Early op amps 741 etc

[newlite4]

This is an interesting thread. At about this time of year, I begin to re-pack the student component trays ready for next years labs. We mainly use the 741 and have done for many years, however demand for better performance is leading us to slowly change over to the TL071. Imagine my surprise when checking the date codes of the 741s to find that a fair number date from 1971 .
Neil

[Tim]

I have found 741's to be fine for audio, as long as they are not asked to provide too much gain. I have used 5532/5534 for stuff, but couldn't tell the difference.
When building projects that specify other OP amps, I can never seem to get them to work properly, and end up reverting to the 741 which ALWAYS seems to work. The 741 is great for low gain applications, buffering,inverting etc.
Mind you I still use BC107's...

[G6Tanuki]

In the context of op-amps, does anyone remember the uA703? It was a relatively simple and early beastie usually in a metal TO-5 style can, designed as a RF/IF-amplifier. CA3028 is similar.

Those of us who couldn't afford Plessey SL600-series RF/IF chips used the 703 quite a bit.

I also liked the RCA CA3020A - which was a sort of "Power op-amp" that could produce 1/2 a watt of output up to about 5MHz - but you needed to heatsink it well.

[Marconi_MPT4]

In the context of op-amps, does anyone remember the uA703? It was a relatively simple and early beastie usually in a metal TO-5 style can, designed as a RF/IF-amplifier. CA3028 is similar.

Have used the lower spec RCA CA3053 recommended for IF amplification. Also surprised to find these in a Revox B760 tuner where four were used in the IF amplifier with a further two in delay line FM demodulator circuit. Metal cans get hot though.

For most non-critical design applications, I tend to use the inexpensive TL072 series. Input offsets are not particularly well matched as it has J Fet input circuit, but careful design largely overcomes limitation. Later versions such as the TI Excalibur range TLE2072 are better.

Rich

[Ed_Dinning]

Hi Yes the CA series was very useful and popular back then. These were the devices of choice after the capabilities of 741's were exhausted. I still have the app notes somewhere.

I also used the 710 comparator as an amp, great slew rate but needed a lot of extra bits to keep it stable.

Ed

[Radio Wrangler]

My final year student project was a PAL colour synthesiser using CA3028s as variable transconductance multipliers and it worked. I've used them for years. Excellent devices.

David

[Gillian]

HI, Update time. I have now built a 741 based audio line level preamp with 2 gain stages.
One pre volume attenuation and one post.
I have kept the voltage gain low on both and I am running them on +- 15V supplies.
They are decoupled by 47uf on each op amp and each supply.
Well all I can say is I thought it was not working at first because I could not hear any noise or switch on pop.
But as soon as I played a CD, wow it sounds fab and all I can hear is the music.
So there we are, still not to be overlooked in my humble opinion for hifi or Studio use.
Best Wishes
Gill

[Ed_Dinning]

Hi Gill, try adding 0.1uF disc ceramics across the E-caps, but directly on the amp pins, leads as short as possible, if you have not already done so. It will get rid of any slight HF present.

These amps work surprisingly well as the Texan and others showed.

Ed

[cmjones01]

Originally Posted by Gillian

HI, Update time. I have now built a 741 based audio line level preamp with 2 gain stages
...
But as soon as I played a CD, wow it sounds fab and all I can hear is the music

That's great - I'm glad it works well. I think you've just demonstrated the major point of this thread, that there's no magic in 'good sound'. It's about understanding the technical limitations of the components in use, and using the right design techniques to handle them. It doesn't matter whether it's an EF86, BC107, 741, NE5534 or something else, the same principle applies.

Chris

[Radio Wrangler]

Ed's suggestion works because of the ESR of the electrolytic capacitor. Two capacitors in parallel at high frequencies make a resonant loop, with the capacitance around the loop (probably just that of the smaller C) resonating with the inductance around the loop. This can give rather odd effects at RF, and dodgy stability.

Done properly, the effective series resistance of the electrolytic capacitors damps the resonance to a very low Q and things work well.

It seems odd that ESR is a good thing... at the right value, when so many of us spend time measuring it to weed out dying capacitors.

Take a look sometime at a supply distribution arrangement in some RF box and then ask yourself about all those chokes and decouplers and how they would twang. The other saviour is the use of lossy ferrite cores so chokes look resistive at HF.

David

[jim_jobe]

Originally Posted by Tim

Mind you I still use BC107's...

Some of us still use OC71s!

[Gillian]

Yes, My next step is good ol germanium.
I have obtained about 50 2SB 54 Toshiba pnp germanium transistors.
I did do a previous thread about building them in op amp topology, that should be cool.
I will be building a recording console, perhaps a 16 channel using which technology produces the best sound in which context. It will be a hybrid of possibly germanium, silicon and op amps if that sounds good to me.
It will incorporate star grounding and power distribution and multi regulators all decoupled individually back the low impedance star ground point.
research continues with interest!
Gill

[Ed_Dinning]

Hi Gill, couldn't you at least stay European and use the OC power series, or the earlier VP*** types?

The Bailey amp in WW many years ago with OC35's and a trifilar wound driver trans did sound good (Si in the pre-amp though)

Ed

[Gillian]

I could do,but a friend of mine had an unopened pack of 50 2SB 54 transistors and I swapped them for a cassette deck.
I think the Japanese ones are good for low leakage current as well.
Gill

[Mooly]

So just how "good" is the 741 ? I couldn't resist a little experiment here, and so I rigged up a two opamp gain stage. Its a unity gain buffer followed by an attenuation network (15k and 2k2) with a second opamp taking the attenuated signal and bringing it back to the nominal level it was at the start by using 15k and 2k2 as the feedback network for this second stage.

You can hear (or not ) for yourself. Do you prefer file A or file X ? These are .wav files. One uses the 741 opamp, the other the OPA134 FET device which is an "audiophile" favourite. File AA and file XX are the same but MP3 versions and so much smaller files.

The links are just dropbox files.

WAV

FILEA
FILEX


MP3's

FILEAA
FILEXX

[Nuvistor]

Well I listened to the wav files on a decent pair of headphones and could not tell the difference. Not at lot I can say after that.
Frank

[Mooly]

It is very surprising how close they are. Used within its limits (modest gain, sensible impedances and signal levels) it can come quite close to the best of what is available today.

[audiomik]

Something which you might like to try when using 741 Op-Amps for Audio is to ofset the output crossover point of the output stage in order to reduce low level signal distortion.

For the Output, add a 680 Ohm resistor paralleled with a 220µF 6.3V cap in series, then a 15k resistor to one of the +/- voltage supply rails. Use a suitable electrolytic coupling cap in series with the junction to the next stage/output.

This has the effect of shifting the crossover distortion point to a greater output voltage, thus reducing low level crossover distortion - easily measurable and was used on a number of mixing desks

Mik

[mhennessy]

A transistor current source is another way of doing that.

[Gillian]

Hi Mooly01 ,Fascinating I can hear a difference and it is on the transient response
I definately prefer file x or xx .
On the timpani and sudsequent reverberation there is much more detail.
As a musician it just sounds more like music.
I will stick my neck out and the contextof the experiment I prefer track ot file X.
Hope it isn't a double bluff or emperors new clothes trick and they are both the same,
Is track X the 741?
Gill