My GEOFX Spyder

[Sanguinicus]

Here's the schematic for my Spyder power supply. Due to the limitations of Altium, I couldn't put an 8-output transformer, so pretend there's 8 supply sections off the one primary 

The AC inlet, fuse and switch will all be in one assembly. Discreet bridge rectifier package instead of diodes. DC sockets will be the standard 2.1mm Boss style ones. Building it on strip board.

And I'm using the 8-output 240V (Aussie) transformer from Weber. The LED turns on when a 2.1mm plug is inserted (I hope). I want to be able to insert DC leads as needed. I think it's an interesting feature (read: eye-candy).

Biggest problem at this point is finding a suitable enclosure that only needs to be as big a necessary. Hammond has suitable box, the 1590DE. Unfortunately, the transformer isn't all that slim (would be awesome if they made a toroidal version).

Basically what I'm after is any glaring errors and feedback, especially with the LED section. It looks fine to me though. Also, will a 1A fuse be enough (slow blow of course)?

[Sanguinicus]

Please Reply!

Anyway, I'm thinking about adding short circuit protection to each output. Any idea how I could do this? Is it simple?

[milosch]

I don't see a schematic.

[R.G.]

Anyway, I'm thinking about adding short circuit protection to each output. Any idea how I could do this? Is it simple?

Almost all three terminal regulators have an internal current limit.

It is more difficult to add a current limit per output. The simplest way is to use a low current three terminal regulator. The LM3187L is suitable, as is the 7809 type regulator, both with current limits about 100mA.

[Skruffyhound]

Hey, I'm going to bump this for you because I'd like to know if your LED idea is viable. Somehow I imagine that there should be some simple FET/Transistor switching before the EE's in the crowd would be happy, but I'm not feeling clever enough this week to hazard a guess and I can't find the page on Geo that I'm thinking of.
R.G. didn't comment on the LED, does that mean it's good to go?
Good luck

[milosch]

I found a surplus transformer locally that would probably work if you can find one.  I used it for my 15V supply for my MXR Flanger clone.  It's an FP20-300 by Triad Magnetics and can be had from Digikey: http://parts.digikey.com/1/parts/998009-transf-10vac-6a-flat-pack-pwr-fp20-300.html

[Thomeeque]

Hey, I'm going to bump this for you because I'd like to know if your LED idea is viable. Somehow I imagine that there should be some simple FET/Transistor switching before the EE's in the crowd would be happy, but I'm not feeling clever enough this week to hazard a guess and I can't find the page on Geo that I'm thinking of.
R.G. didn't comment on the LED, does that mean it's good to go?
Good luck

LED idea is viable. Bit useless, but if you find it funny, why not.. Maybe consider bigger value of R1, 15mA will shine like crazy.. T.

[Skruffyhound]

Thanks  Tomas. Yeah, I'm not sure if it's exactly in this application that I might use it but there is something there. A possible application might be if one had rechargeable 9v cells in ones pedals and wanted to see that there was a connection from the external charger. It's not so much the circuit as the idea that just shorting the LED out with switching jack is fine.

[Sanguinicus]

I thought about switching the led with transistors, but shorting at the jack is an easy way to do it. It is silly sure, but it's a bit of eye candy  . Too late for the regs, already got 7809s rated at 1 amp due in part to me not actually knowing how much current effects actually draw. The short circuit protection sub system is more to appease my curiosity. I'd like to have an led light up when the circuit is shorted.

With the thermal shutdown of the regs, won't the effect fail before the reg shutdown? I mean, it will take a while for the reg to shutdown won't it?

[R.G.]

I thought about switching the led with transistors, but shorting at the jack is an easy way to do it. It is silly sure, but it's a bit of eye candy  . Too late for the regs, already got 7809s rated at 1 amp due in part to me not actually knowing how much current effects actually draw.

Each 7809 cost you about $0.50 or less, correct?  Half a buck is a relatively cheap lesson.    78L09s are about $0.30, making it an $0.80 lesson. Balance that against the possibilty of a short.

The short circuit protection sub system is more to appease my curiosity.

... as long as your transformer secondary can supply the 1A without frying.

I'd like to have an led light up when the circuit is shorted.

Not too hard if you're designing from scratch, but pretty touchy to do when you're already using three terminal regulators. At least touchy to do well.

With the thermal shutdown of the regs, won't the effect fail before the reg shutdown?

Maybe, maybe not. The 7809, in common with most three terminal regulators has both thermal and current shutdown independent of one another. It will shut down at 1A (roughly) even if it's cold, and at 140C chip temperature even if it's only letting through 1ma. The effect has to be able to eat 1A to get the regulator to shut down at all. If the effect pulls 900ma in a fault, the regulator will happily pass that forever if the power transformer, rectifiers and caps can supply it.

I mean, it will take a while for the reg to shutdown won't it?

In thermal, yes. However, overcurrent shutdown is pretty fast, sub millisecond even for laggardly, slow regulators.

[calpolyengineer]

Also, will a 1A fuse be enough (slow blow of course)?

I'm surprised no one pointed this out, but you should not use a slow blow fuse. If that fuse needs to blow, you want it to blow fast. There are a limited number of uses for slow blow fuses, this not being one of them.

--Joe

[R.G.]

I'm surprised no one pointed this out, but you should not use a slow blow fuse. If that fuse needs to blow, you want it to blow fast. There are a limited number of uses for slow blow fuses, this not being one of them.

Actually, it is one of them.

One of those limited uses for slow blow fuses is on the primary of a power transformer. At power, the transformer core must be magnetized, and depending on the remanent magnetism in the core and the phase in the AC line at which it's connected, the power surge to magnetize the core, including any saturation from the phase and remanence can be large. In addition, at power on, the filter caps are at zero volts, so they instantaneously look like a short circuit on the secondaries.

The question is generally moot with small transformers which have relatively high winding resistance, but an AC power safety fuse is one of the places you want a slow blow in general.

Fuses in general, even "fast blow" types, are not in general fast enough to clear in time to protect semiconductors which are not otherwise protected from overcurrents. It's a truism of power amplifier design that you cannot protect a power transistor with a fuse; the transistor almost always gives its life to save the fuse.

I think I can dig up some pointers to fuse application guides if you'd like to read up on it, and I'd welcome any pointers to guides which suggest that one should use fast blow fuses in AC power primaries.

[Sanguinicus]

http://taweber.powweb.com/store/wpdlxfmr2sch.jpg

That's the transformer I'm using. 300mA secondaries. So 78L09s are a better choice. How much current do effects draw anyway? Is 100mA plenty? Just to confirm, the over current shutdown of the regulators will be fast enough to fail before the effect does?

And finally, how big should the primary fuse be?

[PRR]

> the over current shutdown of the regulators will be fast enough to fail before the effect does?

If the effect is pulling excess current, it has ALREADY "failed".

Fuses/cutoffs are to prevent ADDITIONAL failures.

My lamp (or amp) wall-cord is pinched and shorts-out. Lamp or amp is dead. However while the short continues, the wires inside the wall are getting hot, will set fire to the wall. The fuse (should) prevent a wall-fire. If it doesn't, then a prolonged heavy short will heat the overhead wire to the street red-hot, which at my house could fall and set fire to undergrowth then the woods.

Fuses and shutoffs control domino-effect cascade failures.

You would like for a single-effect short to -

* not burn-down the house
* not smoke the Spyder
* power continues to the other 7 effects

"Proving complete fail-safe" is very tedious and will usually result in a MUCH more complicated design. Paradoxically, the extra parts reduce reliability. We usually settle for "no shock or fire", and accept whatever other soft-trouble comes with the parts.

Speculation and consequences:

Run the TO220 regulators with little or no heatsinking. They will carry 1W all day but will shut-down soon at 3W. Say 5V drop (14V-9V) at 2.5W is 0.5A thermal limit.

The PT is eight windings rated 0.3A each (I will ignore the ninth winding; the question is already too complicated.). A single output overloaded with 0.49A (not quite thermally shut-down) exceeds the rating of its winding. However seven windings at say 0.1A and one winding at 0.49A is less total transformer heat than the rated eight at 0.3A. The transformer probably won't smoke for many years, and a pedal pulling 0.49A is probably "sick" and won't stay in your rig that long.

> will a 1A fuse be enough

The point of this fuse is to protect the transformer and wall-cord from shorts, smoke, fire.

In many countries, the line-cord is presumed "protected" by the circuit fuse or a fused plug. US lamp-cord is now bigger than it used to be, so that it will pretty-sure blow a 20A circuit before significant ignition. UK cords have fuses, or selected small-gauge conductors in heavy jackets which will burn-clear without starting a fire.

Transformers heat slow. There's big differences between large transformers and small ones. A big one can pull BIG current for many minutes, often a slow-blow fuse is best. A small one can't pull HUGE current but may smoke in less than a minute. A slow-blow will still protect a small transformer, low-current fuses don't have a large spread between "fast" and "slow" characteristics.

Eight 11V @ 0.3A = 26.4VA
One 9V @ 2A = 18VA
=================
Total 44VA
44VA/120V = 0.37A max
44VA/240V = 0.184A max

Without trial/error over all normal and abnormal operation, we'd normally pick a fuse twice the nominal draw. That's when the transformer is "right size". In this case the transformer was sold over-size so that most customers would never over-load it. I doubt there are many pedalboards sucking 30 Watts!

So I would go 0.5A-1A for 120V lands, 0.25A-0.5A for 240V lands. And personally I would take either fast or slow, whichever was in stock. If both are handy, I'd try fast-blow but carry a sloblo in case turn-on surge turned out to be a problem (worked yesterday and blown today).

[Sanguinicus]

That reply was very informative. I've made the decision (based on RG's recommendation) to switch to 78L09 regulators (100mA shutdown). Is it possible to tell me what a typical current draw for effects is? What are the most current-hungry effects?

[PRR]

> Is it possible to tell me what a typical current draw for effects is?

How long is a "typical" rope? I got a 2-foot rope, I have a 250 foot rope. Saying a typical rope is 50 feet could leave me 200 feet long or 48 feet short.

Many battery pedals draw less than 1mA. I have read of pedals drawing 900mA.

I'd think any pedal that "can" be run on battery usefully will be under 100mA, but there's exceptions to any rule.

Put the voltmeter on the output while you plug in a pedal. If it drops more than say 1%-2%, suspect that pedal is sucking more than 100mA-140mA.

[Sanguinicus]

I suppose you're right. I'll see how I go with 78L09s and swap them out if the need arises.

This forum is bloody great.

[R.G.]

Is it possible to tell me what a typical current draw for effects is? What are the most current-hungry effects?

As Paul notes - no.

However, there is at least on on-line compilation of measured currents for a lot of effects. I can't remember the link now, but perhaps someone else can.

In (very) general terms, the most common effects (distortions and overdrives) use something under 50ma, with about 5-20ma of this being used by the LED indicator. The really power hungry ones are the digital delays, or digital DSP pedals. A simple delay, perhaps PT2399 based, will pull 100-200ma. The commercial digital delays and DSP based effects are the real gluttons. These can eat up to 1000ma each, and are often supplied with their own power adapter, often 9V **AC**.

If you ever get a 9Vac adapter, be very, very careful what you plug it into. An adapter providing AC will kill the polarity diode in the typical effect, then kill the rest of the effect. I know of only very few exceptions to this.

[Processaurus]

However, there is at least on on-line compilation of measured currents for a lot of effects. I can't remember the link now, but perhaps someone else can.

Was this it?

http://www.diago.co.uk/powerstation_compatibility.html

[R.G.]

Was this it?

I don't know if that's "it" but it's certainly one. It's been a long time since I looked. Seems like there were more filled in at the one I saw. But this illustrates the idea very well.