New Discrete OpAmps

[recnsci]

Someone in here (sorry to have forgotten---was it Ed Anderson?) was kind enough to send me some samples of a Sanyo comp pair that might be suitable if you can get your hands on the high beta bracket ones.  Fast as blazes, and the constancy of beta with collector current (often described improperly as "linear beta"), until you run into current-crowding effects, is impressive.

The P/Ns are 2SC3596 and 2SA1402.  I'm not sure how hard they are to get.

http://www.bdent.com/search/part.jsp?partnum=2SC3596
http://www.bdent.com/search/part.jsp?partnum=2SA1402



BTW by "constancy of beta with collector current (often described improperly as "linear beta")" I guess you mean
Ic/Ib is not dependent of Ic, versus Ic/Ib being linear function of Ic. Somewhat analogous to parasitis C variation being
named "nonlinear" where it has liner dependency on voltage.

BTW2 complementary input gm doubling is effect of AB input stage. I've seen here and there debates about
"audiophility" of IC opamps with AB input. While AB input has much more drastic effect than AB out (which "just" induce
variation of Rout over one octave), my hunch is that AB input is much more benign because of nature of distortion (
spectrum rollof and phase of most significant harmonics).

cheerz
ypow

 

[Samuel Groner]

How about Toshiba 2SA970 and 2SC2240?

Good parts, but as for as I know their rbb' is higher than the parts shown, resulting in higher voltage noise.

Is Q8 included just to tame clipping behavior?

I'm not sure it does much for clean clipping; it surely limits the maximum collector current of Q7 during output current limiting. Usually the emitter resistor of this transistor is sufficient but at the high supply voltages we need a better plan for thermal reasons.

Why did you reduce the VAS emitter decoupling capacitor so much?

At least in simulation it prevents oscillation while current limiting. Didn't check the real world behaviour yet though.

Samuel

 

[bcarso]

BTW by "constancy of beta with collector current (often described improperly as "linear beta")" I guess you mean
Ic/Ib is not dependent of Ic, versus Ic/Ib being linear function of Ic. Somewhat analogous to parasitis C variation being
named "nonlinear" where it has liner dependency on voltage.


cheerz
ypow

Exactly---the beta is not dependent on Ic, whereas "linear beta" would mean beta would be a linear function of Ic.

I suppose if pushed one could say that the function "Beta = constant" is a linear function, of the form y = mx + b where m = 0.  But that's perverse and not what we usually mean when we talk of linear functions or linear operators.

And yes, some very good authors always throw in the qualifier "nonlinear" when describing voltage-dependent capacitance effects in things and their effect of generating distortion---tacitly implying that a linear variation in capacitance would not generate such distortion.

Well---it will be a different distortion, somewhat, but it will still be distortion!  ;D

It's true of course that for other than infinitesimal changes, most if not all semiconductor junctions do have a nonlinear variation in capacitance with voltage.  But it is misleading to throw in the "nonlinear" adjective as if it were a necessary condition for generating such distortion.


Brad

 

[Samuel Groner]

I decided that for the SGA-LNA-1 standard compensation is more appropriate (better stability at same or even slightly better distortion performance)--schematic updated. The first post now also includes suggested values for R6 if the SGA-SOA-2 is used at lower supply voltages.

Samuel

 

[okdaniel]

Looks like this thread has been inactive for a while. I've got a quick question about the SGA-SOA-2.

I bought a few DIY kits from diypartssupply.com for the SGA-SOA-2. The inductor in the schematic doesn't seem to be a pat of the PCB. What's the deal there? Is it now gone for a reason, or are my kits modified in some way?

Any advice here would be great. Trying to use these in some mic pres!


Update: I was wrong. The inductor is on  there! So tough to see.
Thanks!

 

[abbey road d enfer]

Looks like this thread has been inactive for a while. I've got a quick question about the SGA-SOA-2.

I bought a few DIY kits from diypartssupply.com for the SGA-SOA-2. The inductor in the schematic doesn't seem to be a pat of the PCB. What's the deal there? Is it now gone for a reason, or are my kits modified in some way?

Any advice here would be great. Trying to use these in some mic pres!

Thanks!

You may get a more documented answer from samuel, but the thing is, this DOA actually works without the inductor; the difference of performance may not be heard in most cases, so someone may have thought of making a quick penny by deleting it from the kit...

 

[Samuel Groner]

I have no affiliation with the sold kit (PCB design isn't mine either).

It works without inductor, but you should expect a penalty in voltage noise, DC offset and distortion. I can't give a good estimation without looking at it in more detail, but the degradation in voltage noise is probably about 10 dB, and that in distortion about 20 dB.

Likely you can just add the inductor at the bottom of the PCB.

Samuel

 

[okdaniel]

Thanks Samuel. I will do that. Looks like it's pretty doable.

I'm trying to figure out how to calculate the phase lag cap value also.

How do I choose a frequency value between 150k and 500k?

How do I take the value for C in the formula and convert that to what I'm assuming is a value in PF capacitance?

Thanks again for answering my first post.

 

[Samuel Groner]

How do I choose a frequency value between 150k and 500k?

No hard rule here; lower values give lower insensitivity to RF interference and in some cases better stability margins. Higher values give better audio-band frequency response.

The formula gives a capacitor value in Farad. One pF is 1e-12, so you multiply by 1e12 to go from F to pF. Most pocket calculators should easily do this, and you can use e.g. http://sim.okawa-denshi.jp/en/CRlowkeisan.htm to verify your calculation.

Samuel

 

[okdaniel]

Thanks very much once again.