Instrument input question for headphone practice amp

[ElectronHerder]

I'm busy designing a simple op amp based practice amplifier that will mix a music signal (from a PC, phone, whatever) with a bass guitar signal and then into a high quality headphone amp.

I've got a reasonable amount of experience with audio electronics, but next to zero when it comes to musical instruments.

For the bass preamp I have an op amp configured for a gain of 4 which feeds into a 50k level control. This will then be summed into the music signal and from there into the headphone amp. Is this a good amount of gain for a bass guitar?

My other question is do the guitar pickups need a bias voltage to work? I.e. do I need to feed a DC voltage back into the bass?

[antonis]

Hi & Welcome.. 

Gain of Bass signal can't be set as stand alone but in respect to music signal amplitude..
Value of Level control should be set AFTER the above setting..
Also, you'll have to identify music signal output impedance for appropriate setting mixing resistors values
(either for passive or active mixing..)

No DC voltage back to pickups unless you are realy looking for troubles..

[ElectronHerder]

Thank you for the reply antonis   .

The music and bass signals each go through their own 50k potentiometers before being mixed, so they can be adjusted individually. I'm doing passive mixing with 10k resistors. Does that sound okay?

Gotcha, no DC bias for the bass

[antonis]

The music and bass signals each go through their own 50k potentiometers before being mixed, so they can be adjusted individually. I'm doing passive mixing with 10k resistors. Does that sound okay?

It indeed sounds OK but a schematic should sound more OK..

[ElectronHerder]

Here ya go:

[antonis]

How many frets have your bass guitar..?? 
('cause it needs tooooooooo many of them to reach 3.4GHz, which is C3/R4 LPF corner frequency..)

Give me some time to modify your schematic a bit..
('cause there are some design "flaws"..)

[ElectronHerder]

Hmmm... given a velocity of 400m/s in the string and a frequency of 3.4GHz, the string would only need to be 56 nanometers long 

Edit: wait... 400m/s is faster than the speed of sound, which might create problems. So let's say 300 m/s. 41.2 nanometers.

[antonis]

R2 value lowered due to bipolar IC1 Inputs
(you may raise its value for higher input impedance but I shouldn't go more than 470k, say..)
C1/R2 & R1/C2 values are apart enough for both filters corner frequencies interaction..
C7 is added for  DC gain rolling off to unity - not essential for particular circuit, due to low gain and -almost- zero pin 3 DC level but better safe than sorry..
R3/C7 HPF has  a corner frequency of about 10Hz where R4/C3 LPF one is of about 34kHz..
(you can set your own bandpass margins by altering capacitors values..)
Although 100k mixing pots should be more preferable, 50k should also do the job..

Last but not least.. I presume we're talking about symmetrical dual supply, aren't we..??
(in such a case, output DC big blocking cap(s) could be avoided by making R2=R4//R3..)

[ElectronHerder]

That all makes a lot of sense - thank you so much for your input antonis !

You are correct, +/- 12V symmetrical supply. I generally stick DC blocking caps in anyway, as no op amp is perfect and I hate scratchy pots. Plus I have a drawer full of 10uF polypropylenes .

If the input impedance of the headphone amplifier is ~20k what should I change the summing resistors to?

[willienillie]

a drawer full of 10uF polypropylenes

Wow, really?  Who makes those?  How big are they?

[antonis]

If the input impedance of the headphone amplifier is ~20k what should I change the summing resistors to?

In theory, yes..

In practice, I shouldn't lower their value less than 5k say 'cause this should be the only "blocking" resistance for IC1/Line1/Line3 very low output impedances, when Level pots set all the way up..

A 5532 as a dual buffer could solve the above issue..

[ElectronHerder]

Wow, really?  Who makes those?  How big are they?

A number of manufactures make them (Wima, Epcos, Kemet, Vishay to name a few). They're 27.5mm lead pitch and 15mm wide so maybe a bit big for use in a pedal. The circuit I'm working on is going into a desktop enclosure to space is less of an issue. Here's an example: https://www.digikey.com/product-detail/en/wima/DCP4I051006GD2KSSD/1928-DCP4I051006GD2KSSD-ND/11314401.

You can get the same capacitance in a much smaller package if you're happy with polyester instead of polypropylene though: https://www.digikey.com/products/en?keywords=MKS2B051001N00JSSD

In theory, yes..

In practice, I shouldn't lower their value less than 5k say 'cause this should be the only "blocking" resistance for IC1/Line1/Line3 very low output impedances, when Level pots set all the way up..

A 5532 as a dual buffer could solve the above issue..

Hmm. So left and right music signal into unity gain buffers and then into the amplifier ?

Edit: would anyone else be interested in building a practice amp like this? I'd be happy to post the PCB files once its finalized if others would like to build one.

[antonis]

So left and right music signal into unity gain buffers and then into the amplifier ?

Yeapp..
(maybe not essential but all time classic prescription for sleeping in peace of mind..)

would anyone else be interested in building a practice amp like this? I'd be happy to post the PCB files once its finalized if others would like to build one.

We havent' see any amp (practice or not) till now..
(specs plz..)

[ElectronHerder]

I'll be using a TPA6120-based composite headphone amplifier. If I remember correctly THD is around 0.0003% at 1kHz into 32ohms, delivers around 1W before the onset of clipping. I'll post plots when I get home. However, the chip is really difficult to solder without the proper tools as it has a thermal pad on the underside.

I'd be happy to design a PCB with a simpler amplifier if anyone is interested. Something like an OPA551 / 552 (200mA so plenty of current, great sounding op amp), or maybe OPA627 / BUF634 combination.

[ElectronHerder]

Distortion and frequency response plots for the TPA6120 amp into 32 ohms:





The same for NE5534 / BUF634 amp:

[antonis]

Good old 5532..!! 

What exactly happens at 50Hz..??

[ElectronHerder]

Good old 5532..!! 

What exactly happens at 50Hz..??

It's 50Hz mains noise - these plots were made with the amps connected to a straightforward bench power supply 

[antonis]

It's 50Hz mains noise - these plots were made with the amps connected to a straightforward bench power supply 

Ahaaaa..!! And your bench PS utilizes half-wave rectification..?? 

[ElectronHerder]

Ahaaaa..!! And your bench PS utilizes half-wave rectification..?? 

Apparently 

It's a decent Rohde & Schwarz unit but the noise still gets through