Yes, C12 is part of the filter, D7, D8 are just for the opamp's input protection.
Last edited:
The power up thing actually makes sense. TL072 will act strange if it's inputs are near rails on power up which would give a wild output. C13 would hold this output voltage for a short period.
I haven't messed around a lot with LF412. It may act the same. Many op amps can't operate with the inputs near VDD or VSS.
If using a separate dual supply for the opamp, then I think you must protect the inputs from excessive voltage.
D7 and D8 will protect the + input. R27 should take care of the - input well enough for a bench test.
No, the +IN to -IN diodes only protect against differential voltage.
One would need to add the diodes to prevent the INs exceeding the rails.
One would need to add the diodes to prevent the INs exceeding the rails.
The external supply +/- would need to reference the same ground as the op amp. If he's using the supply for the NS-OPS, he'll already have a +/- 15V supply referenced to the same ground built into it. Too bad the rail mosfets are bypassed.
Those diode are between the + input and ground, not the - input.
Those diode are between the + input and ground, not the - input.
That appears to be a mistake I had not noticed. The input protection diodes should be between +IN and -IN.
Now does that make a difference to start up behaviour?
separate +/-15v to offset controller
No, this solution just hides the problem.
What i mean.
Amplifier's offset start from about 400mV if the separate power supply start up about 12'' before the main power supply.
Amplifier's ofsset start from 3 volt when separate power supply start up about 5'' before the main power supply.
To my point of view this act exactly as when common power supply was used.
Another find is that offset is 3v when i forgot to startup the separate offset controller power supply.
No, this solution just hides the problem.
What i mean.
Amplifier's offset start from about 400mV if the separate power supply start up about 12'' before the main power supply.
Amplifier's ofsset start from 3 volt when separate power supply start up about 5'' before the main power supply.
To my point of view this act exactly as when common power supply was used.
Another find is that offset is 3v when i forgot to startup the separate offset controller power supply.
Last edited:
What you can try to clear the situation - remove the opamp and ground the output pin, where the opamp's output was, to ground. Start it up and see the offset behaviour. I think, it will start from 3V. Then we know what to do 😉
Grounding the out pin of the missing opamp is equivalent to grounding R28.
That results in -IN pin being grounded via two routes, via R24 and via R28.
But I don't understand why that should give an output offset of 4V. What else is happening?
VZ, come back and explain.
That results in -IN pin being grounded via two routes, via R24 and via R28.
But I don't understand why that should give an output offset of 4V. What else is happening?
VZ, come back and explain.
It may be caused by the difference in input LTP's base currents due to significant difference in base-to-ground resistors' values.
It would not be a problem if the input devices would be the jFETs or the tube triodes, but is a problem with BJTs.
For a test, we can insert an electrolytic capacitor in series with NFB resistor, placed between the -IN and the signal ground.
It would not be a problem if the input devices would be the jFETs or the tube triodes, but is a problem with BJTs.
For a test, we can insert an electrolytic capacitor in series with NFB resistor, placed between the -IN and the signal ground.
But the base current is of the order of 6uA. That was confirmed some while ago and confirmed that the LTP was reasonably in balance (just within that tolerance of ±10%).
6uA across 22k is only 132mV.
changing one side to 11k only creates a difference of 66mV. That very approximately should become the new output offset.
6uA across 22k is only 132mV.
changing one side to 11k only creates a difference of 66mV. That very approximately should become the new output offset.