Hi, recently i went biamped on my system using what i have already, i ended up using a old BIAMP SM23 active ross over. since it was all messed up with inside i made new pcb'S and removed the balanced line in/out circuits, because i felt that using RCA cables (home audio) would mess it all anyway.
Problem is, whenever i raise the gain input pots, i get opamp noise/hiss, that comes inevitably from the source a bit.
So, i post the simplified schematic of the BIAMP thing. yes its old and i know there a tons of better things on the market, no need to argue about that...but for the sake of knowledge on this, what would be the main problem with this circuit that would generate this much hiss noise ?
all resistors are 1% metal films
opamps are LM4562.
one thing i found is that when i reduce R1 100k down to 47k, with nothing attached to the input, the noise reduce by 50%.
thanks for any input.
Problem is, whenever i raise the gain input pots, i get opamp noise/hiss, that comes inevitably from the source a bit.
So, i post the simplified schematic of the BIAMP thing. yes its old and i know there a tons of better things on the market, no need to argue about that...but for the sake of knowledge on this, what would be the main problem with this circuit that would generate this much hiss noise ?
all resistors are 1% metal films
opamps are LM4562.
one thing i found is that when i reduce R1 100k down to 47k, with nothing attached to the input, the noise reduce by 50%.
thanks for any input.
Why do you care about the noise with open input? Normally the input will be driven from some low impedance.
Exactly.
However, that first stage does have unnecessarily high impedance levels.
Unless your source is a tube design, reduce all the first stage resistors (and the pot) by a factor of ten.
Do not change R4 and the subsequent resistors.
The C1 capacitor value will still be ok, but a Nichicon UES of the same value will be better.
Also C2 should now be 100uF 25V Nichicon UES.
However, that first stage does have unnecessarily high impedance levels.
Unless your source is a tube design, reduce all the first stage resistors (and the pot) by a factor of ten.
Do not change R4 and the subsequent resistors.
The C1 capacitor value will still be ok, but a Nichicon UES of the same value will be better.
Also C2 should now be 100uF 25V Nichicon UES.
Last edited:
If you're indicating that noise from the driving source rises with gain, this is as it should be. Noise density should originate from the input source. A test: at desired volume, if you disconnect source and replace source with a short (or 600 ohm shielded resistor) and find reduced hiss, you're in a good place.
the preamp where originate the music is a Nikko Beta30, with ultra low noise opamp. no matter what it hiss a bit t high volumes, and get amplified by the final cross over.
all stages generate a bit of noise by themselves when tested seperatly. the cross over too alone does it a bit.
i dont turn up the input gain stage this much. i even think about deleting this stage as i dont use it much. this would remove one opamp possibily responsible of 50% of the noise imho.
all caps are elna simic II or Fine gold. c1 and c2 are 47uf FG for now.
i will try to lower the input resistors /10 of the gain stage.
thanks.
all stages generate a bit of noise by themselves when tested seperatly. the cross over too alone does it a bit.
i dont turn up the input gain stage this much. i even think about deleting this stage as i dont use it much. this would remove one opamp possibily responsible of 50% of the noise imho.
all caps are elna simic II or Fine gold. c1 and c2 are 47uf FG for now.
i will try to lower the input resistors /10 of the gain stage.
thanks.
Hi Pat,
I advise against changing the input resistors at U1A until there is better understanding of gain distribution and noise input mechanisms.
Do as rayma suggests and set pot R2 to 0 resistance. BTW, the schematic has an odd scheme, as full CW rotation of R2 sets U1A gain to its minimum of 1 rather than convention maximum gain. Perhaps you should confirm with ohmmeter that full CW rotation of R4 shorts pins 1 and 2 of the opamp.
In this configuration, the equivalent input noise wouldn't improve much if you eliminated U1A entirely. The LM4562 opamp you mentioned has 2.7 nV/√Hz, but R4=10K has about 13 nV/√Hz, so deleting the opamp wouldn't change much.
