A/B Box with Buffered Outputs using TL072

[GarryGirthOak]

Hey guys, i tried searching the web looking for such a schematic, haven't really found one sooo i created my own. The idea of the circuit is that I can switch between 2 outputs with separate volume controls which are buffered. I've been told on another audio website that the buffer needs to go after the volume pots.

Question is.

1. Is it electrically okay on principle (disregarding component values)

2. If not, what needs to be changed. I won't accept a answer that involves changing the OP-AMP.

3. Is there other ways of executing this with a TL072.

[GarryGirthOak]

whoops made a mistake.

[duck_arse]

hello and welcome, GGO.

to post images, they must first be hosted somewhere on the web, somewhere like imgur (which we all like) or somewhere else like botophucket/imageshack/tinypic/dropbox/googoo holes, there will be others (which some many of us don't like, for the trouble they cause). then, you take the link your hosting site provides you for each of your uploaded/hosted images, and paste them between the img tags provided by the "img" button, just above the post reply box. my icons have disappeared for a while .....

here it is , back now.

of course, you will have sized your pics to sensible dimensions, both pixels and megabytes, cause we don't all have massive download limits. ("he has issues".)

if you search this forum for A/B box or splitter/blend or similar, you'll prolly find plenty of the blocks to do what you want. people usually just shove the blocks into different orders to suit their needs.

[edit :] imgs icon's back.

[Derringer]

the way I did it, I used a TL074 and added in a bass cut and treble cut circuit on the input using two opamp sections.
The output of that went to the 3pdt switch where I only had one Led to light on the side that I designated as my "red" channel (more volume)
The output of the switch though then went to one of the other remaining opamp circuits. I had one voltage divider for the entire circuit instead of the separate ones you are using. The opamps were set up non-inverting as you have but I used a 10K volume after each opamp ... basically where your R10 and R11 are. I see no reason why not to have the volume controls at the end of the signal.

I see one, maybe two, issues with your circuit. You're going to need capacitors between the switch and your voltage dividers because without them, you'll be putting your reference voltage to ground and something is going to malfunction and get hot.

When I built mine, i grounded the unused output like you did. But i wasn't sure if this meant that i could do away with an anti-pop resistor, so I soldered some 2m2 resistors just in-case. And speaking of anti-pop resistors, you can remove R1. That part of the circuit isn't being switched so R1 is not necessary.

edit: on second look, I don't think your volume control will work as wired. As you rotate the volume control, it's going to change the reference voltage on the non-inverting input from about 1/2V down to 0V ... which again is just not going to work right. I think you should just use 10K volume pots after the opamps ... again like where you have R10 and R11. Make C2 and C3 something big like 1uf or larger

[PBE6]

This circuit looks like it will have some problems. Due to the position of R8, R2 and R8 are now in parallel which means the bias on U1 will change as VOL1 is changed. Similarly with R5 and R9. Also, you'll have a DC voltage gradient through R8 and R9, which could cause the unit to crackle as you adjust the volume. The other bits and pieces seem ok.

I think the circuit can be simplified significantly as follows:

The opamp buffers the signal, and the output is switched between the two volume knobs. LED switching is not shown, but can easily be added. If you're using a dual opamp, you should buffer Vref as well:

This helps keep Vref steady, but it's also the proper way to terminate an unused opamp to minimize issues.

[GarryGirthOak]

The Main issue i am having is the fact i only have a Dual OP-amp at my disposal and i want to utilize both op-amps. But yes the bottom image from PBE6 was the correct manner to design this kind of circuit. So basically what this tells me is to pretty much Move the switch infront of the op-amp.

I have  had serveral designs for this project. The other one utilized one op-amp to act as a booster using neg feedback, the other was used as a buffer, however is it possible to feed the output of a booster circuit into the buffer? Then feed the output of the buffer into the switch circuitry. I'll post a schematic of what i mean

However the switch design i got from a simple A/B box which grounds the inputs but also allows me to switch between to LED's.

[MrStab]

personally, i always try to go for volume controls that are sandwiched between a buffer on either side (as turning a bare pot will increase series resistance~output impedance). that'd double the number of opamps you need, but it could save you headaches one day, from unwanted interactions with other circuits. that said, quadrillions of pedals have the volume pot exposed to the outside, and most people don't have problems.

[MrStab]

just after posting, i realised i had this old post open in a tab, which could help you decide how to minimise:

http://www.diystompboxes.com/smfforum/index.php?topic=81657.msg676814#msg676814

[GarryGirthOak]

[GarryGirthOak]

The Reason for R7 and R8 is because the person who im trying to design this for wants both outputs to have seperate volumes for each output. I was thinking of having R6 as variable but i heard this would cause problems.

[GarryGirthOak]

Orrr, would this version work were i am feeding both inputs of the op-amps at the same time, then making R6 a variable resistor/gain then feeding that after C4. The top op-amp acts as a boost, while the bottom op-amp is a buffer/voltage follower.

[GarryGirthOak]

Anyone?

