OpAmp Measurement Series

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Hi Samuel,
For a number of years I have used an IC that I liked a lot, the MC33178 by TI. It's an interesting animal that combines relatively low noise (8nV/sqrtHz), low current drain (0.42 mA typ) and high drive capability (+/-12v into 300ohms). Surely not the fastest opamp in the world, but it is very stable, tolerant to layout design, in short the ideal all-rounder. After having used several tens of thousands in production, the distributor said the product was discontinued, so I had to use other products. In many cases, I replace it with the JRC2068, except for output stages, where the 5532 is stlll a very strong contender.
I've just found out that the MC33178 is stil current with ON semiconductors, in its Pb-free guise at http://www.onsemi.com/PowerSolutions/product.do?id=MC33178PG

Just to add to the superlative job you've done so far, that would be interesting to see how these two relatively obscure parts, the 2068 and the 33178, behave in your series of tests.
At the time I had conducted all sorts of tests including DIM, but the most important was that nobody ever complained about any restriction in sound quality, although many sound aficionados would churn on its meagre 2V/us slew rate.
I had also used huge quantities of JRC4559, in positions where its low noise (same as 5532) and low consumption (same as a TL072) were appreciated (mainly in the " "gyrators" " in graphic EQ's.
 
A couple I'd like to see added to the pot:

LF411/412
AD711
LM318 (venerable)

It also might be worth looking at a few chips with and without current-sourced bias on the output. I've done some preliminary investigations, and adding the bias almost always cuts down on higher harmonics in the midband HD test, and cuts down intermodulation products in the 19/19.5kHz IM test.

I'll post some pix in the next day or so, after I get some work out of the way.

Peace,
Paul
 
abbey road d enfer,

FYI
The JRC2068 is available in the states from Mouser as part # NJM2068.
It is thought to be an equivalent to the National LM833

Cheers.

ZAP
 
Posted on: Today at 12:32:02 pmPosted by: zapnspark 
abbey road d enfer,

FYI
The JRC2068 is available in the states from Mouser as part # NJM2068.
It is thought to be an equivalent to the National LM833

Cheers.

ZAP
Thanks,
I have no problems finding the 2068; it is available from Profusion in the UK.
It's the MC33178 that is very hard to find. In fact it is available from ON semi on their website, but they only have 287 pieces in stock, so I suspect it's not a really hot item for them; I don't want to use a product that has no second source and very likely to shortage.
The LM833 has never been a serious contender, because it's not as good as a 5532 and much more expensive. Today the 5532 is so cheap (well cost-effective I should say) -we pay $0.34 - that the only reason I don't put it everywhere is that it's current consumption is too high. I have a product with 112 dual op-amps. If I put 5532's, I have to fit an 80W mains xfmr and big heatsinks on the regs. By using TL0's and JRC2068, I can use a 30W xfmr and smaller, cheaper heatsinks.
 
Thanks for the parts suggestions--some of them are already on my to-do-list. I hope to include implementation comparisons (such as the class A bias mentioned) as well, although this might take a while.

Samuel
 
I'm late to the party, but thank you for this tremendous effort, Samuel!

Chips I would like to see you test include the Analog AD8597/9, TI TLE2071/2/4 (TL071/2/4 replacement), OPA227/228, and OPA211(A) if you can get any, Linear LT1113 and/or LT1169, and National LME49725 and LME49723/49743. The ON Semi MC33178 is of interest, but I don't expect much from it.

Also note there is a lot of, uh, inexpensive "studio" gear using Japan Radio Corp. (JRC) op amp chips, e.g. NJM2115, NJM4560, NJM4580, etc. It would be interesting to see how these compare.
 
Thanks for the update, Samuel! Always interesting stuff.

Do you have any idea of the performance of the JFET993 @ +/-22v? The reason I'm asking is that I'm really happy with the sound of these op-amps @ +/-24v, but for the new EQ I mentioned, there's going to be some IC's used too. +/-22v rails wouldn't tie my hands too much, since there are quite a few good IC's that can take +/-22v. The single version of the LME49860 (LME49870) looks good.

I don't think it's going to be a problem with +/-22v rails.. Just keep the levels low enough.
 
I haven't tested the JFET 993 at any other voltage besides what's shown in the PDF; it's pretty save to assume though that the performance at 44 V will be much closer to the results shown for 48 V than to those for 30 V.

Samuel
 
Recently posted at EDN:  Sam Groner’s op amp distortion study  :)

In the comments, [Edit: Paul] Barry Harvey writes  "It's pretty clear that little effort has gone into audio amplifier optimization in the analog IC industry. Part of the reason is cost and volume. The high-fidelity audio market has never been lucrative for IC companies. Today, with the consumers' tolerance of low-quality MP3 music reproduction, there is even less commitment to design quality audio IC amplifiers. Pity."
 
Re: reply #89, the actual quote was:

at 2/19/2009 2:28:05 PM, Barry Harvey said:
So, reading through the measured data in the excellent study it is clear that wide bandwidth or fast slewrate or high accuracy do not ensure good audio distortion. Further, the input stage can have common-mode distortion, potentially. From my work with DSL line drivers, you can design an amplifier with poor accuracy and gain, and be not really all that fast, but still have low distortion even up to 1MHz. It''s pretty clear that little effort has gone into audio amplifier optimization in the analog IC industry. Part of the reason is cost and volume. The high-fidelity audio market has never been lucrative for IC companies. Today, with the consumers'' tolerance of low-quality MP3 music reproduction, there is even less commitment to design quality audio IC amplifiers. Pity.


This is an important distinction. Barry Harvey is an experienced designer for Intersil. Paul Rako is the EDN editor who wrote the original blog item.

wj
 
Quite right - thanks for the correction...  I started with quotes from both Paul and Barry, but decided to keep my post concise.  In the process of editing got the name swapped.
 
Thanks Walt.  I was wondering who Barry Harvey was.

I think the slide to poorer and poorer audio quality, a la MP3 being acceptable now, will eventually bottom out and turn around, but it may be a while.
 
Thanks to Walt Jung for contacting Paul Rako with this matter! I'm glad that this work finds wider audience.

A new revision is online, now with 51 amplifiers. PDF bookmarks added for easier navigation.

Samuel
 
Back
Top