yes.
shorted > 50k pot > amps
Which made me look at the input wiring... looks ok at first glance.
Can someone give me a sanity check. If input is shorted to ground, at any position of the pot, should the resistance (the amps) between input and ground be zero? Its not in my case. its goes as high as 10k.
shorted > 50k pot > amps
Which made me look at the input wiring... looks ok at first glance.
Can someone give me a sanity check. If input is shorted to ground, at any position of the pot, should the resistance (the amps) between input and ground be zero? Its not in my case. its goes as high as 10k.
Last edited:
Read the part "Balancing the Input Bias Currents" @ Working with Cranky Op-Amps or google about input bias currents and bipolar input opamps. With JFET input opamps this problem doesn't exist.
Even in a perfect simulator world the second of the circuits below will have a DC offset varying between 0 and 100 mV depending on the volume setting if the bias currents aren't balanced. In real life it's even worse.
The 1st schematic is how such a circuit could look like. Note that the pot is now 1k, so the input impedance will be quite low. You could tenfold the resistor values and use a 10k pot, but that solution has it's own problems with a much higher noise level and problems finding small enough comp caps.
Another solution is the 2nd schematic with a CCS to balance the input bias current. Note that with this method there's no need for input caps. I just read in a thread here that a JFET at this low current wont act as a good current source. I can't hear the increase in noise they were talking about, and when simulated it's barely noticeable. I don't find it deteriorate the sound at all.
Even in a perfect simulator world the second of the circuits below will have a DC offset varying between 0 and 100 mV depending on the volume setting if the bias currents aren't balanced. In real life it's even worse.
The 1st schematic is how such a circuit could look like. Note that the pot is now 1k, so the input impedance will be quite low. You could tenfold the resistor values and use a 10k pot, but that solution has it's own problems with a much higher noise level and problems finding small enough comp caps.
Another solution is the 2nd schematic with a CCS to balance the input bias current. Note that with this method there's no need for input caps. I just read in a thread here that a JFET at this low current wont act as a good current source. I can't hear the increase in noise they were talking about, and when simulated it's barely noticeable. I don't find it deteriorate the sound at all.
Attachments
Last edited:
yes - but my casing is a little intricate for this one. I was hoping i had made a glaring error somewhere. still wonder if i ts my input wiring. I am using a stepped pot that has no wiring diagram, but the way i have wired it seems sensible on the mulitmeter.
The two input pins of an opamp need to have the same resistance to source the same current to develop the same voltage. It's the input offset current in the specification.
The diamante behaves similarly.
Even when the input transistors are perfectly matched. it's the input offset voltage that ultimately generates the output offset voltage.
If you make the resistance conditions on the +IN identical to the -IN and hold the matched transistors at the same temperature then the input offset voltage at both input pins will be identical and the output offset will be near enough zero.
Look at your circuit diagram and see if the +IN is actually identical to the -IN.
The diamante behaves similarly.
Even when the input transistors are perfectly matched. it's the input offset voltage that ultimately generates the output offset voltage.
If you make the resistance conditions on the +IN identical to the -IN and hold the matched transistors at the same temperature then the input offset voltage at both input pins will be identical and the output offset will be near enough zero.
Look at your circuit diagram and see if the +IN is actually identical to the -IN.
Last edited:
You still have to balance the input currents with one of the three methods mentioned: "balanced resistors", a current source to the input pin or a servo.
I still think a JFET as a current source is the simplest, cheapest and best solution. The surgery is minor and could easily be reversed.
Last edited: