stupid Zener question

[thermionix]

A couple years ago I scored a box of random components on Ebay for cheap.  Older stuff.  There was a package of Archer (Radio Shack) diodes, said 1N914 of the package.  I never noticed any numbers on the diodes and never questioned the type.  The package had been opened, but there were about 50 in there and they all looked the same, so I've been using them as 1N914s in various builds.

Well today I noticed numbers.  1770.  Huh?  Didn't sound familiar, so I googled 1N1770, and found out that is a 9.1V zener.  Oh no!  I tried to remember where I've used these diodes in builds.  I think only in the reverse-power protection spot, and definitely only in 9v powered circuits.

As an example, D1 in this GGG Phase 90 schematic:



So what have I done?  Did I unintentionally regulate my supply voltage?  Is there a downside to this, or should I just leave them in there?  Sorry I know this is basic stuff, I think it's okay, but I don't have much experience with or knowledge of zeners.

[R.G.]

So what have I done?  Did I unintentionally regulate my supply voltage?  Is there a downside to this, or should I just leave them in there?  Sorry I know this is basic stuff, I think it's okay, but I don't have much experience with or knowledge of zeners.

Sorry - you have to listen to some information on zeners first before this makes sense.

A zener is any diode that is operated beyond its voltage breakdown point. The real, no-fooling 1N914 diode specifies that it will withstand at least 100V of reverse voltage without breaking down. It may break down at 125, 150, or 500V, but the makers guarantee that it will withstand at least 100V.

A diode sold as a zener and not as a diode will have a guarantee that it >>WILL<< break over at a specified voltage. That is, for a 12V zener, they have made up lots of diodes that they think will break over at 12V +/- 5%, then they test them and find the ones that actually do break over at the specific voltage. So the difference between ordinary diode and zener is that you know where the zener breaks over, but only that the ordinary diode is "bigger than X".

What happens at breakover? The reason that diodes even exist is that they do not conduct current well at all when their voltage is reversed. They are guaranteed to withstand a certain amount. However, if you continue increasing the reverse voltage, they suddenly start letting current through at some voltage. When that happens, they still hold the same voltage, but let more current through. In theory, if you keep trying to increase the voltage across a broken-over diode, the current increases without limit and the voltage stays the same. In reality, the heat generated burns out the diode.

So in your case, you're using a 9.1V rated zener instead of a 100+ volt rated 1N914. What that means is that when the DC voltage coming into your DC jack exceeds 9.1V, it starts conducting current like crazy, trying to eat enough current to keep the voltage to only 9.1V.

What happens then depends on the zener's power rating and how much power the DC supply shoves into it. A zener that small can only get rid of maybe 1/4W to 1/2W of power without burning out. So if it's, say, 500mW it can get rid of safely, then the current it can stand is 0.5W/9.1V = 54.9ma. If the DC power supply tries to shove in more than 9.1V and the zener conducts, and the supply tries to make the current more than 55ma, the zener will rapidly overheat and die. If the external DC supply can't supply that much current, then the zener is "regulating" the DC that gets into your pedal. Glass body diodes are not all that good as "regulators" because they can't get rid of heat very well.

So - what to do? You're getting some protection, and in fact a little more protection than you'd get with a 1N914, but it's not much protection.  I personally would swap them out for 1N4000 series diodes. A 1N914 is a decent signal diode, but it can't conduct much current either, and my personal preference is to use a 1A or greater diode for reverse protection. But if they're working for you, you could just leave them until they die. If ever.

[thermionix]

Thanks RG!  I'll have to read through that a few times, but it's mostly what I thought was the case.  FWIW google tells me the 1N1770 is rated for 1W.  I always keep a supply of 1N4007s in stock for tube amps, so I can use those.

One pedal I used this diode in is a Phase 90 I built for a paying customer.  He hasn't picked it up yet, so I guess I should correct it before he does.  I don't want him to fry a wall wart if it puts out 9.6VDC and can't handle the extra current, but it's not clear to me if that's actually a realistic threat.

You're getting some protection, and in fact a little more protection than you'd get with a 1N914

This I'm not clear on either.  All things considered, would you say the 9.1v zener is a better choice than a 1N914, but not as good as 1N400x?  I'm also thinking about batteries, which can be as high as 10VDC when new.  Does the zener shorten battery life?  Or only until it's run down to ~9.1v?

[Rob Strand]

This I'm not clear on either.  All things considered, would you say the 9.1v zener is a better choice than a 1N914, but not as good as 1N400x?  I'm also thinking about batteries, which can be as high as 10VDC when new.  Does the zener shorten battery life?

Using the 9.1V zener "as is" on a nominal 9V supply (which could be as high as 10.6V with a battery) is a terrible solution.  If will not only will shorten the life of the battery it will probably blow up the zener in normal use. (RG already mentioned this.)

