help about op amps and more

[mganzer]

Hi guys;
I'm try to figure out how determine if a op amp have high or low current delivery and high/low input impedance.
Anyone knows? I can get this info from datasheet?
I'm trying to apply a parallel opamp circuit (something like a howland current pump) in a overdrive circuit idea and want to mix opamp with high current delivery and low input impedance (master)(like an ne5534 or another) with another with low current delivery and high input impedance(to be determinated).

Another question: If an opamp have supply of 18v instead of 9v they have more headroom. An split supply with +9v and -9v gives same headroom or more? And if this split supply are applied for a circuit like zendrive: The negative supply of opamp is not conected to ground (directly) (assuming using a charge pump) in the circuit i apply the negative supply only in the opamp or entire ground section is negative?

Thanks

[RobA]

For the current and impedance questions, you pretty much have to go to the spec sheets. But, as a general rule, op amps behave like the transistors that make them. FET inputs imply higher input impedance and BJT outputs imply more current drive. There's more to it than that though and you will have to spend time with the spec sheets if you want to get in depth with it. TI has a mass of publicly available application notes on op amps too that are a great resource.

As far as headroom goes, 18V is 18V. The advantage in the -9V to +9V is that it simplifies many circuits. Cascaded filter circuits are much simpler if you don't have to have coupling caps in-between every stage. Also, some op amps just work better in a split rail setting. It allows you to filter the power to both rails at the op amp as well.

In a split power supply (dual rails, +9V and -9V), ground is still ground. The -9V is a separate power rail from ground. I won't say though that the only thing going to -9V is the op amp lower supply. There are other things that will get hooked to ground. But, ground is always separate from the -9V rail and the audio signal is referenced to ground.

[mganzer]

Thank you so much Sr.
I'm already download the spec sheets of a bunch of opa:
lf351, ne5534, opa134, opa604, tl051, tl071, tl081, tlc071, tlc081, lm741. Going print and read. I post the comparison between then.

About rail to rail supply this attached schematic will work? Can i refine the negative supply or charge pump do the job?

[RobA]

Once you are using split rail designs, you can filter the negative supply in exactly the same way you filter the positive supply (with the polarized caps and such oriented the right way, i.e. remembering that the ground is more positive than the negative rail). How much you do depends on the application. Good examples of how to filter for hi-fi applications can be found by looking into the DIY headphone amplifier community. Some of this stuff is directly applicable to clean guitar preamp circuits for example. Sometimes it's complete overkill. It all depends on the application.

On the schematic, you are looking at using IC1A as an inverting amp or filter structure? If so, you most likely don't want Vb. The + input in this case is referenced to ground.

Another note, my personal preference is for the LT1054 for the voltage converter. They have more current capacity are more efficient and have a switching frequency that is above the audible range. I didn't find that to be true of any of the 1044 or 7660 devices even with the boost pin used.

[mganzer]

Is a inverting amp. Schematic is anexed. Do i always use ground as reference in rail2rail?

[RobA]

For op amps, yes you almost always are going to be referencing your signal to ground and not a separate reference voltage. The question is a bit more complicated with transistors.

The synth world and HiFi audio world mostly use split rail designs. Looking into DIY sites for these areas could be really helpful for you.

The other thing is that most example circuits in op amp spec sheets and application notes are split rail. It's usually going the other way that's a pain.

Here is a link to a great reference on op amps,
http://www.ti.com/lit/an/slod006b/slod006b.pdf

It can get pretty dense, but there are tons of examples for all sorts of different op amp uses and you can often get a lot out of it by just going to the chapter that's specific to what you are working on. The historical overview is pretty cool too.

[mganzer]

This guide is great RobA! Thank you. You have some material about capacitors and filterings? I'm want to going deep on it too.

[RobA]

Filters are a big subject. It depends on what type of filter you want to build. The classic guitar filters were be done in some interesting/ingenious ways. I think the best way to learn those is just look at schematics of the  effects you are interested in and then breadboard the parts of the circuit to manipulate them and hear what they do.

There's a vast library available on the design of op amp filters. TI and National Semi put out bunches of app notes on filter design. All of it is available on TI's website now. Here is a three part app note called An Audio Circuit Collection. The first part is part of a bigger doc.
http://www.ti.com/lit/an/slyt023/slyt023.pdf
http://www.ti.com/lit/an/slyt145/slyt145.pdf
http://www.ti.com/lit/an/slyt134/slyt134.pdf

I came to doing audio from DSP, so my foundation in filters is actually in the digital world. The basic reasons for using particular types of filters is the same in both domains though. This site has a bunch of good info on the application of filters in a musical setting,
https://ccrma.stanford.edu/~jos

Capacitors are the basis for religious wars. Entire cultures have been destroyed by fighting over which capacitor is best.

ESP has pages on capacitors and filters along with much else in the articles and projects pages.
http://sound.westhost.com/articles.htm
http://sound.westhost.com/projects.htm

Geofex has tons of good info on guitar effect, especially in the "technology of" and FAQ.
http://www.geofex.com

[mganzer]

Thanks again for your great help. This article jumps on my eyes and helps me to ensure my point: I'm try to learn real eletronics and put all mojo bullshit aside.
http://sound.westhost.com/articles/capacitors.htm

Now i really think  about to get some  (full range value pf-uf) tdk mlcc (trought hole) capacitors from mouser in place of various types to simplify the work. Aparently these mlcc have various advantages over film types and i'm thinking maybe  in guitar-audio(or at least stompboxes application) have no major audible diferences. The major disavantage is dieletric absortion of ~.5%.
This is a good pratice in your opinion or have a whole bunch of capacitor types is better for experiments?

I will spent a big time read all these docs huh?

[RobA]

Yes, I think it's a good idea to experiment with caps.

MLCC's come in a variety of types. According to articles I've read, the only type that are appropriate for audio are the C0G varieties. This is noted in multiple spec sheets and app notes from places like TI and Wolfson. I had an article linked that did actual measurements and quantified what occurred, but it seems to have been lost on the site I had linked before. The main point though is that other types of MLCC's are not stable under voltage. The capacitance changes with changing voltage. When they have a bias voltage on them and a signal moving through them the actual capacitance moves with the signal. This leads to non-linear distortions and thats really not good. But, the C0G's are only available in smaller values. So, I stick with film caps for decoupling and for larger valued caps in filter circuits.

What it comes down to is that it's not possible to make a perfect capacitor. Each of the types has certain strengths and weaknesses. Using the type that best fits the job is good practice. I think the importance of that article you linked above is to not get caught up in falling for the expensive means better nonsense that pervades the marketing of audio caps. There are good film caps that do the job of audio decoupling cap very well (better than MLCC's) and are still quite cheap. The only reason I can think of for not using them is in a case like very small surface mount components where the physical size limits you to MLCC's.

[mganzer]

Fooling around right now i discovered a series of tdk mlcc are C0G From 100pf to 68nf:
http://product.tdk.com/capacitor/leadmlcc/en/documents/leadmlcc_commercial_midvoltage_en.pdf
Its just for information.

Select a good stash of components is driving me crazy. In Brazil is quite dificult and overpriced mission (except if you satisfied by chinese crap).
I want to select (with minimum rate of 25v):
Resistors 1206 package with 1%.
Resistors 1/4W 1%
Capacitors NP in wide pf range
Capacitors NP in wide nf range (plus 1uf)
capacitors polarized uf range alum (thinking also in tantalum smd in 1, 2.2, 3.3, 4.7, 10, 22, 33, 47 uF)
Series of handy diodes (rectifiers, silicon, germanium, schottky, zener)
Series of handy transistors (npn, pnp, fet)
Series of handy opamps (single, double, quad).
Use in breadboard and single sided layouts with smd components mixed with radial caps.

[RobA]

I think the C0G's are available up to 100n. That seems to be the limit. TDK have a line of film caps marketed under the brand Epcos that are good and cheap. Panasonic and Nichicon make good electrolytics in their FR and PS and FG series. I use the FR and PS caps powers supply stuff and the FG's in the audio path. I tested the FR and PS caps against a whole bunch of caps in a voltage doubler/negative rail generator circuit based on the LT1054 and they were measurably more efficient.

One thing to remember, almost all the electronics are made in the far east. Getting the stuff from suppliers over there can be a great price break for the exact same goods. You do have to be careful though -- lots of knockoffs too.

Avnet and Mouser both ship to Brazil I think. I have no idea how much they charge to ship there though. Avnet's prices in the US are mostly very good. For lots of components Tayda really is very hard to beat.

[oldhousescott]

Operational amplifiers are idealized devices, mathematically designed to "operate" with certain characteristics in circuit. Those design rules are:

-- infinite input impedance
-- zero output impedance
-- zero voltage between the + and - inputs

Actual components achieve this with different topologies internally, but essentially behave the same way in-circuit. Differences become evident at either end of the usage spectrum, whether for lowest noise and distortion amplifying small signals (like mic preamps), or when signals get near the rails as with many OD designs.

[RobA]

I think it's actually pretty important to be pedantic here, so here goes. Op amps are real world devices that are composed of transistors, diodes, resistors and capacitors. Those components behave in circuit exactly the way they would if you built the circuit from the discrete devices. The only practical difference electronically is that the components are very well matched to each other.

The idealized op amp equations are first order approximations of the behavior of an op amp. They are useful for getting the analysis started. But, you still need to get down to what the actual device is and does to design a circuit. Even in the process of designing an op amp itself, meeting the requirements of the ideal op amp equations would only be one of the design goals.

When working on pedals, either modding or creating new design, you are mostly working from known building blocks where you already know that an (any) op amp works in a certain circuit. What's important to the design or mod is what does a particular op amp give you versus some other op amp. What you need then is to be able to read the spec sheets. And, since there are a seemingly endless number of op amps available, it's good to know some general guidelines about what general families of op amps give you so you can narrow down the search.