LM10 questions

[thermionix]

I have one I want to try on the breadboard, maybe Dist+ or similar.  If I don't use the internal Vref, do I have to do anything with the unused pins?  I'd assume not, but not sure.

If I do use the internal Vref, do I just throw a 1Meg from pin 1 to pin 3?  What do I do with pins 8 and 5?  I have no idea what "reference feedback" or "balance" are.

I tried looking it up before asking here, couldn't find anything I could understand.

[Rob Strand]

Nat Semi App Note AN211 is like a cookbook for that chip.   It also explains the pins,
https://datasheet.datasheetarchive.com/originals/library/Datasheet-03/DSA0043038.pdf#search=AN211+national

The chip is like two opamps except one opamp has a DC reference permanently connected to its +input.

If I do use the internal Vref, do I just throw a 1Meg from pin 1 to pin 3?  What do I do with pins 8 and 5?  I have no idea what "reference feedback" or "balance" are.

Balance is like an "Offset Null" pin.

The DC amplifier/opamp has a 200mV DC reference permanently connected to the +input.  The Reference Feedback pin is the -input for that opamp.  If you connect resistors to the Reference Feedback pin it sets the DC gain.  This works exactly like gain resistors for a non-inverting amplifier.  The DC gain then multiples the 200mV DC reference so the DC output of that opamp can be set by the user (as a DC reference voltage).

For the AC amplifier you might want to check the common-mode range.  If it's a funky value you might have to set the DC bias point to a voltage other than Vcc/2.  I just can't remember the details.

[thermionix]

Oh my bad, I thought it was a single opamp and the Vref was separate.  Didn't know why it was depicted with a triangle symbol.  I'll try to see if I can wrap my head around the datasheet info, thanks for that.

For the AC amplifier you might want to check the common-mode range.  If it's a funky value you might have to set the DC bias point to a voltage other than Vcc/2.  I just can't remember the details.

I think I understand you here, some opamps can't swing to equal distance from either rail.  I'm not wording that right.

So, in the case of a Dist+ if I just used the regular no-Vref opamp, would I have to do anything with the unused pins?

[Rob Strand]

I think I understand you here, some opamps can't swing to equal distance from either rail.  I'm not wording that right.

It's actually inputs in this case.  I found the statement about it on the first page of  AN211, last paragraph.  Common-mode range includes -V.  Hmmm nothing about the positive side.***

So, in the case of a Dist+ if I just used the regular no-Vref opamp, would I have to do anything with the unused pins?

For the DC amp: I'd be inclined to connect pin 1 to pin 8 as it is a known good configuration used in many of the example circuits.
For the AC amp: I'd leave pin 5 open.

[EDIT1: ***
OK it's in here.  V+-0.85V. So the inputs can go from V- to V+-0.85V.   In other words you can do what you want.
http://www.ti.com/lit/ds/symlink/lm10.pdf
]

[thermionix]

Thanks again!  Sorry I'm still quite a noob with this stuff.  The datasheet mentions performance similar to an LM108, but internally compensated.  Maybe it would sound cool in a Rat.

[Rob Strand]

Thanks again!  Sorry I'm still quite a noob with this stuff.

No problem.  It's a weird looking chip until you decipher it.

It occurred to me you could do something sneaky and use the DC amplifier for AC.
You would wire it up as an inverting amplifier.
However, you need to use the same trick that's used on the NE570/NE571 chip to get the DC output biased right.   Check out fig 2 of,
http://experimentalistsanonymous.com/diy/Datasheets/SA570.pdf

You will need to add 2xRDC, CDC and R4.  You calculate the values using equation for VOUTDC to the left of that figure.  However for the LM10 you use 0.2V for Vref not the 1.8V used for the NE570.

PS: I don't know how noisy that chip is.  Especially the DC amp.
---------------
EDIT: Looks like using the DC amp as an AC amp has been done.  See figure 5 in,
http://www.ti.com/lit/an/snva530/snva530.pdf

The biasing here is simpler from what I described but you can't vary the feedback resistor without upsetting the DC ie. it's best for fixed gain.

[PRR]

> you could do something sneaky and use the DC amplifier for AC.