[Derringer]

how much gain do you want?

for instance, if you make R4 = 1K and make R5 into a 10K pot wired as a variable resistor, you would end up with a max gain of 11
then you could just omit C2 and R6 so that your bias voltage reaches the non-inverting input of the second opamp (it won't work as drawn because there's no bias voltage reaching the non-inverting input due to the blocking capacitor C2)

But Like PBE6 pointed out, you really don't even need that second opamp in the signal path. The output from the first non-inverting opamp is already buffered. You should just use the "other" opamp to buffer a vref and change up your biasing arrangement to comply.

and again, you do not need R1. You don't need a pulldown resistor here because the input of the circuit is not being switched. Your input impedance is being set by the parallel resistance of R2 and R3.

[Transmogrifox]

I honestly like this first general scheme the best.  It makes the output of both channels low impedance so you can drive long cables (if this is an issue).  You can handle the input impedance just by using large (like 1Meg) pots.

The only issues I see with it as it is drawn is you're set up for quite a bit of switch pop.  The best plan is to set the switch out ahead of the decoupling cap and tie signal path on all ends of the switch to ground through something like a 4.7Meg resistor.  Each op amp will have its own decoupling cap on its input.

There is a problem with how you show the pots wired since they pan between 4.5V and ground, which changes the bias as the pot is turned.  You would want the pot to go one leg to the input decoupling cap, wiper to op amp input, and other leg to the 4.5V resistor divider.   As long as the DC change across the pot is minimal then the pot won't crackle when you turn it. 

To avoid the chance of pot crackle, you put it on the grounded side of the input decoupling cap, then a fixed value resistor to bias the op amp from the resistor divider on the other end.

Just my 2 cents worth -- If I were doing it for myself your first idea is really close to the way I would have done it to keep output impedance low.

Also, add something like a 100 to 470 ohm resistor at the output of the op amp.  This protects the op amp from going unstable at high frequencies due to cable capacitance, etc and also reduces the probability of RF energy from being demodulated in the op amp outputs and becoming audible.

[GarryGirthOak]

Well here is the final Schematic.



I should've checked this forum on removing R1

R11 is removed.

This circuit works well. It sounds at least musical, but however to my liking cleans up the mids but retains low end. Here is also a Veroboard layout.



I am thinking of adding some unusual modifications. 1 is to have a SPDT switch after each "gain" pot (R7/R8) which engages/bypasses a simple low pass RC filter with 12 different positions, each one adding higher values of capacitance to ground.

Mod 2 is to have the ability to link one output to another. The application of this is to simply be able to instantly change the output of your signal into one amp. However the way i am wiring it, you would literally be able to switch out 1 into out 2/ Out 2 into Out 1. I can't see why there would be a problem as this switch is after the one that selects the output. So there will never be a time where both pots are sending signal to the same output.

I will consider redesigning the circuit Transmorgrifox quoted. Driving long cables wasn't a issue. The box was purely designed for being able to have a A/B box with adjustable outputs going 10db past unity. When i drew the first schematic i was very naive to building anything more complex than just a bazz fuss.

[Derringer]

cool man,

for your variable high pass filter idea, instead of a 12 position switch which is effective but takes up a lot of space in an enclosure, I'd suggest you take a look at the input on this circuit that Gus made a while back.

http://www.aronnelson.com/gallery/main.php/v/gus/plus.GIF.html?g2_imageViewsIndex=1

the image is big, so after you click the link, just right click on it and select "view image" which fits the image into the browser frame (at least it does on my browser.)

You'll need a dual opamp to achieve each variable high pass filter your make. With something like a 500K bass pot, a 10K limiter resistor and a 0.01 uF cap you'll have a -3dB roll-off range that goes from about 30hZ up to about 1.6kHz

[GarryGirthOak]

Thanks for that schematic. I might use that in future builds I wanted to ad a little uniqueness to the pedal by having a rotary switch. However the Filter is low pass, not high pass, which is what i've normally seen varitones used for.

Might not be a bad idea, could i perhaps use this on the input?

[Derringer]

doh' ... you did say low pass.

And yeah, if you like rotary switches then go for it.

but you could still use the dual opamp setup, stick a 100K pot or larger wired as a variable resistor in series between the output of the first opamp and the input of the second and place a cap to ground between the pot and the second opamp's input. You wouldn't even need to add another biasing resistor because the second opamp's input would be DC connected to the 1st opamp's output that already produces the bias voltage.

so if you used just a 100K pot and a 0.01 uf cap, you could have a -3dB rolloff from above infinity Hz (if you don't use a limiting resistor) all the way down to the rolloff starting at 160 Hz.  Might be a good idea to use a limiting resistor like 1K or 470 Ohm though so that the end taper of the pot doesn't get all wild.

On my AB pedal, I incorporated both of these circuits, the HPF and the LPF into the front end between a pair of opamps for some subtle EQing when neccessary.

[GarryGirthOak]

Cheers Derringer! I'll will definitely add this modification in another version, also cheers for explaining to me how to do it and including the values. At least i can make use of the other op-amp instead of terminating it. I spent the last week trying to get the right values for the main circuit.

To install this mod, would i remove the jumper after R2/R3 to the inveting input of the second op-amp. Also would i install the Pot before C2 or after? finally where does the output of the second op-amp connect to?

I am learning allot from this project

In this version i just have a simple SPDT switch with a .01uf cap on one lug and a .068 cap on the other where C1 is. The larger cap is for "bass" and the smaller cap is for guitar, works well.

[Derringer]

I am learning allot from this project

Awesome! Always a good thing!

modifying your schem is probably the easiest way for me to explain so ....