Nurse Quacky unpleasant high frequency feedback.

[nguitar12]

So I finished Nurse Quacky build everything work fine and it is pretty usable. However it will produce some kind of high frequency feedback sound when the attack and threshold set low, which is pretty unpleasant to the listener.

http://vocaroo.com/i/s1ZCqN3ZNw5S
(WARNING: turn you speaker/headphone down to hear the clip.)

I have uploaded the sound clip. Normal setting from 0:00 to 0:14. You will here high frequency feedback after 0:15 which is low attack and low threshold setting with hard strumming.

How to get rid of these noise? Thanks

[Kipper4]

I'm not listening to that. Naway.

Schematic so as we dont all have to go look for it





You could try making the two 4nf7 in the filter bigger. try values upto 10nf.
You can make the 51ohms bigger. This is the minimum attack resistor.
Try a 50~100pf cap in parallel with the filters 470k NFB loop gain resistor.
Mess with the range pot?
After this I'm kinda stumped.

[anotherjim]

Feedback often means there is positive feedback going on somehow.
You did build the 4.5V supply for the +in of the filter amp?

I can see a possible way if the rectified envelope at the attack control is leaking enough AC signal, probably needs some accidental coupling between input and output of the whole circuit to make that into a feedback path.

Maybe the 51R in front of the attack pot could be bigger, so it equals whatever value that pot is at the point that it won't quite squeal.

[Kipper4]

Ok I listened to it.
Now might be a good time to show some pics of what you built because that's not normal behaviour for an autowah.

Do the debug thing, post some pics.
Lets try and get it sorted.

[nguitar12]

Feedback often means there is positive feedback going on somehow.
You did build the 4.5V supply for the +in of the filter amp?

I can see a possible way if the rectified envelope at the attack control is leaking enough AC signal, probably needs some accidental coupling between input and output of the whole circuit to make that into a feedback path.

Maybe the 51R in front of the attack pot could be bigger, so it equals whatever value that pot is at the point that it won't quite squeal.

http://3.bp.blogspot.com/-Sv0Do_EStC4/UwLzfFwkKOI/AAAAAAAAAXo/tnA1DYi9f1Q/s1600/NURSE+QUACKY+schematic.png

I build according to the above schematic. Yes I made the 51R bigger is will solve the problem at certain threshold but when the threshold set lower the feedback occur again. It seem like when the signal hit some point it will trigger the feedback

[nguitar12]

Ok I listened to it.
Now might be a good time to show some pics of what you built because that's not normal behaviour for an autowah.

Do the debug thing, post some pics.
Lets try and get it sorted.

I am so sorry it is on the breadboard and the wire is totally a mess and I don't thing anyone will get some hints from that

[duck_arse]

we love looking at breadboard photos. did you use LM358?

[Kipper4]

Ah I see your schematic and raise you.

Make double sure you did Not make the 470k feedback resistor go from the + input to output.

It should go from -(minus) input to output.

Such a mistake might make it oscillate.

[nguitar12]

did you use LM358?

Yes I am

Ah I see your schematic and raise you.
Make double sure you did Not make the 470k feedback resistor go from the + input to output.

Just confirmed I am connecting the 470k feedback resistor correctly.

You could try making the two 4nf7 in the filter bigger. try values upto 10nf.

Not yet tried this mod but what is the principle behind that? Thanks

[Kipper4]

I don't think changing the filter caps is going to alter this sound now. That was before i listened to the sample.
Upping the values of the filter caps would bring more of the bassier notes back in the filtered signal.

Surely you have made a mistake somewhere.

The way I see it. even if you had just the 51ohm attack resistor and no pot (effectively like you turned the pot anticlockwise)
It should not cause feedback. The attack resistance should merely lengthen or shorten the onslaught of the wah effect.

voltages check out?

[nguitar12]

I don't think changing the filter caps is going to alter this sound now. That was before i listened to the sample.
Upping the values of the filter caps would bring more of the bassier notes back in the filtered signal.

Surely you have made a mistake somewhere.

The way I see it. even if you had just the 51ohm attack resistor and no pot (effectively like you turned the pot anticlockwise)
It should not cause feedback. The attack resistance should merely lengthen or shorten the onslaught of the wah effect.

voltages check out?

I increased the 220R right under the 3904 it seem that the problem improved now. It seem that the sweep will not reach the "trigger point" now no matter how hard I strum the strings. Does it make sense?

[PRR]

> the wire is totally a mess and I don't thing anyone will get some hints from that

Messy wire is a "hint" to the CIRCUIT that output should sneak back to input and be amplified around again and again to infinity.

High-gain audio circuits (even fairly low-gain like this) really love to sneak-back and oscillate.

[duck_arse]

sometimes, some people, with some transistors, report better/more sweep if the transistor is in backwards, so C to the 220R and E to the 4n7//47k//4n7.

what does the voltage peak at on the 22uF?

[nguitar12]

sometimes, some people, with some transistors, report better/more sweep if the transistor is in backwards, so C to the 220R and E to the 4n7//47k//4n7.

what does the voltage peak at on the 22uF?

This circuit will work even  with a reversed transistor? Gonna try it later
I can't measure the voltage peak on the cap my meter is not responding fast enough to do this.

[anotherjim]

A BJT still works with E & C swapped, but with much less Beta (current gain). That can be an advantage when you have an input control range that's quite wide. High current gain results in the transistor saturating early in the sweep. Reducing base current with a bigger resistor doesn't always help, you can get reduced effect for the lower end down the control sweep. Lower transistor gain can be just the ticket.