Consider the enclosed power supply.
The loads on the positive and negative rail are not equal.
The supply transformer is 24volt centre tapped and the data sheet says it is capable of 210 mA RMS on the secondary.
Theoretically it ought to provide +/- 17.5 V but it provides about +/-15.7 V.
The power supply powers two OPA2134 op amps.
In addition, the positive rail feeds two additional LM317 regulators, one at 3.3V and one at 5V (for a PCM2134 DAC and a CS8416 input chip)
Both downstream LM317s output correct (and theoretical) voltages.
With this arrangement the power supply's positive rail drops to +10.5 V.
This is a bit concerning, and adjusting the resistor values around the LM317 doesn't do much to increase the output.
Based on these observations, I'm thinking maybe the transformer is undersized? Maybe a 36VCT transformer would be better?
The loads on the positive and negative rail are not equal.
The supply transformer is 24volt centre tapped and the data sheet says it is capable of 210 mA RMS on the secondary.
Theoretically it ought to provide +/- 17.5 V but it provides about +/-15.7 V.
The power supply powers two OPA2134 op amps.
In addition, the positive rail feeds two additional LM317 regulators, one at 3.3V and one at 5V (for a PCM2134 DAC and a CS8416 input chip)
Both downstream LM317s output correct (and theoretical) voltages.
With this arrangement the power supply's positive rail drops to +10.5 V.
This is a bit concerning, and adjusting the resistor values around the LM317 doesn't do much to increase the output.
Based on these observations, I'm thinking maybe the transformer is undersized? Maybe a 36VCT transformer would be better?
Have you measured the current consumption on the positive rail?
Because if the regulator's input voltage drops too much under load, than the problem is the current, not the voltage.
Because if the regulator's input voltage drops too much under load, than the problem is the current, not the voltage.
You need more input voltage, at least 21VDC under full load. This implies at least a 30VAC with CT secondary,
or two 15VAC windings. Perhaps a 36VAC secondary ( 2 x 18VAC ) would be best. In this case, be sure that
the regulators can dissipate the needed power under load.
You really need higher voltage out of the transformer, at least 30vCT if you only need a few mA or preferably 36vCT. Remember the law of conservation of energy means there is no free lunch: At 0.21A from the transformer the filtered DC voltage on the output of the bridge can't be more than 12V less about 1V across the bridge = only 11Vdc. The regulator can't make up voltage.
Thanks - that's what I figured. I'm going to try with a 10VA 36VCT transformer. Hoping for better results.
I'm a little slow getting to you on this -- maybe you've already solved it. If so please feel free to ignore this 😉 . .
If not, what is the voltage drop across the 10 ohm resistor in the positive leg? Do we have any info on a possible cause of the asymmetry? You can run a whole pile of op-amps on a supply like this, typically without more than a few percent (maybe up to 10% or so) difference in current demands between the two rails. I think a failed or leaky part in the power supply is more likely.
Maybe pop out the fuse(s) and measure voltages. Also, confirm that the 10 ohm resistor really is 10 ohms.
Cheers
If not, what is the voltage drop across the 10 ohm resistor in the positive leg? Do we have any info on a possible cause of the asymmetry? You can run a whole pile of op-amps on a supply like this, typically without more than a few percent (maybe up to 10% or so) difference in current demands between the two rails. I think a failed or leaky part in the power supply is more likely.
Maybe pop out the fuse(s) and measure voltages. Also, confirm that the 10 ohm resistor really is 10 ohms.
Cheers
Hi . If transformer is with center tap ,yours is , load can be symetric ar assymetric , this loads transformer windings equally . Measure voltages at lm317 and 337 inputs , with your complete load it should drop both . If only positive rail drops ,it means too high drop on 10 ohm resistor ,or too small capacitor value for that current .
Set output voltage to 12v instead of 15 and see how the regs behave.Most audio op amp applications don't need more than that anyway.
Thanks for all the advice. The problem was solved this weekend on the arrival of a 10VA 36VCT transformer.
The smaller one just didn't have the mojo (or the voltage) to get the job done.
Current-wise the power supply is still imbalanced with the +ve side drawing more current than the -ve side, but that is because there are more downstream devices on the +ve side.
But voltage-wise it is outputting exactly + / - 15 V as expected.
The smaller one just didn't have the mojo (or the voltage) to get the job done.
Current-wise the power supply is still imbalanced with the +ve side drawing more current than the -ve side, but that is because there are more downstream devices on the +ve side.
But voltage-wise it is outputting exactly + / - 15 V as expected.