Buffer Amp - unity Gain

[Reffc]

Thought I'd have a bash at a buffer amp which will become the basis of a preamplifier when I get round to adding a selector, multiple inputs and ALPs pot. For now, I just wanted a buffer between my current valve pre and SS power which has relatively low input impedance of 5K1. I really wanted to see if the buffer would make any difference to bass (which is pretty much ok as it is).

I've decided to use the NE5532 low noise dual channel Op-amp as a unity gain stage, with a simple upstream circuit of a shunt resistor, series cap and resistor and a 100K load resistor. I've added some diodes as input protection, connected between signal and positrive and negative voltage rails.

power supply for now is going to be two 9V batteries.

The start this morning:



Front panels prepped, connectors fitted, switch and "power on" LED fitted



Battery connected to test switch/LED



Rear panel view:



Starting to populate left channel on prototyping board:



Both channels populated:



Underside view; copper wire along centre is the star-earth:





Assembled for testing:




Test measurements:

Input impedance across left channel = 985 KOhms

Input Impedance across right channel = 1mOhm

Difference is the variation in the input shunt resistors...now both measure exactly 1mOhm as I changed the sus one (at least not as tight a tolerance one).


Output DC voltage measure across output signal to ground:

Left channel = 93mV

Right channel = 94 mV

That's somewhat higher than I was expecting although with both channels being fairly equal, at least it may point to choice of components rather than a soldering fault or something. I've put a 100nF ceramic cap very close to the Op amp and between positive and negative rail inputs to the op amp so had hoped that this would keep DC to a handful of mV, but the reading is within acceptable ball park and is a reflection of the op-amp used I guess.

I'll report back on how it sounds when I muster up the courage to stick in the system!

[Dave the bass]

1mOhm....1 milli Ohm....thats almost a short circuit!!! 1MOhm thats more like it.

Out of interest how did you measure the i/p Impedance of the cct?

DTB

[Reffc]

Yes, sorry Dave, I meant 1MOhm, measured between star earth and input signal bucket on RCA sockets.


I'd been a numpty and forgotten to connect the mid point between the batteries to the star-earth hence higher than expected DC offset and distortion when I tried it.

Whipped the top off, saw the fault, felt silly, connected it and re-measured. I measured DC offset between output signal and ground on the output RCA jacks. Now reads a shade under 15mV for both channels.

Just tried it between Croft preamp (300Ohms output impedance) and Albarrys (5K1 input impedance) and wow, what an improvement. They sounded pretty good before as Greg and Nigel have noted, but now, the slight leanness to the top end has gone, instead replaced by a fuller mid range and more natural top end. bass has filled out to (like it needed to!).

Very low noise floor too, no hum...a great little circuit which did exactly what was wanted...result

[andrew Ivimey]

Its always sweet and nostalgic to see veroboard and, if I remember correctly, the 5532 was so much better than the equivalent of the double 741 package I used to use (something like 5584 - dunno) but having relatively recently swapped 5532 for opa6 ... (bloody hell, is this alzheimer's or what - there are plenty of folk her who know which i.c.s I am talking about) it does make the 5532 sound dull grey and lifeless in comparison.

p.s. I am a bit embarrassed about not remembering the exact i.c.s that I am talking about. Suffice to say there are enough peeps out there who can fill in the dots.

[Greg]

I can't fill in the dots Andrew refers to but I can add some subjectivity to this discussion. I have no idea about a dull and grey sound from a 5532.

Indeed, I commented on the good sound Paul had when we listened but to be sure, although very dynamic, it did become for me, a bit fatiguing during the day. With absolutely no disrespect to Paul, on getting home, I didn't want to listen to my system because my ears were, I'm afraid a bit fragile. The next day I listened to my own system and marvelled at the transparency and three dimensional presentation in a way that was beguiling and making me wanting simply more. As ever, my system contented me and I could listen to it for hours.

So, my point is, maybe this is just yet another of those synergy issues. 5532 may cause a dull sound for you but in Paul's system, maybe they bring the balance needed.

[Reffc]

I agree Greg which is why I built the buffer...the reason for the leanness was simply that a capacitor coupled valve pre with relatively high output impedance was trying to drive big SS monos with relatively low input impedance and experience has always taught me that this pairing can be too lean. The completed buffer amp addresses that very well. Imaging has also been improved as the balance has been restored to the music. The acid test was that I found myself listening to the music this evening, getting what I'd describe as an emotional musical experience, and not once was drawn to the hi-fi producing the music. Bass was also fuller and more natural. That thin mid was brought forwards and filled out...the whole sonic picture is now just very natural. It was yesterday's lengthy session with the Albarrys (the first I've really put them through their paces) that convinced me that a buffer was necessary and it works spectacularly well, so I'm pleased enough.

[Greg]

Well done Paul! Maybe I should visit you again soon

[grog]

OPA627 perhaps? it's quite good,


also consider the Nelson Pass B1 buffer,
ridiculously simple, and excellent,

i'm using the more complicated DCB1 version with discrete shunt regs and no coupling caps,
sounds amazing!
circuit below, actually it's not exactly the circuit, but Pass's own schematic complicates it by squeezing both channels and PSU on one diagram, (posted in the second image for completeness...)

[andrew Ivimey]

Thanks Grog - that's the one. Some of us tried other types of 'new improved ICs' but we all tried replacing the 5532 in a whorl of enthusiasm around a DAC module from Hong Kong that was (and still is!) cheap and cheerful. The module's output was a 5532 and the concensus was that this was it's only weak point.

And if my memory serves (ha! no chance) ... an AVI preamp I had many years ago was full of NE5532s and the cognoscenti preferred the preamps made with disrete transistors blaming the 5532s for making the amplifier less sonically desirable. I had an opportunity to compare one of these transistor pres with i.c.s and it did sound better on the day.

'synergy' or 'serendipidy' ... or even just subjective preference. I wonder.

[Reffc]

Thanks for the infor Grog

I couldn't say whether the OPA 627 was any better as the 5532 seems to do the job perfectly. It has been around a long time now and is still a very good circuit for the job. Not sure if I understand this subjectiveness about it being "grey" sounding as all it has allowed is for me to benefit from what the Croft Valve pre is capable of without really adding its own sonic character...it's certainly not "grey" sounding

One thing I see from the simplified circuit, posted above, or rather don't see, are any shunt capacitors to prevent switch-on and switch-off crack or pop from the speakers. Perhaps the inclusion of a 100nF 25v electrolytic between input switch and voltage rails would be a useful addition? I need to add these to mine as well.

Greg, you're more than welcome for a re-match any time.

[Mike H]

dull grey and lifeless ? Mine aren't...

[Reffc]

The 5532 is easier to work with in many respects than the 627 as it is very stable (as well as being low noise). The 627 can need very careful circuit implementation. I've been told by those using it commercially for a number of years that it takes a good deal more care in choosing the right application for it.