Nat Semi App Note AN211 Figure 38: Microphone Amplifier

> has been done.

1977.

The "ref" is rigged for gain of 100. Vol-pot, then opamp rigged for gain of 10.

[PRR]

> What do I do with pin... 5?

Look at the guts. If you leave pin 5 open, Q1 and Q2 have equal loads, which is probably what you want.

If you had VERY strict DC goals, you might want to dial-out the few mV DC error. The Balance pin can upset the input balance to offset the error. Knowing who designed it, it may do exceptionally well. However we are Audio engineers, and mV of DC is beneath our notice. Ignore (leave open) pin 5.

[Rob Strand]

Nat Semi App Note AN211 Figure 38: Microphone Amplifier

I posted the link and didn't realized it was in there!

The "ref" is rigged for gain of 100. Vol-pot, then opamp rigged for gain of 10.

Your left to right schematic is *far* easier to understand. 
That's the way schematics should be.

[Rob Strand]

If you had VERY strict DC goals, you might want to dial-out the few mV DC error. The Balance pin can upset the input balance to offset the error. Knowing who designed it, it may do exceptionally well.

The LM10 err's towards instrumentation for sure.   Many people say it's one of Bob Widar's masterpieces.  The reference circuit is very cool, and kind of unique as it's 200mV not a 1.25V band-gap.

[thermionix]

I saw the microphone amplifier, the only audio circuit example.

However for starters at least, I'm wanting to try just the AC amp, as if it were a single like a 741.

For the DC amp: I'd be inclined to connect pin 1 to pin 8 as it is a known good configuration used in many of the example circuits.

Re-reading this, I've re-confused myself a bit.  Do you mean connect 1 and 8 only if I'm actually using the DC amp?  I can just ignore them if I'm only using the AC amp?  Or I should have them connected regardless?

I just don't want to damage the chip, want to make sure I hook it up correctly before applying power.

[Rob Strand]

Re-reading this, I've re-confused myself a bit.  Do you mean connect 1 and 8 only if I'm actually using the DC amp? 

When not using the DC amp, connect pins 1 and 8, then just leave pins 1 and 8 alone.
What that does is configure the DC amp to have a 200mV output but we don't
do anything with it.

To use the DC amp as a DC reference you would use a circuit like the bottom right hand corner of
figure 16 in AN211.  The extra components are R5 and R6 which sets the DC output voltage to
something other than 200mV.

[thermionix]

Thanks once again Rob!

When not using the DC amp, connect pins 1 and 8, then just leave pins 1 and 8 alone.

That's what I initially thought you meant, but like I said I re-confused myself.

To use the DC amp as a DC reference you would use a circuit like the bottom right hand corner of figure 16 in AN211.  The extra components are R5 and R6 which sets the DC output voltage to something other than 200mV.

Baby steps, man.  I suppose it would be a waste of an LM10 to only use the one "regular" opamp, but for now I'm just wanting to mess around on the breadboard.  I fully expect that in a Dist+ circuit it will sound pretty much the same as a 741.

But that's tomorrow, for now it's quite late and I'm going to bed.

[thermionix]

Hey it works!  I actually followed the GGG schematic for the Dist+ as opposed to the one I posted above, with a 500k gain pot and 50k volume.  I used Si clippers as I don't like germs in this circuit.  I also used all film caps.

I think I like the LM10 better than the 741, it sounds only slightly different, maybe a touch clearer and shaves off a bit of the highest highs.  I also tried a TL071, which was the worst of the three, too much high end, thinner overall tone.

[Rob Strand]

Hey it works!  I actually followed the GGG schematic for the Dist+ as opposed to the one I posted above, with a
500k gain pot and 50k volume.
I think I like the LM10 better than the 741, it sounds only slightly different,

Cool.   I could only speculate why it sounds better.

[thermionix]

I can't guarantee that the difference I heard between the LM10 and 741 wasn't just a result of my ears being in a different position in relation to the speakers, it took a couple minutes to swap the IC on the breadboard because I have resistors straddling it.  But the LM10 does sound very good in this circuit.  The TL071 on the other hand clearly had too much high end, for my taste.  All told, I think switching the output cap from electro to film did more for the sound than the IC swap.