Diode Polarity Protection

[Joe Hart]

I read the article on GEOFEX (I just wanted to get that out of the way!). I want to use the diode to ground thing and my understanding is that this is fine with a 9V battery, but if I connect the effect to AC the diode will fail and I could mess up my effect. But if I use an AC adapter, doesn't the circuit still only see about 9V DC? The article doesn't actually use the word "adapter" and that's the reason for my question. So can I use the diode to ground polarity protection trick with an effect that has a battery snap and a 9V adapter jack (like most effects have) or no? Thank you!
-Joe Hart

[theundeadelvis]

Could it have said a 9 volt AC power supply instead of a DC supply?

*Edit* You did say AC. Sorry!  

[R.G.]

But if I use an AC adapter, doesn't the circuit still only see about 9V DC? The article doesn't actually use the word "adapter" and that's the reason for my question.

OK. I meant to say "9Vac adapter".

If you connect an adapter with a 9Vac output to a typical effect with a 9Vdc input, even with the diode from ground (anode) to +9V (cathode), then the AC adapter will kill first the diode, then the rest of the circuits typically. What it does is put a large amount of current through the shunt diode and burn out the shunt diode, then it kills all the rest.

I've seen diodes in that situation that burned a black spot on the board before giving up.

So can I use the diode to ground polarity protection trick with an effect that has a battery snap and a 9V adapter jack (like most effects have) or no?

Yes, you can (and may  ) use the reverse biased diode between ground and +9V for an effect that has a battery snap and 9V adapter jack.

But it will not protect your pedal from a 9Vac adapter plugged into the jack. It protects the pedal from reversed DC adapters and batteries, not AC power sources.

If your question is can you use it for protecting from AC adapters, the answer is no, it offers little or no protection from that.

[Ice-9]

I may be wrong but I also suspect that the diode shunt to ground protection offers very little protection from 9v DC adapters also.

With a 9v battery reverse connected the diode will conduct the current through that until the battery give up its power which will also get very hot, but the battery will eventually fail and the diode has done its job.
With a 9v DC adapter, the current through the diode will continue until either the adapter burns out, or the diode dies. If the diode dies first which is most likely then your effect dies.

As mentioned in posts above the shunt to ground protection diode is not suitable for AC adapters at all.

[Gurner]

I may be wrong but I also suspect that the diode shunt to ground protection offers very little protection from 9v DC adapters also.

With a 9v battery reverse connected the diode will conduct the current through that until the battery give up its power which will also get very hot, but the battery will eventually fail and the diode has done its job.
With a 9v DC adapter, the current through the diode will continue until either the adapter burns out, or the diode dies. If the diode dies first which is most likely then your effect dies.

As mentioned in posts above the shunt to ground protection diode is not suitable for AC adapters at all.

I'd imagine 99% of the time is what you're trying to protect against with polarity protection  is someone connecting a battery the wrong way, then realising his mistake ("huh, it doesn't work....oh damn, silly me") before putting it right.

I just wish the electronics world had adopted a better way of 'keying' 9V batteries & they're associated connectors becuase as it stands at the moment, it's easy to connect a 9V battery in reverse no matter how transient this reversal may be....there's no physical barrier to stop this (though I'd imagine diode manafacturers might be pleased with the present arrangement!)

[Joe Hart]

Thanks everyone. I don't think I would ever connect an AC source to it, so it was just one of those "accidentally touched the battery to the wrong terminals" -- although in reality, it's more a "totally just in case" thing because I don't think I've ever put a battery in the wrong way or plugged in the incorrect adapter. But the pedal I'm building uses a part that I doubt I could replace (mojo reasons, but I also haven't found a part that sounds like it, either, so...). Thanks, again!
-Joe Hart

[R.G.]

I may be wrong but I also suspect that the diode shunt to ground protection offers very little protection from 9v DC adapters also.
...
With a 9v DC adapter, the current through the diode will continue until either the adapter burns out, or the diode dies. If the diode dies first which is most likely then your effect dies.

A lot depends on the adapter. Most regulated adapters will have a current limit which keeps them from supplying damaging currents. For the simple three-terminal regulators, this limit is a bit over 100ma, 500ma, or 1A depending on the regulator series. The common types of protection diodes are equivalent to the 1N4000 series and can carry 1A indefinitely without damage. Some more sophisticated adapters not only current limit, but do what's called "foldback" limiting, where the current actually reduces once it crosses the limit threshold. These are even easier on the diode.

The adapter I'm most familiar with because I do engineering support for it will sense an overcurrent, shut down entirely, wait for a second or two, then try to bring things back up. The long term fault/short circuit current for it is close to zero.

