I got a free Onkyo TX-84 from electronic salvage and replaced all the electrolytics. While I was in there I replaced all the NJMxxxx chips with OPA1612's. The sound is incredible but there is constant hiss even with the volume all the way down. The hiss volume stays the same even when the volume knob is maxed out. So it looks like the chips might be oscillating. I added 0.1uF Kemet ceramics to all the Ve- and Ve+ of the chips to ground but the hiss is still there. Is there anything I can add to the circuit or modify to stop the hiss? I'm using headphones with the receiver
View attachment tx-84.pdf
View attachment tx-84.pdf
The opamp is rather meant for high gain stages (turntable, mike) in gain of 1 stages they require a >220pf across the feedback resistor (10k), or respectivly bigger for lower ohm. they easily oscillate at 15MHz. Try to put your finger on the opamp and see if the noise gets changed, that is a clear indication.
Surprised you went to the 'trouble' for such an 'upgrade' -- I only saw what, 4 or 5 op-amps in the whole thing? And several of them can easily be avoided -- VCR2 output driver, S.S., phono preamp ..
+1 bansuri on the finger/op-amp tip. 220pF seems kinda heavy, though; but I still have some PDF reading to do.
Maybe a pic of how you fitted the Kemets would give a clue.
Cheers
+1 bansuri on the finger/op-amp tip. 220pF seems kinda heavy, though; but I still have some PDF reading to do.
Maybe a pic of how you fitted the Kemets would give a clue.
Cheers
So now you hear 0.0001% distortions through a 0.04% THD power amp...That's incredibly sound.There can't be any incredible sound with a hiss over it in the first place.I suspect you actually like white noise so maybe you should let it be...
I agree with Dreamth this doesn't sound like HF oscillation but more like white noise being generated , you usually hear this on more sensitive components being installed in an old amplifier with low spec passive components like cheap carbon resistors .
Easy to prove if you have an oscilloscope.
Easy to prove if you have an oscilloscope.
A single 0.1uF Kemet ceramics between the Ve- and Ve+ of the chips rather than two to ground is usually recommended.
With the provision that the rail supplies are obtained from some 22...47 ohms resistors and the electrolitic filter capacitors before them are at least 5 cm away from the op-amp.
I never heard of such a provision. For the electrolitic filter capacitors, the most close to the op-amps are usually not more than 220 µF and should not cause problems with 100 nF in parallel.
Last edited:
I would fit the correct op amps as a start and then trace the noise with my oscilloscope if it is still there.
There is only one, Q351
There is only one, Q351
Have you considered amplifier current noise? The OPA1612 has worse current noise rating compared to for example the good old NE5542.
The op-amps are not necessarily bad just because the part number start with NJM. Also they are not necessarily "one size fits all" just because the part numbers start with OPA.
The fact that the original design uses four different op-amp for different purposes indicate that he knew what he was doing.
I brief look at the schematics show that many/all of those op-amps have hi value resistors on their inputs.
Since the noise is independent of volume control my first guesses would be to look at the only two op-amps that are after the volume control. The NJM4560 that buffer the volume control and the NJM4558 used for tone control. The NJM4558 can be bypassed usig the tone-defet button.
Maybe re-inserting the original op-amps or replacing these with something like NE5532, OPA2134, or something else with current noise less than 1pA/rtHz and voltage noise less than 8nV/rtHz would do the trick
Does the character of the noise change as you switch the tone controls in/out?
Last edited:
I'll try the 220pf cap first and see if it affects the hiss. Its worth tinkering with because I never heard an old receiver sound this great. Even the phono stage is incredible
Last edited: