My GEOFX Spyder

[Sanguinicus]

So the 100mA regulators might not be enough for some pedals. Should I go higher? I plan on building a pt2399 based delay soon.

[G. Hoffman]

If you need more current, you can run two of them in parallel to get double the available current.  If you need 18V, you can run two of them in series.



Gabriel

[PRR]

> what a typical current draw for effects is?

No.

Running some math on the "Powerstation" data says the _average_ is 150mA. However many of the high-current (to 2000mA!) pedals use (as R.G. says) _AC_ supplies, and some odd voltages (to 40V!), so will need their dedicated warts. These 23 pig-pedals using 1000-2000mA _AC_ skew the average, but would NOT be suckled from your 9V Spyder.

Casting-out the sort-end where a lot of pedals are showing odd supplies, the _average_ is 50mA, 116 models at 10mA or less, 400 pedals (about 90%) under 140mA, and only one at 350mA.

Obviously this _averaged_ data does not reflect "typical reality" (whatever that is). Popularity ("typicalness") is not factored: Some of those pedals are found on every pedalboard, others are unknowns. 62% have no mA number.

[PRR]

> run two of them in parallel to get double the available current.

R.G. or others may have tried this and can refute me.

But direct parallel of two regulators is usually "bad".

It works (mostly) on batteries because they sag. If the batteries are very similar, the higher one sags until the lower one takes part of the load.

Regulators "don't" sag (much) until they shut-down. And they are not precision. If you have two "9V" regs but one is really 9.0V and the other is 8.8V (well within specs), the 9.0V one will try to carry the whole load, the 8.8V one will be confused and trying to make itself more-off. If the 9.0V reg would sag.... but 78L09 sag is less than 0.1V before it punks-out. When the 9.0V one does over-heat and shut-down, the 8.8V reg is free to carry the load alone. Of course it will shut down about as fast as the 9.0V one did. By that time the 9.0V reg may have cooled and come back.... at best you have a 9.0V/8.8V/9.0V/8.8V/9.0V/8.8V wobble, and probably periods where both regs are resting and output is zero. Worst-case, the shut-down reg with the other reg *forcing* its output high breaks-down and dies. (One of the few ways to kill these handy chips is to put voltage where no voltage should be.)

Series works only if the raw sources (transformer windings) are isolated. In this case, they are; also separate wall-warts may be seriesed. However out of context, somebody may try to series two outputs which are internally paralleled, which is a dead-short which may be bad.

[Sanguinicus]

All of this new info! Confuses me more and more. Anyway, 78L09s will be fine? Or should I aim slightly higher?

[slacker]

So the 100mA regulators might not be enough for some pedals. Should I go higher? I plan on building a pt2399 based delay soon.

I can't speak for all of them but simple PT2399 based pedals like the Tonepad Rebotes or my Echo Base don't draw anywhere near 100ma, the Echo Base is about 30ma max. That's using a "super bright" indicator LED which means you can run the LED at less than 1ma.

Fancier ones like the PT80 might draw more.

[askwho69]

since someone show some illustration- can i ask what would be the difference between AC and DC? if i will build a good power supply.. what should be the common AC or DC? What if i use AC to a pedal that is requiring a DC does it produce hum? damaging the pedal? or vice verse. . . 

[Sanguinicus]

You will kill the pedal. AC will slowly fizzle the input diode then the rest of the circuit.

[Sanguinicus]

Slacker, I would have thought the Echo Base was much fancier than the PT80, which is the one I wanted to make. Now I'm not sure which delay to build. I don't want something overly analog-voiced.

I think I might try to get LDO regulators rated a bit higher.

[G. Hoffman]

> run two of them in parallel to get double the available current.

R.G. or others may have tried this and can refute me.

But direct parallel of two regulators is usually "bad".

It works (mostly) on batteries because they sag. If the batteries are very similar, the higher one sags until the lower one takes part of the load.

Regulators "don't" sag (much) until they shut-down. And they are not precision. If you have two "9V" regs but one is really 9.0V and the other is 8.8V (well within specs), the 9.0V one will try to carry the whole load, the 8.8V one will be confused and trying to make itself more-off. If the 9.0V reg would sag.... but 78L09 sag is less than 0.1V before it punks-out. When the 9.0V one does over-heat and shut-down, the 8.8V reg is free to carry the load alone. Of course it will shut down about as fast as the 9.0V one did. By that time the 9.0V reg may have cooled and come back.... at best you have a 9.0V/8.8V/9.0V/8.8V/9.0V/8.8V wobble, and probably periods where both regs are resting and output is zero. Worst-case, the shut-down reg with the other reg *forcing* its output high breaks-down and dies. (One of the few ways to kill these handy chips is to put voltage where no voltage should be.) [/url]

Huh.  The Voodoo Labs one I've got (which uses 78L09's for most of the outputs) recommends it, if I remember correctly (though now I can't find it on their website, so maybe I'm just remembering it wrong - easily believed!).  Maybe there is more going on in theirs than I'm aware of.  Maybe a series resistor to give any mismatch somewhere to go?  Though then you get into a problem with sag.

Series works only if the raw sources (transformer windings) are isolated. In this case, they are; also separate wall-warts may be seriesed. However out of context, somebody may try to series two outputs which are internally paralleled, which is a dead-short which may be bad.

Indeed, though I was talking about this specific situation where they ARE isolated, and it would work fine.  But yes, don't try it if you don't know the outputs are isolated! 

All of this new info! Confuses me more and more. Anyway, 78L09s will be fine? Or should I aim slightly higher?

It probably makes at least some sense to look at how it is done in a comercially available unit.  The Voodoo Labs Pedal Power 2+ (which I've got) has four 9V 100mA outputs, which I believe are exactly the same as the circuit on the Spyder, and use 78L09's (I've looked inside mine, and they are absolutely TO-92 semiconductors, and since I don't see any zener diodes I don't think it is a discrete regulator circuit).  The other outputs are two 9V 250mA outputs using a circuit I haven't bothered to figure out (I believe the regulators are also 78L09's, but there is obviously more to the circuit than that - they are certainly TO-92's, but I haven't looked to see if they have a model number), and two 4-9V 100mA that are basically the same as the Spyder article's variable voltage version, though without the output impedance control (I believe - they may have just put a voltage divider on the output). 

So, at the very least, you can probably use 78L09's for MOST of your outputs.  If you have something DSP based (a big digital delay or similar), then you can build one or two outputs with a 7809 or whatever other design you want.  Then again, if you already have the 7809's, you are probably OK to just use them.  As PRR already pointed out, if a pedal is drawing enough to shut down the 78L09, it is probably already broken, and if it is damaged enough to cause a serious problem will shut down the 7809 too.  The only major value I see to using a 78L09 is saving space - which is a very real advantage, but you can decide if it matters to you.


Gabriel

[Sanguinicus]

Settled then. 5 LDO outputs. 3 high current outputs with 7809s. Cant find regs that are 9v and around 250mA.

[aquataur]

I just made one of those spyders. I took eight 2VA transformers, that were the smallest they had. Although certainly not the most elegant solution, this was the easiest. Look up my web-pages for some piccies: aquataur.at.tf.

I used a wooden case to avoid any discussion on electrical insulation, although in retrospect it would have been sane to avoid stray current induction in my single coils. It is also not the smallest case, but if I compare this to yours, R.G., it is small

It uses 78L09 chips that deliver 100mA. If you consider pedals like fuzz faces, diode clipper based overdrives, wah pedals and the like "typical" then a typical effect device consumes max 10-20mA, so each single spyder circuit has ample power overhead. Practically all battery driven stomp boxes use currents smaller than that, otherwise the battery life would be ridiculously low.

A full-blown digital effect in comparison easily chews up an ampere, a tube based effect may also typically ask for an external supply. Some of those non-battery operated effects do also ask for different or AC voltages, beware!

Having turn on the LED upon plugging, I don´t know how useful that is... I´d rather see a power good signal to know if the circuit is intact. But feel free to disagree...

have fun,

-helmut

[Sanguinicus]

I think an earthed metal box is a must. I'm having a hard time finding one that will fit the transformer while being as small as possible.

[G. Hoffman]

I think an earthed metal box is a must. I'm having a hard time finding one that will fit the transformer while being as small as possible.

DIY is seldom very useful for "as small as possible."  When you are making several thousand of something, you can take the time to prototype and get things as small as possible, but if you are only making one you are better off accepting something bigger.



Gabriel

[Sanguinicus]

Very true. Although I usually find myself cramming stuff in tiny spaces. I'm pretty good at making use with very little space. Besides Hammond, does anyone know of another box manufacturer?

[R.G.]

Be
very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very, very
careful cramming a lot of AC power stuff in small spaces. A tiny box is not worth dying from electrocution for.

[G. Hoffman]

If having a tiny box is really important to you, I would say you are probably better off buying a Pedal Power 2+ or something along those lines.  (I know there is at least one other commercially available transformer isolated power supply, I just don't remember what it is since I don't own one.)

Oh yeah, and DON'T FORGET THE FUSE!!!!


Gabriel

[Sanguinicus]

It's gotta be DIY And the fuse, I got an IEC power entry unit with switch, fuse and inlet.

[Sanguinicus]

This sounds crazy, but will I need different patch leads? If all the shields of the leads are connected that voids the isolation of the supplies right?

[Sanguinicus]

Finally going to start building this. IK've got a few more questions though.

Firstly, because a transformer is an inductive load, should have have flyback diodes connected to it to protect the rest of the circuit in case of a voltage spite when it turns off?

What about protection diodes at the start of the circuit just in case something goes wrong during operation? How would I go about this?

Thanks!