Yesterday I replaced the five TL082 opamps in a 2 way crossover with five OPA2134 from BurrBrown. I've added 0,1uF under each opamp, between V- and V+.
The gain in SQ is jaw dropping (noise floor, clarity, detail, smoothness... all has improved dramatically) and there's no way back!
Someone on this forum raised the subject of possible problems with the PSU being unable to deliver enough current for the new opamps.
From what I can read here and there, TL082 draws half the current of OPA2134 (5mA vs 10mA/chip).
As said above, I have 5 opamps. They're the only active devices in the circuit post regs (there's just one LED pre-regs, from a distinct enroulement of the power trafo).
Regulators are LM317L + LM337L in TO92 package. Max current is 100mA.
There are 4 x 470uF pre-regs.
As I understand it, I might run into problems depending on how much MAX current the OPA2134 will effectively use on top of their idle current.
In this case, the first opamp is an input buffer, and each of the other 4 sees the input impedance of a power amplifier (it's a 12dB/oct crossover, so there's one opamp for each channel: LP-L, LP-R, HP-L, HP-R).
In the context of my system, speakers are all hornloaded, bass included, and average sensitivity is around 104/105dB.
So even if the power amps have an input sensitivity of 1V, the chances for me to actually need 1V is practically zero, since it would drive my speakers at levels unsafe for my hearing - and incompatible with having neighbours anyways.
What I need is someone doing the maths roughly and telling me if he thinks I'm still in the safe zone.
I guess I am, but I wanna be sure...
The gain in SQ is jaw dropping (noise floor, clarity, detail, smoothness... all has improved dramatically) and there's no way back!
Someone on this forum raised the subject of possible problems with the PSU being unable to deliver enough current for the new opamps.
From what I can read here and there, TL082 draws half the current of OPA2134 (5mA vs 10mA/chip).
As said above, I have 5 opamps. They're the only active devices in the circuit post regs (there's just one LED pre-regs, from a distinct enroulement of the power trafo).
Regulators are LM317L + LM337L in TO92 package. Max current is 100mA.
There are 4 x 470uF pre-regs.
As I understand it, I might run into problems depending on how much MAX current the OPA2134 will effectively use on top of their idle current.
In this case, the first opamp is an input buffer, and each of the other 4 sees the input impedance of a power amplifier (it's a 12dB/oct crossover, so there's one opamp for each channel: LP-L, LP-R, HP-L, HP-R).
In the context of my system, speakers are all hornloaded, bass included, and average sensitivity is around 104/105dB.
So even if the power amps have an input sensitivity of 1V, the chances for me to actually need 1V is practically zero, since it would drive my speakers at levels unsafe for my hearing - and incompatible with having neighbours anyways.
What I need is someone doing the maths roughly and telling me if he thinks I'm still in the safe zone.
I guess I am, but I wanna be sure...
You haven't specified the voltage difference over the regulators, but as long as you don't burn your fingers on the TO92 packages, I wouldn't be worried at all.
Hans
The current quoted is the quiescent current when zero output current is flowing.
Using 5 @ 5mA gives a quiescent load of ~25mA.
Changing to 5 @ 10mA gives a quiescent load of ~50mA
You can check this by adding in a temporary 1r0 resistor after the two regulators and measuring the Vdrop across the 1r0.
Each millivolts is equivalent to 1 mA.
Using a 199.9mVdc scale on your DMM, you should get a reading between 45.0mVdc and 55.0mVdc
Then you need to add on the extra current when a signal is present.
I would expect a significant extra load when processing louder signals.
You can add little flag heatsinks to To92 devices.
As far as I know the To92 regulators also have over temp protection, just like their bigger brothers & sisters.
Using 5 @ 5mA gives a quiescent load of ~25mA.
Changing to 5 @ 10mA gives a quiescent load of ~50mA
You can check this by adding in a temporary 1r0 resistor after the two regulators and measuring the Vdrop across the 1r0.
Each millivolts is equivalent to 1 mA.
Using a 199.9mVdc scale on your DMM, you should get a reading between 45.0mVdc and 55.0mVdc
Then you need to add on the extra current when a signal is present.
I would expect a significant extra load when processing louder signals.
You can add little flag heatsinks to To92 devices.
As far as I know the To92 regulators also have over temp protection, just like their bigger brothers & sisters.
- Status
- Not open for further replies.