so im using the headphone amp whammy as my preamp and have noticed that the thing works and sounds better at gain of 5 or greater, from probably the opamps being more stable. But x5 is too high for my setup and i need to attenuate without changing the gain of the circuit.
The setup of the whammy is as follows-
10k pot>input cap>100kohm to gnd>1k ohm signal/100k ohm to ground>opamp input pin
I changed the ratio of the 1k ohm/100k ohm at the opamp input to 70k/30k to work as an Lpad and the bass became feeble, possibly because the input wants higher value resistor to gnd to set its impedance?
I suppose i could put an input resistor before the pot to attenuate it but if theres no harm then id rather attenuate at the opamp's input to reduce all the preceeding noise.
What should i do?
The setup of the whammy is as follows-
10k pot>input cap>100kohm to gnd>1k ohm signal/100k ohm to ground>opamp input pin
I changed the ratio of the 1k ohm/100k ohm at the opamp input to 70k/30k to work as an Lpad and the bass became feeble, possibly because the input wants higher value resistor to gnd to set its impedance?
I suppose i could put an input resistor before the pot to attenuate it but if theres no harm then id rather attenuate at the opamp's input to reduce all the preceeding noise.
What should i do?
Redundant resistor is redundant.
Relocate 100 pF C2/C7 to go between opamp pins 2/1 and 6/7, respectively. Should make things a lot more stable. You may not even need that much.
If possible, add a bit of resistance in series with the outputs, like 1-10 ohms.
While your basic idea was correct, I would rather avoid low division ratios if you need high input impedance, output impedance becomes rather too large. Instead, you can try populating or adding the resistors indicated as R39 and R40 with anywhere from 10 to 47 kOhms.
Which opamp type are you using in there? The schematic indicates OPA2604, while the kit seems to ship with an RC4580.
Optional tweaks:
If you wanted to accommodate an opamp with highish input bias current (e.g. 5532), the feedback would have to be changed from just R to R || (R-C) in an attempt to balance out DC resistance seen by both opamp inputs.
I am also fairly sure that using the same value (499 ohm) gate stopper resistors for both P and N FETs does not yield optimum output stage distortion performance... usually the Ps have higher G-D capacitance and require correspondingly lower driving impedance.
An optocoupler as a thermal runoff and bias control device - someone must have had a genious spark. Now turn that upside down for a total new class A design!
R39 r40 makes most sense, thank you.
However with embarassment i must say from my suboptimal layout im getting transformer noise from the output
Something before the opamp must be pickung it up because if i ground the opamp input the noise goes away. Thats why id rather attenuate at the opamp to kill the noise.
Im using lowly 5534 btw. And even with the decomp cap at pin 5-8 (22pf) it still sounds vetter at gain 5+. Maybe its the change in impedance? It just sounds smoother and mkre dynamic and once you notice the change its hard to go back
However with embarassment i must say from my suboptimal layout im getting transformer noise from the output
Something before the opamp must be pickung it up because if i ground the opamp input the noise goes away. Thats why id rather attenuate at the opamp to kill the noise.
Im using lowly 5534 btw. And even with the decomp cap at pin 5-8 (22pf) it still sounds vetter at gain 5+. Maybe its the change in impedance? It just sounds smoother and mkre dynamic and once you notice the change its hard to go back
I'd rather be fixing the root cause, as this may be indicative of a larger problem. Can you show a few pictures of your input and volume pot wiring and the layout in general? In which range of the volume control does the "transformer noise" appear? The problem could be
a) electrostatic, coupled from transformer windings into sensitive high-impedance circuit nodes (if so, maximum hum would appear near maximum volume setting) - to be fixed by better shielding of said problematic high-impedance nodes and/or the transformer itself, as well as transformer orientation (you can experiment with this).
b) magnetic, coupling into internal ground loops (if so, maximum hum may appear at minimum circuit impedance, i.e. at lowest volume and at maximum)
c) power supply noise injected into a non-zero ground return impedance by rail decoupling capacitors (another that should get louder as you turn up - you can fix this by establishing a separate, dedicated power ground return with wires running back to the power supply alongside audio ground; that said, the 47 ohm resistors in series with the rails should normally be keeping this issue at bay already)
You may want to draw the grounding layout for the whole circuit (input jacks and housing included). This would help spotting ground loops.
Last edited:
Thank you for the detailed response. I believe my problem is 'b' magnetic because not only for the symptom u wrote but when i move away the transformer the noise disappears completely. I am using a rather large-ish ei for the job and bolted right next to the pcb which is probably not smart..
Also isnt 47ohm kinda high? Ive always stuck with 10ohm but ill swap it out for 47 in this circuit
Also isnt 47ohm kinda high? Ive always stuck with 10ohm but ill swap it out for 47 in this circuit
How many VA? EIs are well known for their stray fields. Location and orientation are going to matter (experimentation required) but I would prefer not to have any ground loop to couple into to begin with.
My first suspect to look at would be the input wiring. Does it follow the build guide? You should be using insulated jacks, and if you didn't have 2-conductor shielded audio lead but rather two separate cables, these should be kept twisted together (the same actually extends to outside cabling).
Are you using one of the premade boards? Just so we're on the same page. Going by pictures, those seem to be connecting input and output ground together, so are best used with star ground on board and insulated input/output jacks. If you have a board layout of your own that would potentially add a whole bunch of variables.
The voltage drop of <200 to a little over 400 mV depending on opamp strikes me as anything but critical... in fact given the relatively high supplies typically suggested (e.g. +/-21.5 V... explains the OPA2604; note: extra heatsinking may be a good idea), you may easily be able to afford 100 ohms.