Protection from incorrect power supply?

[raulduke]

There's some current leakage with Schottky diodes. In some cases, the amount of current that gets through won't be an issue. In others, it could be. One thing I've done is to use a 1N4001 in a crowbar after the 1N5817. I still don't think that this is a perfect solution, but it could help in some situations and it's pretty cheap in cost and complexity.

What about PTC's for high current protection?

You see these a lot now in other equipment, but not often in guitar stuff (including amps).

They are relatively cheap, and can withstand fairly high maximum currents before breakdown.

You can get 50ma/100ma variants that could be in the right ballpark to trip before you damage any of the expensive stuff in a circuit.

[RobA]

I think PTC would work fine as long as you left enough space for the heat dissipation and people knew to wait for it to come back online after it blew. I haven't tried it in the case of an AC input, so might the repeated trip cycles for the thing blow it if someone left the AC plugged in too long?

How about a bridge rectifier made from 1N5817's? If you did it right and had enough capacitance after, you could plug anything into it of the right voltage level and it should work. Hmm, I think I'll give this a try.

[RobA]

I tried the PTC in series preceding the 1N5817 using only the Schottky for reverse polarity protection. If the circuit tries to draw any current, the voltage at the output of the diode drops to essentially zero and the current across the PTC is unmeasurably low. I don't think the PTC is really going to get you much except in the case of a failure of the Schottky. If you wanted to get extra safe, you could go PTC into 1N5817 with a 1N4001 crowbar following. This will give decent protection against AC as well, since it is really a half-wave rectifier. But, at the cost of just 2 more diodes...

I built a bridge rectifier using 1N5817's. For either polarity of 9V DC input, you get the correct polarity out with a minimal voltage drop. The voltage drop across the rectifier is from 9.42V to 9.19V. Put an AC voltage on it and you still get (a noisy) correct polarity voltage out. If you pop on about 470µF of capacitance (~1200µF is actually better) you get a pretty clean output voltage. The warning here is that an input of 9VAC is going to end up with about 12VDC. So, you would need to know that the following circuit could take that voltage. But, this would give perfect polarity protection (noting that you would still need isolated supplies to mix polarities in different pedals) and very good AC protection (as long as the circuit can take the output DC).

If you can live with a bit more of a voltage drop, using a W06G or DB102 bridge rectifier would simplify the circuit a bit more.

[RobA]

Well, of course there has to be another downside. Thinking about this a bit more, I think you would have to use the bridge rectifier idea with every pedal or you would end up with different potentials tied to ground. Is there a way around that issue?

[raulduke]

Yeah whenever I have seen PTC's used they are 'first in line'. Like a fuse; you want it to be the thing to pop first!

Could a combination of a schottky and a zener be a good combo. You could then get over-voltage protection from any high dc voltages, or those caused by the rectification of AC?

I guess the simple solution is; plug the right bloody PSU in!

It's amazing how guitarist think they can plug any old PSU in and have a pedal work!

[GrindCustoms]

-9VDC at input, output -0.92VDC

Are you measure the open circuit voltage? This seems to me to be the leakage current times your mulimeter impedance (usually 10M) implying 92 microamps of leakage current.

The data sheet  for the Fairchild version shows a maximum of 500 microamps (0.5 mA) at 25 degrees C.

The main question is - will this small amount of current damage anything? I think not, but don't have data to back it up.

No matter how much AC i send through the diodes, tried up to 24VAC, it will pass through and you also get about half DC voltage of the AC supply, wich makes sense since AC is the total of a negative and positive wave and DC being just one side of the 0V.

Both type can flow through... so i'd say that it does'nt protects against AC.

Not sure what you mean by this - can you explain your test setup - I will spend time at the bench to see what I find.

Andrew.

Edit: Just went to the bench found a 12VAC wall wart. I connected it to a 1N5818 diode then to a typical RC power line filter you see in pedals. (a 47 ohm series resistor and a 100uF cap to ground).

At the filter cap I saw the DC rectified voltage (about 20 volts as expected  12VAC (rms) * 1.41.

The over voltage may cause a problem to a circuit which was designed for 9VDC, but I see no issue with reverse voltage in this case.

I simply put the diode on my Test rig that has a variable -24VDC/24VDC and variable 0-24VAC and bringing the different source of power to the diode and taking the measure at is other end.

Testing the diode in «stand alone».

[electrosonic]

From what I can tell I will stick with the series 1N5818 schottky diode.

Lastly, there is the p-channel mosfet method as shown on the geofex site.

http://www.geofex.com/article_folders/mosswitch/mosswitch.htm

It uses BS250P mosfets which are reasonably cheap on ebay (~$0.50 each or less). I just noticed that the Sunking 2 uses this method.

http://www.madbeanpedals.com/projects/Sunking2/Sunking2.pdf

Andrew.

[RobA]

Yeah whenever I have seen PTC's used they are 'first in line'. Like a fuse; you want it to be the thing to pop first!

Could a combination of a schottky and a zener be a good combo. You could then get over-voltage protection from any high dc voltages, or those caused by the rectification of AC?

I guess the simple solution is; plug the right bloody PSU in!

It's amazing how guitarist think they can plug any old PSU in and have a pedal work!

I think PTC into 1N5817 followed by a 10V (for a 9V pedal) zener would be pretty good. If the zener started to draw to much current from over voltage, the PTC would trip and that should solve the problems with the zener overheating. Could be a good solution.

[RobA]

[...]
I simply put the diode on my Test rig that has a variable -24VDC/24VDC and variable 0-24VAC and bringing the different source of power to the diode and taking the measure at is other end.

Testing the diode in «stand alone».

Can you try it with a bit of a load after the diode? I tried it with using an LED with a 4.7k resistor and this seemed to be enough drop the output current (and voltage) to an unmeasurable level. But, my meter is the greatest in the world.

[raulduke]

Yeah whenever I have seen PTC's used they are 'first in line'. Like a fuse; you want it to be the thing to pop first!

Could a combination of a schottky and a zener be a good combo. You could then get over-voltage protection from any high dc voltages, or those caused by the rectification of AC?

I guess the simple solution is; plug the right bloody PSU in!

It's amazing how guitarist think they can plug any old PSU in and have a pedal work!

I think PTC into 1N5817 followed by a 10V (for a 9V pedal) zener would be pretty good. If the zener started to draw to much current from over voltage, the PTC would trip and that should solve the problems with the zener overheating. Could be a good solution.

I agree, that could a nice simple protection scheme. I had a look and ptc's come too around 50p for the 100ma trip varieties

[GrindCustoms]

[...]
I simply put the diode on my Test rig that has a variable -24VDC/24VDC and variable 0-24VAC and bringing the different source of power to the diode and taking the measure at is other end.

Testing the diode in «stand alone».

Can you try it with a bit of a load after the diode? I tried it with using an LED with a 4.7k resistor and this seemed to be enough drop the output current (and voltage) to an unmeasurable level. But, my meter is the greatest in the world.

Sure can!

So you have the greatest meter in the world?

[RobA]

[...]
Sure can!

So you have the greatest meter in the world?

I don't know how come, but I do that way too often. I seem to have problems typing "n't".

[GrindCustoms]

So i added a load after the diode, 4K7 resistor and super bright 5mm blue LED.

I get the same voltage reading on my DMM when supplying negative voltage and AC.

And i've tried 5 different 1N5819, Fairchild, so no knockoff or whatever.

[RobA]

So i added a load after the diode, 4K7 resistor and super bright 5mm blue LED.

I get the same voltage reading on my DMM when supplying negative voltage and AC.

And i've tried 5 different 1N5819, Fairchild, so no knockoff or whatever.

Thanks for trying this. Seems I do need to get a better DMM. I should try this on an O-scope too.

So, in some cases at least, a crowbar 1N4001 post the 1N5817 would be needed to block the leakage current from the reverse voltage applied to the Schottky. If a zener were being used to block over-voltage, it might be enough to reduce the -V too.

[GrindCustoms]

So i added a load after the diode, 4K7 resistor and super bright 5mm blue LED.

I get the same voltage reading on my DMM when supplying negative voltage and AC.

And i've tried 5 different 1N5819, Fairchild, so no knockoff or whatever.

Thanks for trying this. Seems I do need to get a better DMM. I should try this on an O-scope too.

So, in some cases at least, a crowbar 1N4001 post the 1N5817 would be needed to block the leakage current from the reverse voltage applied to the Schottky. If a zener were being used to block over-voltage, it might be enough to reduce the -V too.

IIRC, Zener when supplied negative voltage if not re-oriented, do nothing.

Curious, about the voltage drop with a 1N4001 going to ground after a Schottky in serie... give me a sec will try it out..

Dinner hour for Science!