Did Diodes Die?

[sundevil_1997]

So I stopped by radio shack, and bought 6 of the diodes that are supposed to drop the voltage by about .75 V each. I then soldered 4 of them inbetween the 12V connection and the fan, and I also inserted a switch and soldered 2 more of them serially. So here's the idea: At one switch setting, there would be 4 of these diodes between the 12 V and the fan, so that should drop it about to 9V...then at the other switch setting, it would add 2 more diodes, so it should drop it down to about 7.5 V.

So I hook it up, plug it into a spare PSU I have sitting around, and break out the multimeter. At the first switch setting I had....10 V. Hmmm...that seems a bit odd. With 4 diodes serially in the way, how did I get 10V..unless they're all worth .5 V each instead of .75? Ok, try next switch setting. 10V. Ok....I may not be a genius, but I KNOW it's going through 6 diodes now, and still only giving me 10V??

So here's what I think may have happened. I'm a soldering NEWBIE! I'm lucky the house didn't burn down. Several of these diodes I soldered end to end...and at one point while holding the iron onto the lead (waiting for it to heat up enough to melt the solder I was holding down next to it), I nearly burned myself by the heat coming from the lead on the OPPOSITE side of the diode. In other words, a whole lotta heat had traveled through the diode. I figure that can't be good.

Can the heat from a soldering iron render the diode useless? And if so, what is better soldering ettiquette? Also, what would you suggest as the best way to connect all those diodes? Here is a diagram of what I'm trying to do:

[lenny]

Can heat kill diodes? Yes.

How to avoid it? Solder as quickly as possible, use alligator clip between soldering point and diode as heatsink.

I'm not experienced with soldering either (I'm a software engineer ) so I'll leave it to someone else to give you tips on soldering.

You may want to experiment (with a new set of diodes) using a prototyping board (can't find a photo : here's a drawing) or alligator clips to see if it works as designed. I know voltage drop across a diode is not a fixed value, but I can't recall the specifics.

[sundevil_1997]

I'm not experienced with soldering either (I'm a software engineer ).

Hahah..I have the same affliction.

I did a google on "how to solder", and it came up with a reasonably informative site....which included the alligator clip idea.

I remember having a prototyping board (we called it a breadboard) way back in college. Not a bad idea to get another one, I guess...it's a heck of a lot nicer to discover something doesn't work BEFORE I turn on the 40000 C degree soldering iron.

[peteamer]

sundevil_1997, Lenny is 'spot on' with what he says.

May I add...

1: Before you begin to solder.. make sure your soldering iron is fully hot... (to lessen the 'contact time' ) (I prefer the tip of the tongue or nose method... but other people are probably a lot more intelligent and just wait a suitable (x + y - t + ambient temp. x 72.234secs) amount of time... )

2: Using your Mulitimeter/New Member... (very old joke.. ) measure between each diode and you should be able to see if one or two are not behaving like they should and go from there...

Hope this helps some ...

Pete

[lenny]

I remember having a prototyping board (we called it a breadboard) way back in college.

Oh yeah, breadboard! That's the word I was looking for! Been too long, getting too old Thanks for refreshing my memory

[sundevil_1997]

1: Before you begin to solder.. make sure your soldering iron is fully hot... (to lessen the 'contact time' )

Hmmm...well, I have no particular method to know. The iron contains a light to indicate it's on, but that's about it. Contact time might certainly be an issue...constantly I had to hold the iron onto the surface for around 20 or more seconds before the solder (right NEXT to it) would finally melt.

2: Using your Mulitimeter/New Member... measure between each diode and you should be able to see if one or two are not behaving like they should and go from there...

My Multimember is rather simple...it does voltages, continuity...maybe resistance, but that'd be it. I tried testing a few diodes even before I used them. I didn't expect to see a voltage drop when I test it for DC voltage, and I didn't. I can't remember if I tested them for continuity or not (I'd expect that to test positive in only one direction). So I'm not sure how I'd check to see if they work or not.

[sundevil_1997]

Oh yeah, breadboard! That's the word I was looking for! Been too long, getting too old Thanks for refreshing my memory

Heh..you know, of the two names given to it....breadboard, or prototyping board...ONE of them makes sense to me.

Well, all I can say is thank GOD for cheap electronics. Burning out under $2 in diodes is not bad. I think getting a breadboard, some new diodes and all that and testing it out first is a good plan.

Does anyone have a better suggestion than soldering diodes end to end? I was doing that and using heatshrink to cover them all...but would it be better if I put them all on an electronics board instead of just connected directly?

