Swapping Op-Amps... you have checked to see it's stable haven't you ?

Thank you all - I can't find a schematic for the (older, or original not the new "Super") Eastern Electric Minimax DAC but I will continue to look. The LME9940 is a "common" (DIY, smart? effective? is another issue) swap (with the brown-dog adapters) for the stock I/V conversion stage and output. The stock opamps are I/V 2 x NE5532. After this opamps, there're 2 outputs stages, one based on a single tube (12AU7) and other SS with the opamp NE5534.

Scanning the net reveals many- many - comments about them (LME49990) running hot (in all kinds of audio circuits. It is *apparently* common. The specs allow for a range up to about 185F, but I didn't understand why my right and left channels seem to run at different temps. It could be my laser temp. reader is not consistently hitting the hot-spot on the U2 and U7 due to the angle at which I have to take a reading.

OPERATING RATINGS(1)
Temperature Range
TMIN ≤ TA ≤ TMAX –40°C ≤ TA ≤ 85°C
Supply Voltage Range ±5V ≤ VS ≤ ±18V

But, the LME4990 run so hot (occasional reading up to 140F on U1) that I might worry about putting the Minimax lid back on top and driving ambient up inside the unit.
 
Last edited:
actually the lme isnt exactly the best choice for IV, its a great chip for low noise VAS but not really fast enough for a ES9018 trans-impedance stage IMO. it seems its been chosen due to its (well earned) reputation as a nice low noise opamp for not much money, rather than being useful for the job at hand.

that being said, being on an adapter in a socket that isnt really very well decoupled (not close enough and only using audiophile approved leaded film caps, rather than something usefully small and low inductance...) the right side is very close to the clock, could mean you are picking up stray RF. is it the one on the right in this pic next to the clock thats hotter? I wouldnt worry too much about a tiny opamp heating up the chassis ambient too much, but if they really are that much different, then there may well be something undesirable going on
 
Thanks would you recommend anything over the I/V stage stock NE5532?

actually the lme isnt exactly the best choice for IV, its a great chip for low noise VAS but not really fast enough for a ES9018 trans-impedance stage IMO. it seems its been chosen due to its (well earned) reputation as a nice low noise opamp for not much money, rather than being useful for the job at hand.

that being said, being on an adapter in a socket that isnt really very well decoupled (not close enough and only using audiophile approved leaded film caps, rather than something usefully small and low inductance...) the right side is very close to the clock, could mean you are picking up stray RF. is it the one on the right in this pic next to the clock thats hotter? I wouldnt worry too much about a tiny opamp heating up the chassis ambient too much, but if they really are that much different, then there may well be something undesirable going on
 
Administrator
Joined 2007
Paid Member
I would always recommend an analogue scope (using a CRT) and one that has (and everyone has their own ideas here) at least 20 to 30Mhz bandwidth with dual trace. Higher if you can get it but if not that will sufficer for most diy (and more) needs.

No idea on prices tbh although auction sites list dozens :)
 
LME49990

I haven't detect temperature issues with the LME49990 used in I/V of my Minimax.

There may be more suitable opamps for the function, but it's the best direct drop option I've already tried. I have tried the opamps bellow that are ranked in my preference order:

- LME49990 (similar to 49720, but more smooth highs)
- LME49720
- NE5532
- OPA2604

For output, the model chosen is ADA4627-1BRZ.
 
Administrator
Joined 2007
Paid Member
Instability is always a possibility swapping opamps and is why you should always check with a scope. Without that check you just don't know for sure and as this thread mentions at the start, its not always audibly obvious that a device is oscillating.
 
new Rigol 'scopes and others have become ridiculously high value propositions in just the past year or so - with GHz sample rate the 50 MHz dual channel sufficiently oversamples for a "analog feel"

buying a new, working 'scope seems like a very practical way to go today
 
www.hifisonix.com
Joined 2003
Paid Member
Thank you all - I can't find a schematic for the (older, or original not the new "Super") Eastern Electric Minimax DAC but I will continue to look. The LME9940 is a "common" (DIY, smart? effective? is another issue) swap (with the brown-dog adapters) for the stock I/V conversion stage and output. The stock opamps are I/V 2 x NE5532. After this opamps, there're 2 outputs stages, one based on a single tube (12AU7) and other SS with the opamp NE5534.

Scanning the net reveals many- many - comments about them (LME49990) running hot (in all kinds of audio circuits. It is *apparently* common. The specs allow for a range up to about 185F, but I didn't understand why my right and left channels seem to run at different temps. It could be my laser temp. reader is not consistently hitting the hot-spot on the U2 and U7 due to the angle at which I have to take a reading.

OPERATING RATINGS(1)
Temperature Range
TMIN ≤ TA ≤ TMAX –40°C ≤ TA ≤ 85°C
Supply Voltage Range ±5V ≤ VS ≤ ±18V

But, the LME4990 run so hot (occasional reading up to 140F on U1) that I might worry about putting the Minimax lid back on top and driving ambient up inside the unit.

If this is running as a 'summer' on a current output DAC, then what you may be seeing is stray capacitance between the summing node and GND that's causing it to break into oscillation. Also, keep in mind that any noise coupled between the actual summing point (i.e. where the resistors meet) and the inverting input will be amplified by the full loop gain of the opamp - and in the case of those LME devices, this could be over 100 dB at a few kHz
 
Great post about Op Amp "Singing" (Oscillation)

Its more years than I care to remember - however, advise about Op Amp compensation
can often be obtained from application notes or even data sheets themselves.

The NE5532 is an Op Amp that uses bi polar transistors
Many of the Other Op Amps mentioned here have jFET's on the input.

Zin of an Op Amp with a jFET input stage would be significantly different.
Rin of a jFET Op Amp would be in the order of 10^12 or 10^13 ohms
and Cin would be in the order of 10 pF

Its important to remember that at the x1 setting, a scope probe adds about 15pf // 1 M ohm.
So be careful not to disturb the very circuit you are trying to measure.

Using the x10 setting on the probe will help.
However, a wire added to the scope probe will act like a receiving antenna.
If its place near the circuit it should pick up any "singing"
Mooly was bang on showing oscillation was hiding in the blurred trace.

One other comment - the chassis of a scope is earthed.
So a scope can only measure voltages with respect to 0 v.
.
 
Last edited:
Administrator
Joined 2007
Paid Member
Good point on the scope leads introducing extra capacitive loading which can in some cases tip a circuit over the edge. Worst offenders are the simple "coax and crop clip" type leads that are surprisingly common. If in doubt then it can pay to solder a lowish value resistor to the point of interest to provide isolation and then probe the free end of that.