Cap values - materials

[Keven]

so i'm going through the BOM for a 6 band EQ from general guitar gadgets. and i can't seem to find that elusive .39uF film cap. all i see is .33uF. i'm sure it won't matter much, but in this case, it's for pinpointing the actual frequency each band uses, there's a boost cap and a cut cap from what i remember from electronics, if the values are off from a tiny bit, it won't matter much right? it'll just be a broader EQ band and that's all?

also, as far as polarized capacitors go, i'm wondering if i should try tantalum caps for this project instead of aluminum, anyone know anything decent about them? i'll actually order both the tantalum and aluminum, they seem to be mostly Crf caps and power supply filter caps (they're the 10uF caps in this schematic, http://www.generalguitargadgets.com/pdf/ggg_eq6b_sc.pdf ) there's also a pair of them by the opamps.. maybe tantalum just by the Opamps and aluminum by the electric current



but maybe some tech guy here can enlighten me on how they work, i'Ve read they perform better on higher frequencies, but have yet to find a good electronics handbook about that!

thanks!

[jpfamps]

Farnell sell 0.39 µF film caps.

The circuit works by using a simulated inductor, called a gyrator circuit, working in concert with a cap and resistor to form an RLC filter.

Using a 0.33 µF will raise the cut/ boost frequency, although whether this will practically make much difference, who knows.

Tantalum caps are used because: they have lower ESR at high frequencies (although if you are really bother by this you can use a lower value film cap in parallel with an aluminium electrolytic); are smaller; don't dry out so shouldn't change in value over the life time of the equipment; and allegedly more tolerant of higher temperatures.

They are more expensive than aluminium electrolytics, and do not tolerate reverse polarity very well. They tend to fail short circuit. So if you do use them I would advise making sure that the wrong polarity connot be applied to your pedal (a good idea any way). It was also harder to get ROHS compliant tants, although I expect that this is now changing.

My personal experience of working with equipment with tantalum caps in, is that they often do fail, although I can't obviously say why.

[gwEm]

tantalum capacitors do give better sound quality, and are smaller. they also dry out less than other types of electrolytic capacitors. you have to watch the polarity on these capacitors, like any other electrolytic, though the actual polarity can be harder to identify.. they also work for a time with the polarity reversed, which an fool you into thinking everythings ok.

reliability with heat is meant to be better and they are more reliable in general, though they will fail faster when charged or discharged with high currents. this may be why they fail in valve circuits more..? in low-voltage/solid state stuff i've never seen one fail if inserted correctly.

[jpfamps]

I've seen tants fail in old pedals. I always assumed due to reverse polarity, which can be caused by various component failures, although is usually pilot error.........

Usually replacing the part with a modern aluminium electrolytic is fine, which I expect is due to the fact modern electrolytics are much better than they older parts.

In valve gear I've only seen tants in the solid state parts of amps, mainly in switching circuitry.

Tants also can have better tolerance than aluminium electrolytics so were used where the value of the cap was/is more critical.

If I wanted the best possible cap, I would use a film cap, unless space limitations and/ or cost were prohibitive.

[Keven]

even a film cap for polarized purposes?

also, if i want the best possible cap for the CA CB area of the schematic, is it worth it to dish out on sprague orange drops? or will some panasonic low volt caps work just as well? i can't get some of the lower uF values on the spragues so i'm getting panasonics for the smaller ones, but on most everything else i can get the values i want.

also, is the TL074 opamp the best i could get or could i just get a bit better performance with another one? it's a matter of getting the best components i can get for this build, as i've got another one built with off the shelf parts and they're nice, but i'm curious and want to learn/compare, hence the tantalums for this build.

[jpfamps]

even a film cap for polarized purposes?

also, if i want the best possible cap for the CA CB area of the schematic, is it worth it to dish out on sprague orange drops? or will some panasonic low volt caps work just as well? i can't get some of the lower uF values on the spragues so i'm getting panasonics for the smaller ones, but on most everything else i can get the values i want.

also, is the TL074 opamp the best i could get or could i just get a bit better performance with another one? it's a matter of getting the best components i can get for this build, as i've got another one built with off the shelf parts and they're nice, but i'm curious and want to learn/compare, hence the tantalums for this build.

For the 10 µF caps, the polarity is only an issue because electrolytic caps don't like to be reverse biased. Film caps don't have a polarity so this is not an issue.

For decoupling the power supply (the 10 µF caps to ground from A & B), I expect you won't here much difference with different cap types as the main purpose of this caps is to prevent HF oscillation in the op amps. Op amps are very good at ignoring power supply ripple on their rails, so I would think that 10 µF is over kill here anyway. In fact you might be better off with a smaller film cap, say 100 nF, which will have lower impedance at HF than an electrolytic.

The other 10 µF caps in the signal path can be replaced with 10 µF film caps, and these should give better performance. However you might find that you can use cheaper, say 2.2 µF film caps, without any affect on the bandwidth of the pedal. To reduce low frequency distortion in electrolytic capacitors larger values are used. The output stage driving a 10k external load (ie quite low) will start to have a significant roll off below 3.2 Hz! Using a 2.2 µF film cap would raise this to 14.4 Hz, which is still below the theoretical limit of human hearing (20 Hz) and well below the lowest note on a bass guitar in standard tuning (40 Hz).

I would not use Sprague Orange drops.  There is no need to use very high voltage caps here.