[peteamer]

sundevil, with your multimeter you should be able to 'see' a constant voltage drop between each diode.

To use your diagram as an example:...

1: Connect the multimeter black wire to the wire that goes straight from one molex to the other...

2: use the red end of your multimeter as a probe...
i.e. : on the 'other' wire at the 'PSU molex' end, you should read ~12V,
at the wire/connection between the first and second diodes (from the 'PSU molex) you should read ~11.3V,
at the wire/connection between the second and third diodes (from the PSU molex ) you should read 10.6V and so on...

If you don't see the same drop between each and every diode, one has gone 'short circuit' and is doing nothing to regulate/drop the voltage.

If your soldering iron is taking ~20secs to melt solder right next to the iron, I would wager than the iron is not getting hot enough before you attempt to use it or is of too lower wattage to heat the wire and solder sufficiently quickly. I have used low wattage irons in the past that can take nearly 5mins to warm up...


Hope this helps some... let us know how you get on and we can help more.


Pete

[BrianE]

Does anyone have a better suggestion than soldering diodes end to end? I was doing that and using heatshrink to cover them all...but would it be better if I put them all on an electronics board instead of just connected directly?

I've never used them, but you can also buy pre-drilled PCBs with copper strips running along the holes. Just figure out how to place the diodes serially and solder them in. Breadboards are more for temporary experiments and I wouldn't trust it to hold in the long run because they're just friction fit.

[cmcquistion]

Check out the link in my sig. There is an easier way to do that and it won't involve burning anything:)

[sundevil_1997]

To use your diagram as an example:...

1: Connect the multimeter black wire to the wire that goes straight from one molex to the other...

Ooooooooooooo....I get it. Stupid me. I thought you meant check the diode individually, not connected to any circuit at all. now it makes sense...just work my way from the 12V down to see which rogue diode isn't dropping the voltage like it should. gotcha.

If your soldering iron is taking ~20secs to melt solder right next to the iron, I would wager than the iron is not getting hot enough before you attempt to use it...

Well, from that google'd page, I've discovered a few things I'm doing wrong. For one thing, I haven't cleaned the iron tip in I don't know how long. Gotta find a sponge (and a soldering stand, for that matter). Also haven't been "tinning" the tip.

No idea the wattage. Next time I'll plug it in, then go watch a Family Guy...by then it should be hot enough.

The annoying thing was, like when I was soldering one of the diodes to a piece of wire...the insulation on the wire was melting! But the solder (which was much closer to the contact point) still wasn't. I don't think it's a problem of the type of solder, since it's what came with the iron. But still.

Thanks for the help, all of ya. I feel emboldened again.

[sundevil_1997]

Check out the link in my sig. There is an easier way to do that and it won't involve burning anything:)

Gosh darnit...I was just getting mentally prepared to put my life and skin into jeopardy again, and you go and give a perfectly simple alternative. Now, do I wimp out and go with the wonderfully versatile terminal block idea? Or do I get back up on that bucking soldering iron again to show it who's boss????

*Checks availablilty*

Cool, the Radio Shack nearby has terminal blocks in stock. Woohoo!

*Looks over his shoulder at the soldering iron, coiled up like a snake about to strike*

Some day, iron.....some day you and I will have it out once and for all!

[peteamer]

I feel emboldened again.

(I don't actually know what embothingy is... but it sounds like it's making you feel good... so it's cool... )

The solder you probably need is only ~1 maybe 2mm(ish ) thick, any thicker and it too absorbs heat faster than the iron can provide it...

However... cmcquistion's suggestion is the best yet ... I was impressed when I first saw it... and if I had a memory (am hoping to upgrade to goldfish memory soon...) would have linked you straight to it ... (bring on the WIKI...)

Good luck on either route... and let us know how you get on.


Pete

[sthayashi]

Cmcquisition's got a very good suggestion there.

As for soldering, you may just want to plain get a new tip. Tips wear out and it happens.

Second, though I don't know how many levels of bad this was, in the past, I've used wet paper towels in lieu of sponges (learned it from an EE professor from China. He was very talented given the level of ghettoness he used. For a while he stripped his wires using an old pair of office scissors).

Third, and again I'm not sure how many levels of bad this is, but usually I would melt solder onto the tip of the iron first (and AWAY from the component), and then touch the tip to the lead while continuing to feed the solder. Pull away when enough solder is on there. If you're not using wire at ALL, you may want some solder flux to apply to the leads.

Fourth, almost all multimeters can measure diode voltage drop. It's usually the same as the continuity test.

[sundevil_1997]

Cmcquisition's got a very good suggestion there.

