1
\$\begingroup\$

I'm designing a color organ. My circuit so far:enter image description here

Quick description: Input signal is .2V pk-pk. Voltage after amp is 4V pk-pk. 2N7000 needs 2V to turn on, so it should turn on when the frequency I want is passed through the filter, but any other frequencies are decreased in amplitude.

My problem lies after C2 at node Va. What I expect there would be an amplified wave of my input centered on the x axis (if the y axis was measuring voltage.) Instead what I find is my amplified input shifted up by 1 or 2 volts. So even though my AC signal decreases at node Vb as frequency increases, it is still biased by that +1 or +2 volts, so the MOSFET stays on at all frequencies.

I've tried a number of things so far. First, I tried discharging the caps, but that seemed to have no affect. I tried adding a large resistor in parallel with C3, but that didn't help much either. I really don't know what's going on here or how to fix it, so any help would be greatly appreciated. Thanks in advance!

EDIT: I also chose large values for my filter because previously it was too small and the lack of impedance was causing a voltage loss at higher frequencies. I am also aware of the nasty clipping my amplifier gives me at the moment. I plan on fixing it later.

\$\endgroup\$
2
\$\begingroup\$

1000 µF is a ridiculously huge value to use as the coupling capacitor at the output of the amplifier, especially given the high input impedance of the filter(s). Assuming this is an electrolytic capacitor, its leakage current is undoubtedly enough to explain the bias you're getting.

Assuming that the net input impedance of your three filters together is on the order of 100 kΩ, a ceramic capacitor of 0.1 µF for C2 would give you a lower cutoff frequency of 16 Hz, which is plenty low enough for this application.

\$\endgroup\$
  • \$\begingroup\$ Thank you so much for your suggestion! I swapped out the 1000uF cap for a .47uF and it worked perfectly! Originally I thought that the cap I was using was interfering with my filter, so I opted for a big one to let all frequency through. I guess I didn't need that after all. Thanks again! \$\endgroup\$ – Gigaxalus Feb 27 '16 at 4:28
1
\$\begingroup\$

The node Va has no DC path to ground, therefore the voltage is not well defined. Usually a circuit simulator automatically inserts a conductance to resolve this problem and issues a warning.

\$\endgroup\$
  • \$\begingroup\$ Is there an easy way to fix this? \$\endgroup\$ – Gigaxalus Feb 26 '16 at 20:32
  • \$\begingroup\$ I'm afraid I don't really understand what you are trying to achieve. Generally, you would need a bias network similar to the one used for the bipolar (R1,R2). \$\endgroup\$ – Mario Feb 26 '16 at 20:35
  • \$\begingroup\$ What I would like is no bias. I would want just the signal frequency and amplitude to be responsible for turning on the mosfet. \$\endgroup\$ – Gigaxalus Feb 26 '16 at 20:37
  • \$\begingroup\$ You need a well defined operating point. That could be a large resistors that pulls the gate to ground. \$\endgroup\$ – Mario Feb 26 '16 at 20:45
  • \$\begingroup\$ I tried putting an 8M resistor in parallel with C3 to do that, but that didn't help. I tried with 1M, that helped a bit, but not very much. I tried with a 220k and that worked, but it stole too much voltage from node Vb, so it barely turned on. Maybe there is some other goldilocks value I have to find? Thanks for your help by the way, I appreciate it! \$\endgroup\$ – Gigaxalus Feb 26 '16 at 20:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.