The Panasonic caps will be fine. In fact the spec of many PCB mount parts are much better than many axial components due to the economy of scale. My guess is that you will notice more difference by using film rather than electrolytic caps, than between different types of film cap. Polypropylene film caps are generally regarded as the best of the "non-exotica", ie sensibly priced caps.

Regarding op amps, the TL07* series are JFET input op amps. In practice this means they have very low input current, so will not have significant DC offset issues. This probably isn't an issue in this circuit. TL07* series also have fairly low current requirements.

The main issues with TL07* op amps are that the are: relatively high noise, relatively high distortion at higher frequencies, and will invert the phase of the signal if the output is driven to near the negative rail, which will obviously sound terrible. Having said that, unless you are building a distortion pedal, then you it is probably advisable to avoid any clipping in the device anyway.

The NE5534 op amp is generally regarded as the best of the sensibly priced opamps for audio. It is lower noise and distortion than the TL07*, however I don't think you can get this in a quad package. It also consumes more power than the TL07* series.

I expect noise performance to be the biggest issue with op amp choice.

Incidentally the 100k resistor on the input will be a significant source of noise and I might be tempted to experiment with a lower value here.

[jpfamps]

I should add that if you want to experiment with op amps I would use IC sockets so that they can easily be swapped over.

The best advice is to have a go and see (hear) what happens.

The great advantage about experimenting with pedals is that the worst than can happen is that you will kill some quite cheap parts.

[Keven]

excellent advice! thanks! this is for an knob driven EQ pedal that i'll stick after the preamp in my tube amp so i can get better midrange response out of it all, so i'm bound to be using a very narrow frequency band and a fixed volume pot to even it all out. it's also alot easier to use only panasonic caps as they have a better choice and are really cheap.

i'll experiment with different values 'till i get it right :D

as far as resistor values go, i'm currently buying 1/4w 1% metal film ones, they seem to have a much better tolerance that the 5% carbon film. am i right in thinking this? or does it not matter if it's 5% or 1% %

[jpfamps]

Metal film are lower noise than carbon film so I would use them.

1% tolerance is more satisfactory from an accuracy point of view, although I expect that you wouldn't be able to hear much of a difference using 5% parts.

[Keven]

i've found a nice replacement to the TL074 in the Burr Brown OPA4134. i've yet to try it out but i've seen the BB stuff in numerous pedal upgrades so i figure the quad package must be just as good.

http://focus.ti.com/lit/ds/sbos058/sbos058.pdf here's the datasheet.

just for clarification, +1 -1  is the same as +A -A right? 2=B 3=C 4=D ?

[jpfamps]

i've found a nice replacement to the TL074 in the Burr Brown OPA4134. i've yet to try it out but i've seen the BB stuff in numerous pedal upgrades so i figure the quad package must be just as good.

http://focus.ti.com/lit/ds/sbos058/sbos058.pdf here's the datasheet.

just for clarification, +1 -1  is the same as +A -A right? 2=B 3=C 4=D ?

+1,-1 is the same as +A,-A.

The OPA4134 looks a very good option (unless it's horrendously expensive). It's a FET input which much lower noise than the TL07* series and according to the data sheet doesn't experience latching issues when driven near its rails.

I've not found this op-amp available in a DIL package though (mind you I have only looked at Farnell and RS), only surface mount.

Can you rearrange your layout for dual op amps? This would give you more flexibility with trying different devices.

I will reiterate that I would use a DIL IC socket so that op-amps can be changed easily.

[Keven]

Well the layout i have here is a fixed PC board, but i was thinking of going with a perfboard or a turret board for my second one, so it'll make it easier to use dual opamps (well, 4 actually) and sockets and try things out for prototyping, and then getting a PCB professionnally made. I'll analyze the circuit a bit further and maybe go with dual opamps standard opamps as i don't really need 5 bands, so if i can cut it down to 3 bands and a gain control i think i could get away with two OPA2134 on sockets and a smaller enclosure.

or i can probably rearrange the whole thing to use 4 dual opamps, and then get different ones for better frequency response (the one with best high freq response for the high end, etc...) i understand how the circuit works (mostly) so changing connections won't be so hard. it's just a matter of mounting it to a board so it works and it's usuable live, I don't plan on selling this thing :D

I'm no electronics pro by any means, as you can tell by my questions, but this is a personal project that i'd like to see to the end, whatever time it may take!

[jpfamps]

All the op-amps discussed here have more than enough bandwidth for audio purposes, so the frequency response will not be an issue.

The main issues will probably be noise and headroom.

You can also try bread board as a way of prototyping.

Making your own PCB isn't that tricky once you have a design finalized. Maplins sell the kit to do this with, or if you don't mind going mail order Rapid Electronics are very good (and free postage over £35 too).

Getting away from quad op amps will give you much more options.

[Keven]

any stateside alternatives that you know of? i'm not on your side of the pond!

and yeah, after doing much research, it does seem like the main issue will be noise and headroom, though i plan to power this EQ with 18V supply if i can get my hands on one, or dual batteries at worst.

[jpfamps]

Didn't realize you were in the US!

Most of the US guys use Mouser or Digikey. Small Bear Electronics are also very good for pedal parts.

Op-amps generally have exceptional power supply rejection ratio (PSRR), so your power supply will have to be spectacularly poor to cause noise to be injected into the pedal.