As for soldering, you may just want to plain get a new tip. Tips wear out and it happens.

Heh...I figured it would have been clear from the posts that this tip has NOT been worn out. This is the 2nd time I've used this thing.

For a while he stripped his wires using an old pair of office scissors).

Heck, I use wire cutters right now for my strippers because my strippers SUCK.

Thanks for the reminder that I don't need to go out tonight specifically to buy a soldering sponge. It's easy to get tunnel-visioned when trying something new. ("But the site said use a SPONGE....")

Hmmm..well, I'll try the multimeter again. Only so many ways to hook it up.... I usually just listen for the beep when I do continuity, I haven't looked at the actual display.

Thanks!

[flyingsherpa]

... Also haven't been "tinning" the tip.

tinning the tip is crucial, i think that is your big mistake (tinning causes MUCH better/faster heat transfer). also, tinning tells you when the iron is hot enough. so while it is warming up, you try to tin the end... if it won't tin, then it is not warm enough. if it tins but it takes 4 seconds, it is still not ready. i know mine is ready when it tins almost instantly when i touch solder to it. keep practicing, it really does get easier the more you do it and it can really come in handy... i've even started soldering the ends of my bicycle cables to prevent them from unraveling.

but you should still use the gator clip heatsink too whenever soldering small, fragile parts like tiny diodes.

[sundevil_1997]

*GRUMBLE*

I swear, I'm getting tired of buying wrong parts.

So I go back to Radio shack...they're out of 1N4001 diodes because I bought them all a few days ago. Fortunately, they have a 25 diode pack for a couple bucks, and in there is an assortment of 1N400x type diodes. No problem, I figure.

So I get the terminal block...I hook up the multimeter..and I'm ready to go.

I wire the 12v lead to the first terminal block, and from there I hook a 1N4002 diode to the 2nd block. I break out the multimeter and read. With no diodes, I get 11.95. Close enough (there's no load on this PSU, and it's relatively old...just serves as a good test PSU). So I check the voltage on the 2nd term block, so going through 1 diode. 11.70. What? So, a drop of .25 volts? Hmmm. So, I wire up another of the same diode into the 3rd block. Check the voltage. 11.45 V.

Ok...so contrary to everything I've read from everyone, I somehow got the magic batches of 1N400x diodes that only drop .25 V each instead of .75 V?? I'm NOT going to wire up 20 diodes just to get 7V.

What in the world is going on here???

[gr8r-x]

Q and A time...

A. You can use a steel wool ball to clean the tip. This helps reduce the heat time between cleaning and can prolong the life of the tip.

A. Tinning the tip is VERY much a good idea. Also, the idea of putting solder on the tip before the wire can work, but the solder will burn quickly if left to long. Burnt solder == bad.


Q. Won't the drop across the diode be in proportion to the draw of the circuit as a whole? So wouldn't you have to measure the diode WITH a fan in circuit? Didn't sound like you were doing that... I've never used diodes to regulate voltage, so am unsure of the practical usage.

[sundevil_1997]

A. Tinning the tip is VERY much a good idea. Also, the idea of putting solder on the tip before the wire can work, but the solder will burn quickly if left to long. Burnt solder == bad.

This is something I've not fully understood. So, leaving the tip on solder too long causes it to burn, right? But isn't "tinning" just putting solder directly on the iron? Wouldn't THAT solder burn?

In my experience, when I let the solder get too close to the iron on the work, the solder seemed to almost flash to vapor. Seems like the iron almost gets too hot to put and keep solder on. But either way, I definitely do need to clean it...it's way dirty.

Also, aren't you supposed to put a coating of solder in the iron tip before unplugging and storing it? Or should you clean it before storing it?

Q. Won't the drop across the diode be in proportion to the draw of the circuit as a whole? So wouldn't you have to measure the diode WITH a fan in circuit? Didn't sound like you were doing that... I've never used diodes to regulate voltage, so am unsure of the practical usage.

Gah, it's all the little things they don't tell ya....

I HOPE what you say is true. Even at only $2.58, I'd still be miffed to have a pack of 20 diodes that would be useless to me. It kind of makes sense what you're saying...I was just measuring voltage of the bare wire ends. I'll plug a fan into the end tonight and then measure it.

It's funny...the whole goal here was to replace the 7V trick because I was concerned it would eventually kill my PSU. Well, after last night's frustration, I instead rewired for a 5v connection, and the fan works fine at that voltage. I still have to monitor temps to see if it works ok, but at least the crisis is averted for now. Heh.

