National LM 1972

mlloyd1

Member
Paid Member
2001-02-25 7:10 pm
Northern Iliinois
I have some of this but haven't yet used them (too many other projects - Spring and summer are too short!).

I wanted to build a few line stages with different mechanisms and parts (LM1972, CS3310, AD7112, discrete resistors and relays, etc.) to do the gain adjustment so I could compare and determine what I want to use in my next preamp design.

I have heard of no commercially available products that use the National part, but I have seen a few products that use the competitive part by Cirrus - the CS3310.

What were you planning to use for control: a micro controller like a PIC or AVR?

Michael
[email protected]
 

mlloyd1

Member
Paid Member
2001-02-25 7:10 pm
Northern Iliinois
Janne:

According to the data sheet, what you propose would work just fine. However, what the datasheet doesn't always tell you clearly is that the distortion will drastically increase with the lower impedance loads and also at higher frequencies. Using a buffer of some sort after the op amp helps minimize the effect. I use a buffer after all my op amp circuits and have NEVER gone back.

Michael
 

paulb

Member
2001-06-01 4:53 pm
Calgary
I've used an Analog Devices VCA circuit (I believe it was the SSM2018) with disappointing results. I get pops from AC line noise. When I traced it with a scope, I found that the VCA was actually amplifying the noise on its power supply rails to its output. Cleaning up the rails made it acceptable, but I'm now very leery of using electronic circuitry for a potentiometer.
 
hmmm.. yeah, i'm very leery of it as well. On the upside, the newer devices (in particular the CS3310) actually implement a passive resistor ladder on the silicon, so in this regard it's just like a stepped attenuator. The questions in my mind are:

1. how good are these polysilicon resistors compared to say SMD metal film types?
2. how good are the FET switches they're using to select the tap point on the resistor ladder? ...are they linear enough?
3. how good is the output buffer?

I would imagine that the resistors are ok, if they can be sufficiently isolated from noise on the substrate. And, of course Crystal should be perfectly capable of making a sonically transparent output buffer (even if it takes some external circuitry to ensure this). The killer is these FET switches. They are basically the same thing used in IC analog MUXes... has anyone used an IC MUX for audio before? Opinions on their sound quality???
 

mlloyd1

Member
Paid Member
2001-02-25 7:10 pm
Northern Iliinois
For some useful commentary on the CS3310, see this link:
http://www.jeffrowland.com/dacs.htm

PaulB mentioned the SSM2018 VCA. I would never use a VCA as a gain control element in my preamp. Yes, I know about Dolby and the highly regarded Meitner preamp (of the 80's?). I just wouldn't use 'em in anything besides a homebrew analog synth. At least for now - THAT Corp has produced some parts that might be worth looking into.

I'm probably going to get shot even more for this, but I think FET switches are OK to use if carefully applied (correct layout, appropriate topology minimizing voltage swing across the switch, etc.). Some of the newer parts from Maxim and Analog Devices have VERY impressive specs on distortion, resistance variation, charge injection, etc. I'm thinking MOSFET switches suffer from some of the same early misapplication and bad press that OpAmps went through in the early days. Back in the day, LM741 parts were being used in some preamps, the preamps measured poorly and sounded bad. Consequently, ALL preamps using OpAmps were labelled as bad sounding. Slowly, Rowland, Levinson and others started properly using some of the "audio appropriate" devices, making some good sounding gear, but no reviewers ever really came back and said "OK, we forgive OpAmps".

My 2cents. Your mileage may vary ....

Michael
 
I have to agree with your comments about opamps... when I first started getting serious about hifi, I was indeed skeptical of opamp sound, as I'd read all the bad press. But, the preamp I currently use is the best one I've built yet, and it's based on the OPA134 (with external output transistors inside the feedback loop).

...yes, I even think it's better than the SRPP tube buffer I experimented with, and definitely leaps and bounds ahead of the BUF-03FJ of Corey Greenberg's preamp, or any of the discrete MOSFET topologies I've tried. One thing I really like about solid-state: DC coupling - it can really make a big difference!

Back to volume controls though...

I think I'll just go ahead and order some CS3310's and try them out. In the meantime, I'm also considering a "digital" version of the log-approximation scheme used to turn a linear pot into an "audio taper" pot. Instead of using a pot though, I'm going to use a set of resistors connected in series, each one bypassed with a relay. Each DPDT relay shorts out one resistor on the top leg, while simultaneously un-shorting a like resistor in the bottom leg. This allows only 8 relays + 16 resistors to give 256 steps of linear attenuation (or 6 relays for 64 steps etc...). Add the resistor to ground, and you've got a log approximation attenuator, with good economy. We'll see if i actually build one of these attenuators, but I'll definitely report back on the CS3310s.
 
let me see here...

ok, Nov 1991 and the follow-up article was Nov 1992. The power supply design is ok, but I'd have to recommend going with a different buffer. If you visit my homepage, you can see the circuit I'm using right now (KAP-4), directly in place of the BUF-03FJ. Only one problem... there's an error on the schematic: the opamp is shown upside down (+/- terminals reversed)! Unfortunately, nbci is being a bunch of $#!+heads and since May I have been unable to change or even delete my own website!! I'm very annoyed... Anyway, here's the appropriate URL: http://members.nbci.com/_XMCM/hifizen/preamps.html