9.1V is a nominal voltage 

From this data sheet:
http://datasheet.datasheetarchive.com/originals/scans/Scans-004/Scans-0097628.pdf

The zener current will be 50mA with voltages as low as 8.2V and upto 10.0V.
The zener current will be at most 0.52mA at 7.3V.
So even at 7.3V (80% of the nominal voltage rating) the current still hasn't reached an insignificant level.

You just can't use it on a 9V rail.

You would need two in series plus some sort of series resistance to stop them heating up with the maximum expected voltage (say 24VDC?).

[thermionix]

Actually, I think this thread is essentially moot.  I don't think these are zeners I have, I don't know why they have 1770 printed on them (actually 17 over 70).  I tried a few with a brand new battery (9.96VDC), both directly and hooked to an SHO that's on my breadboard.  I'm not seeing any regulation down.  I even found one that had the band on the wrong end!  I guess I should try with a higher voltage supply to know for sure, but at this point I'm thinking they're just regular signal diodes.

[thermionix]

Okay Rob, thanks for the clarification!  I was typing while you posted.  I'm now going to see what one of these does to 18V just to be sure.

[Rob Strand]

Okay Rob, thanks for the clarification!  I was typing while you posted.  I'm now going to see what one of these does to 18V just to be sure.

A 30V supply with a 10k resistor can narrow things down a bit.

They may well be 1N914's.    It could be a date code but have to admit it's not common on diodes.   The other option is it's some kind of OEM marking like Peavey, Fluke, Tektronix and many other companies used back in the day.

[thermionix]

Yep, not 1N1770 zeners.  No regulation of 18V.  Some kind of date or manufacturing codes.  Damn Radio Shack!  LOL.  Wasted an evening, but at least I don't have to change parts in pedals.

[Rob Strand]

Some kind of date or manufacturing codes.  Damn Radio Shack!  LOL.  Wasted an evening, but at least I don't have to change parts in pedals

Don't worry, I've spent many hours chasing down part markings.

[thermionix]

I'll blame the jeweler's loupe.  I would have never seen those tiny markings without it!  The 1N914s would have always been 1N914s in my mind, because that's what it says on the package.

[rankot]

I usually put 11V/3W zener diode in series with appropriate PTC as a overvoltage and reverse voltage protection. It can withstand 270mA, which is more than enough for most of my pedals, and I choose PTC which is rated according to expected pedal current consumption. Like this:

[Rob Strand]

The 1N914s would have always been 1N914s in my mind, because that's what it says on the package.

There's a good chance they are, or at least something similar.

1n914's do vary from unit to unit.  IIRC you could see anything from 0.6 to 0.65V at 1mA.

[PRR]

> a package of Archer (Radio Shack) diodes, said 1N914... about 50 in there

I had that pack. RS bought anything similar to '914, often private-spec and private number. (Say Knight needed diodes that did not have to meet all '914 specs, they would offer a lower price for low breakdown voltage or high leakage, and the factory would cull their rejects and put Knight's part number on.)

The pack I had was notable in that old '914 was specced for 60V, most of mine broke-down at 95V to 125V. Which was perfect: I had a need for a 300V Zener and three selected junk-pack diodes got me 303V. They worked for years.

Unless you need to know, I do NOT advise this. Get a 300V supply and a 470K resistor. (Watch your fingers!) Feed the diode. Measure diode voltage. If it is 50V-150V, it is no-way a 9V Zener, and if in a small glass pack, is probably something from the '914/'4148 barrels. They churned them out like cookies then sorted for chewy/crunchy or voltage or leakage, to sell under many registered part numbers and many other "we don't need that good" private orders.

[thermionix]

No dates that I recall, but the package (gone) looked "1980s" to me.  Many of these diodes appear deformed, I've only used the better-looking ones, and generally checked Vf before use.  Now I know I need to double-check polarity too, after finding one mismarked.

To my satisfaction at least, I've already ruled out 9.1v zeners, as 18V was unaffected.  My only source for ~300V is my amp, and I'm not tempted to induce a partial short of any kind in there.

[EBK]

What you really need is an arrangement that allows you to place the diode-under-test across the top of a low-voltage radial electrolytic cap such that if the diode drops enough voltage, the cap doesn't explode.   

Ok, really I just enjoy watching stuff like this:

[PRR]

> My only source for ~300V is my amp

Won't hurt the amp (if done right).

With 470K in series, the only thing you could kill is a human.

And even that is doubtful (unless you puncture your skin to reach the juicy parts); but I agree with not doing it without a good need-to-know.

[duck_arse]

What you really need is an arrangement that allows you to place the diode-under-test across the top of a low-voltage radial electrolytic cap such that if the diode drops enough voltage, the cap doesn't explode.   

Ok, really I just enjoy watching stuff like this:

I was happy to see that motherboard blow. the led was a disappointment, tho. I like when they crackkk!

[amptramp]

If you do burn a LED, there may be some arsenic to avoid.  Gallium arsenide is the typical red LED but arsenic is in a lot of formulations.