Need a little help with VREF

[NC30]

Hello people,

Nice forum you've got going here. I'll take a good look around and surely stick around. 

Anyway... I have a query regarding the VREF point in an overdrive circuit. I'm used to seeing it before transistors and op-amps, for reasons I understand, but the pedal I'm studying has VREF connections after any boosting, at points I would normally expect to see a ground connection (filter, output level). What exactly is this connection for?

Can I post the name and schematic for this commercial pedal? I'm kinda unsure after reading the forum guidelines.

Any help appreciated. Cheers.

[Earthscum]

VRef is (note: usually) half your supply voltage. This gives the maximum amount of signal swing throughout the circuit.  Since most of the stages will want to bias around this point, you can use a VRef to ensure they are all biased along a common rail.

I'm guessing that the first couple stages are "self biased", in that there is a resistor network set up to cause the input to settle at a certain bias, independent of the rest of the circuit. An example of that would be if you used a simple transistor boost in front of some op amp circuit. The "boost" would probably be self biased for a couple reasons (feedback, needs a different VRef, etc) while the Op amp would most likely want a half supply reference. Another example would be using CD4049's and Op Amps. The 4049 will self bias, and (again) the op amp would want half supply bias.

That help at all?

(also, you can name the circuit, I believe there's no harm there... we can usually look up the schem if needed, Free Info Society has a nice database).

[NC30]

The circuit is the 1992 Marshall Bluesbreaker overdrive pedal. Straight forward in most respects.

I understand why a transistor and an op-amp will need biasing, but surely after that, you would want to remove all DC content?

I am quite new to this, by the way.

Cheers for the help.

[Earthscum]

Aah, I get it... ok, you are wondering why the Tone Stack section goes to VRef instead of Ground, most likely? If you follow the output of the Op Amp, follow it across all resistors, and you'll notice that you actually have a DC coupled connection to ground, so you use a VRef to keep the output from swinging to one rail more easily that the other, and keep symmetry. There's probably a lot more to it. The newer DS-1, for example, uses a 1k to ground. This causes some kind of loading on the chips they used, leading to a smoother sound.

If you see something connected to VRef that looks curious, just follow the DC path. If you see, say, a T-filter (resistor-(cap to ground)-resistor), the cap will sometimes be coupled to the VRef. I think it helps with loading symmetry. I'm no EE by any stretch of the imagination, so hopefully it kind of makes sense (and definitely not any kind of official explanation).   

[NC30]

This is starting to click. So, in a way, that VRef IS temporarily the "ground"...

So it would be roughly the same if the DC was decoupled immediately after the IC and the VRef points were replaced with ground... Right? And you think it's done this way instead to better retain symmetry?

Cheers.

[Earthscum]

See if I can get my head on the cap thing: If you have a t-filter as I described above, and your cap is going to ground, the cap is always going to have a  charge of 1/2V+, and it will swing up or down. This is a case you would want if you are using an electro cap, or any other polarized cap (especially Tantalum). If you have non polarized caps, and you connect them to VRef, then they will hold no charge. In this case, the cap charges up, loses charge at VRef (or "zero-crossing"), and has to charge back up again on the swing the other way.

If I remember this correctly, the way a cap discharges to zero and flips charging poles keeps impedance between the positive and negative swings symmetrical, where a partially charged cap will have an "average" impedance, but the swing towards ground has a different impedance than the charge towards V+. () I do know this issue comes up once in a while with Tube Screamer type circuits where there's a question as to whether the decoupling (shaping) cap in the Neg feedback network goes to Ground or VRef. You might try digging a bit through some of those threads.