor.

 

[abbey road d enfer]

Interesting possibility; can a 4148 function as a constant current source?

No. But there are constant-current diodes that look very much like a 4148.

 

[abbey road d enfer]

This is the circuit of the AT8533 power module for their mics using the 2SK660 FET; the FET connects directly with the TB3M connector shown at the left of the schem. The drain is on the top pin, and source is on the bottom.

AT had and has a huge range of mics where the only electronics in the mic head are the K660 and a couple of RF caps between source and ground, and drain and ground.

So the FET is being biased by the 1k resistor and diode between the source and the drain - near as I can tell . . .

Hard to tell, because on the schemo, the source is grounded...

It's the only circuit like this that I've seen, and it sounds much better than their other power modules which simply bias the FET with a source resistor.

OK, if we consider the source is not grounded, the capsule is buffered with an emitter-follower, i.e. without the gain associated with common-source topologies that are predominant. As a result, the overload margin is significantly higher and distortion lower. OTOH, the noise performance may be not so good, which does not matter much because the capsule noise is dominant in such a small diaphragm.

 

[k brown]

Perhaps I should have simplified my question to: does anyone know of a through-hole equivalent to the SMD 1SS226? Khron seemd to think a 4148 would be fine here.

 

[Gus]

http://www.eicsemi.com/datasheet/1SS226.pdf
note 1.2VDC forward voltage at 100mA you would need to do a test at lower current(the current the circuit works at) to measure the forward drop

 

[abbey road d enfer]

I believe the original type is closer to 1mA than 100.
http://www.datenblatt-pdf.com/data/ic.php?q=MX5283

 

[Khron]

http://www.eicsemi.com/datasheet/1SS226.pdf
note 1.2VDC forward voltage at 100mA you would need to do a test at lower current(the current the circuit works at) to measure the forward drop

I can't help wondering if that's per diode, or both in series. According to the graph on page 2 on that datasheet, it looks a lot like a normal diode, dropping 0.8-0.9V @ 100mA. Considering this mic is phantom-powered (right?), i kinda doubt that's much of an issue; just depends on whether they're using only one diode of the two that are in that package, or both in series.

 

[k brown]

The schematic for the PM using the 1SS226 shows the two in series.

Yes, the AT8533 and AT863R are phantom power only.

 

[abbey road d enfer]

The schematic for the PM using the 1SS226 shows the two in series.

Can you post this schemo?

 

[k brown]

I did.

The only diff between the 8533 and the 863R is the diode - and the fact that 863R is all SMD, and 8533 is through-hole.

 

[abbey road d enfer]

I did.

Where?

 

[k brown]

Image in first post.

The schem that uses the 1SS226 is exactly the same as the image I posted, except for the diode.

 

[abbey road d enfer]

Image in first post.

Sorry, this schemo has mistakes that make it incomprehensible.

 

[k brown]

Quite sure I copied it down accurately.

I actually have seen this way of biasing a FET before, and had forgotten about it. The Shure SM91 power module has the exact same topology, and differs only in the use of a resistor (R115) in place of the 'mystery diode'. If you unclutter and straighten out this schem, you'll see it's the same, but with different values.

Plus, I copied down three other AT PM schems independently of each other, working from the original AT docs, and they all look exactly like what I posted. Not likely I made mistakes every time.

Look Ma, no source resistor.