What makes AC adapters especially dangerous is that they are usually not current limited in any way.  They're just the bare secondary of a transformer.

[Ice-9]

I may be wrong but I also suspect that the diode shunt to ground protection offers very little protection from 9v DC adapters also.
...
With a 9v DC adapter, the current through the diode will continue until either the adapter burns out, or the diode dies. If the diode dies first which is most likely then your effect dies.

A lot depends on the adapter. Most regulated adapters will have a current limit which keeps them from supplying damaging currents. For the simple three-terminal regulators, this limit is a bit over 100ma, 500ma, or 1A depending on the regulator series. The common types of protection diodes are equivalent to the 1N4000 series and can carry 1A indefinitely without damage. Some more sophisticated adapters not only current limit, but do what's called "foldback" limiting, where the current actually reduces once it crosses the limit threshold. These are even easier on the diode.


What makes AC adapters especially dangerous is that they are usually not current limited in any way.  They're just the bare secondary of a transformer.

There are also quite a lot of unregulated DC adapters about that also just have a couple of rectifier diodes and a cap after the secondary winding., I have come across lots of these in the UK, So it would be prudent to buy a decent DC adapter like the ones  R.G. mentions.
Side note OT to R.G. , did you get to the bottom of the problem of 240v AC being on the pins after unplugging the adapter and giving a Little shock if you happen to touch the pins as its unplugged on the 1spot.

[R.G.]

I think so.

I believe that it's a "feature" of the input noise prevention filter. All switching power supplies generate switching noise, which radiates as radio interference, using the AC power line as one of its possible antennae. It's standard practice to meet radiated EMC regulations by using a common mode choke and high-voltage rated capacitors across the incoming AC power line.

I believe that by pulling adapter out of the wall plug, it interrupts the AC power going in at random parts of the AC power line voltage. If you catch it high, the input filter caps are held at that high voltage until it dissipates down. It is a small charge, but can be noticeable. The stored energy is small, but medium voltage, and it is enough to make you feel it if you catch it just right - or just wrong, depending on how you feel about electroshock.  Kind of like static sparks in dry weather.

That's the best theory I can come up with so far.

[Ice-9]

I think so.

I believe that it's a "feature" of the input noise prevention filter. All switching power supplies generate switching noise, which radiates as radio interference, using the AC power line as one of its possible antennae. It's standard practice to meet radiated EMC regulations by using a common mode choke and high-voltage rated capacitors across the incoming AC power line.

I believe that by pulling adapter out of the wall plug, it interrupts the AC power going in at random parts of the AC power line voltage. If you catch it high, the input filter caps are held at that high voltage until it dissipates down. It is a small charge, but can be noticeable. The stored energy is small, but medium voltage, and it is enough to make you feel it if you catch it just right - or just wrong, depending on how you feel about electroshock.  Kind of like static sparks in dry weather.

That's the best theory I can come up with so far.

Thanks for the detailed reply, Now that the adapter is permanently plugged into a distro board unplugging is no longer a problem but it is nice to know what the cause was.

[amptramp]

I have some 1N56 high-conductance germanium diodes that can be used in series with the effect supply for effects that take low current and are not that picky about voltage.  They have some reverse leakage, but probably not enough to do damage and you could always use a diode across the power to ground after the series diode if leakage current is a problem.

[Hides-His-Eyes]

1N5817 in series. Much more sensible.

[caress]

1N5817 in series. Much more sensible.

i use that sometimes, though a 1N5818.

[boogietone]

1N5817 in series. Much more sensible.

i use that sometimes, though a 1N5818.

Most pedal circuits specify a 1N914/4148 diode in series with the 9Vdc supply. Are the above "better" and why?

[amptramp]

1N5817 in series. Much more sensible.

i use that sometimes, though a 1N5818.

Most pedal circuits specify a 1N914/4148 diode in series with the 9Vdc supply. Are the above "better" and why?

The 1N5817 family are Schottky diodes with a forward drop of 0.33 volts.  However, the recerse current leakage may be a cause for concern with some circuits:

http://www.datasheetcatalog.org/datasheet2/a/0swl0u9yxwszejqog8artdo5rk7y.pdf

If forward voltge is not an issue, the 1N4148 offers better specs for reverse voltage and leakage current:

http://www.nxp.com/documents/data_sheet/1N4148_1N4448.pdf

The 1N56 is listed here among other early germanium diodes, but you should note that the number is so old that it has been reassigned to a silicon zener spike suppressor diode:

http://www.datasheetarchive.com/1N56-datasheet.html