pinnerite wrote:
Ten years ago I bult a low-noise pre-amp based on a 1970's article in Elektor magazine.
It used a TBA231 dual op-amp IC.
It didn't work but I did not have time to test all the copper runs etc before I wa over-whelmed with 'work'.
This week I took it out and started fault-finding.There was one open connection, poorly soldered. I then checked the power links throughout the board and they were fine.
Next i connected in a little waveform generator to the input and one of those timy LCD screened chinese oscilloscopes to the output. Nothing.
So with no shortcircuits, I concluded that the capacitors were OK whch only leaves the IC.
I do not have another TBA231 or equyivalent but have no idea how to test the chip.
I am a bit out of date with electronics since my heydays
So any suggestions will be welcome.
TIA
This is a bit more modern than the TBA231/LM739.
https://www.ti.com/lit/ds/symlink/lme49720.pdf
That's a dual op amp which seems to take high rails
voltage and is intended for driving 600 ohm or 2K ohm loads.
The 600 ohm value, I gather, was a telephony headphone
value of long ago (suited to the vacuum tube era perhaps).
I might have had such a headphone at home when I was a kid.
The lme49720 has a GBW of 55MHz. Which means if you built a gain of 1000
preamp, the bandwidth of the circuit would be 55KHz. Compare
that to an LM741, where a gain of 1000 circuit has a bandwidth
of 1KHz and you can't even make out human speech properly in
such a situation. The LM741 is compensated for unity gain,
and has no external compensation network. The LME49720 is the
same way (there's no pins at all assigned for compensation
networks on it).
Op amps, in the canonical configuration, work via virtual grounds.
The input pins should be very very close to being the same voltage.
You can connect one input to a mid-rail ref ("ground" on your
dual rail supply), and the feedback sent around the feedback path
will bring the second input pin very close to the voltage of its mate.
Consequently, just building an amplifier and putting a small signal
into it, should be more than sufficient for verifying parametrics.
There's no distortion, because feedback removes it. Since your
preamp has tremendous gain, you'll need to take a resistive divider
and divide down your bench level test signal, to a suitably low
value for test. A phono cartridge might have 2mV p-p, so a
1mV test signal is suitable.
Look at how much lower the offset voltage is on the
op-amp above (not that this matters for audio amp
circuits). To get that low a decade or two ago, you
almost had to go with chopper stabilized amplifiers.
The offset voltage has to be that low, because of the
gains involved.
*******
On your current circuit, I would start by verifying
pin 14 and pin 7 have the V+ and V- values you expect
from your dual rail supply. The reference pin should
be at a mid-rail voltage (between V+ and V- values).
The output should not have an excessive load. If the
datasheet keeps referring to 600 ohms, then the circuit
is not going to be happy with 32 ohm headphones to drive.
Whereas a power amp with a 10K ohm input impedance,
is a sufficiently light load for a 600 ohm output and won't
upset it.
If you place a heavy enough capacitive load on an op amp,
it will "motor boat" and make a put-put-put sound at
a few hertz. One of my colleagues at work tried that.
Being a hobbyist after my own heart, he attempted to
"squash the erroneous behavior" of his op amp circuit,
by putting a large cap on the output. The app notes
will instruct you to place a 50 ohm resistor between
the output of the op amp and the capacitor, if you
really really want to be doing that :-) That's an example
of load-isolation to maintain gain/phase margin.
Paul