Note: Both my and rayma's posts mention adjusting R2. Schematic in fact has R2 is a 1M and R7 is the pot in question.
I advise against changing the input resistors at U1A until there is better understanding of gain distribution and noise input mechanisms.
Do as rayma suggests and set pot R2 to 0 resistance. BTW, the schematic has an odd scheme, as full CW rotation of R2 sets U1A gain to its minimum of 1 rather than convention maximum gain. Perhaps you should confirm with ohmmeter that full CW rotation of R4 shorts pins 1 and 2 of the opamp.
In this configuration, the equivalent input noise wouldn't improve much if you eliminated U1A entirely. The LM4562 opamp you mentioned has 2.7 nV/√Hz, but R4=10K has about 13 nV/√Hz, so deleting the opamp wouldn't change much.
Note: Both my and rayma's posts mention adjusting R2. Schematic in fact has R2 is a 1M and R7 is the pot in question.
Last edited:
If the noise with open input doesn't matter, there is no point in reducing R1. Reducing the other resistors does reduce the noise of the first stage when it is set to a higher than minimum gain - if it is the first stage and not the signal source causing the noise.
the pot is inverted on the schematic, my fault, but is wired properly on the board. full ccw makes it 0 gain.
opamps are not tlo but lm's
after some further test with shorted input of the cross over, the noise reduces a lot so the Nikko is in doubt there. my findings is the noise comes from the nikko and is amplified by the cross over. the nikko has a LM4562 as well as the output stage....
opamps are not tlo but lm's
after some further test with shorted input of the cross over, the noise reduces a lot so the Nikko is in doubt there. my findings is the noise comes from the nikko and is amplified by the cross over. the nikko has a LM4562 as well as the output stage....
would swaping to LT1115 help at all or i am toward more problems ? the LT has 3x less noise figure per data sheet (0,9 for LT and 2.7 for LM)
In the Nikko you mean, as the Nikko apparently dominates the noise per post #12?
Depends entirely on the circuit of the Nikko. The LT1115 has a lower noise voltage, but a higher noise current (about 3.4 pA/√Hz, the 1 pA/√Hz stated in the tables in the datasheet is measured under unrealistic conditions).
Depends entirely on the circuit of the Nikko. The LT1115 has a lower noise voltage, but a higher noise current (about 3.4 pA/√Hz, the 1 pA/√Hz stated in the tables in the datasheet is measured under unrealistic conditions).
the nikko has a strange circuit with fets, i dont know. from now, all seperate units seems to have a bit of noise. i would tend to replace them all at some point. i replaced the njm2043d because it was dead, by lm4562 as well. the schematic is in two part, sorry on the far right R733-c711 goes to the cross over BIAMP thru rca cables.
Last edited:
here is a better picture showing the tone control circuit as well (the other is the beta50 without the proper tone control like the beta30)
The R23 and R24 pots are intended to adjust the relative spectral balance between LF and HF paths. Make sure that one of these pots is at maximum. (BTW, what is value of the pots?) Set the R7 pot at minimum which is unity gain.
Do the power amps have their own level controls? Set them to a fixed position as low as you can so the listening level is controlled solely by the preamp.
All these measures are directed toward getting the best noise figure available from the preamp. You want the noise floor to be controlled by the preamp and the original program material.
Do the power amps have their own level controls? Set them to a fixed position as low as you can so the listening level is controlled solely by the preamp.
All these measures are directed toward getting the best noise figure available from the preamp. You want the noise floor to be controlled by the preamp and the original program material.
The preamp volume control is before the tone controls. With the volume control at 0, are you content with hiss levels? That's as good as it will get.
nope. thats the problem. with volume pot at 0, all others set to have acceptable listening levels (with still cross over gain at 0), LF and HF set to have flat responce from my SA3050 RTA, i get a bit of hiss.