interesting power up problem...

[giantsteps]

Hey guys...

Working on a tube screamer that has an interesting issue.

Sounds fine and everything functions...except...

When I connect a power supply or battery to the pedal, it takes 4-5 seconds before signal goes through the pedal...almost like it has to charge up...I've never seen this before...

Interestingly enough, the scope shows plenty of signal going through the pedal at all times during connection, but with an amp connected it fades up.

Any ideas?

Thanks,

Chris

[amptramp]

Is this an original tube screamer with the JFET switching or is it a true bypass modification?  The ones with JFET switching may require time for the floating gate to pull up and turn on.  Check the schematic here:



Two 2SK30A's are used.  When the left one turns on, the effect is bypassed.  When the right one turns on, the effect is functioning.  The gate goes high to turn the JFET on.  When you first switch the unit on, the voltage at the 1N914 diode cathodes in series with the JFET gates comes up with a time constant of a little over 47 milliseconds.  But the gate does not necessarily follow anything but leakage current when it is switched on and if the leakage is higher in the JFET than the diode, the gate may stay down for an inordinate time.

Some people who use this kind of switching add an op amp connected as a voltage follower to the JFET source whose output goes through a high-value resistor from the follower to the gate.  Since the voltage at the ends of the gate resistor is the signal voltage, some current flows to charge the gate, but this is a fixed voltage, so the current drain is constant.

[giantsteps]

Thanks! It has the JFET switching...turns out there was some DC on the output...

Best,

Chris

[amptramp]

Electrolytic caps are considered maintenance items, just like batteries but with a longer life.  If you are showing significant output voltage across the 10K resistor from the output to ground, replace the 10µF output cap.  You would need 100 microamps to move it by a volt.  That would be a lot of leakage.