[scruzbeachbum]

Testing diodes:
http://www.elexp.com/t_test.htm

[Aleksi]

one of the previous posters is on the right track.

I checked your schematic and was also wondering "Why the hell doesn't that work?". If there isn't any load =>GND in the circuit after the diodes (by this I mean a fan) the circuit is open and the diodes or far that matter any other component will not work.

Try this to get the basics sorted out:

Take two wires that give you +12V and GND. Solder a diode to the 12V wire and then solder this wire to a fan's voltage supply. Than connect the grounds. If this works OK (the fan should be fed ~11.3V ath this point) add another diode after the first one and measure again. Simple ASCII schematic

+12V----|>|----fan +

GND------------fan -

As you can see, if there is no fan connected, there is just a diode dangling from a 12V wire and the circuit isn't closed without connection to the GND.

Hope this helps.

[sundevil_1997]

I checked your schematic and was also wondering "Why the hell doesn't that work?". If there isn't any load =>GND in the circuit after the diodes (by this I mean a fan) the circuit is open and the diodes or far that matter any other component will not work.

Ya know, if I'd paid attention in my EE classes, I would have been smart enough to figure that out last night.... Those classes just convinced me that electricity worked by magic.

I'm DEFINITELY in the "knows enough to hurt himself" category of electronics and soldering right now.

I'll let ya know how the testing goes tonight once I throw a fan into the works.

Thanks everyone for the help.

By the way, since this is the fans and CONTROLS forum, maybe a sticky post on top about "Build your own" might be useful. I know I see posts constantly about diodes, rheostats, 7V trick, etc. Just an idea.

[peteamer]

Those classes just convinced me that electricity worked by magic.

You mean... It DOESN'T !!! ... ...

I'm DEFINITELY in the "knows enough to hurt himself" category of electronics and soldering right now.

Sorry Squire... but that's unlikely to change much... even with a 'lot of experience' we all do things that... remind us of basic (forgotten) principles... and common sense.


Good Luck none the less.

Good thought on the Sticky, any takers on the task?...

[sundevil_1997]

Good thought on the Sticky, any takers on the task?...

Well, I've learned a fair amount trying to do this, I'd imagine I could collate that info into a post. Then what? Don't we have to throw it at a mod to get it stickied?

[peteamer]

Provided it does what is needed and covers enough usefull stuff, accurately and well, with links if nec., SPCR Mods are more than good enough to 'pick it up' as they browse around.

[Mar.]

Hah... As an undergrad EE major I found this topic pretty entertaining.

We learned to solder in high school... I assumed everyone knew how?

[scruzbeachbum]

me, too.

[BrianE]

Heheh, I asked my father about this one, since he knows a lot more about this than I do. Like Aleksi, he says you can't measure the voltage drop with the circuit open like that because (assuming you're using a digital multimeter) the meter will draw extremely little current when it completes the circuit. Current is extremely low because of the meter's massive (mega-ohm) resistance built into it. He suggested hooking up your fan or a 1k ohm resistor to complete the circuit and allow you to measure it.

The voltage drop for silicon diodes like this isn't a perfect wall. Search on the internet and you can probably find a graph showing the relationship between current and the voltage drop of a diode. After hashing it out a bit, we figured that if you're drawing super tiny amounts of current you probably will not get the expected .75V drop per diode, but something significantly less, which could explain why you're getting crazily high readings.

Very basically, the circuit needs a few milliamps of current going through it in order to work the way you expect. (From discussion I believe that the voltage drop is not actually something purposely designed into the diode, but rather it is just a by-product of the fact that you are pumping voltage through a semi-conductor instead of a conductor. Silicon diodes drop ~0.7V, germanium diodes only drop ~0.2V.) He thinks this is a rather unusual way of regulating voltage, but thinks that it should work once actually hooked up to a fan/load.

I guess we'll find out......

[sundevil_1997]

He thinks this is a rather unusual way of regulating voltage, ....

Heheh...ya, I'm sure it is, to someone who actually understands this stuff. Perhaps "regulating" is too strong a word. Maybe more like dumbing it down...beating even.

But anyways, the rumors are true....hooked up a fan, then read the voltages at the various terminals, and sure enough, it was dropping like it should. After 2 diodes, I was already down to 10v. The 12 V source was around 11.8, though, so that explains how it drops so fast.

Thanks for reminding me of the most BASIC of electronic rules. I'm glad I got it working....it will certainly make that big pack of diodes useful now.

Oh...and my digital multimeter has a specific setting for measuring diodes. Didn't even see it before. Ah well, live and learn. I think I will dedicate myself to a sticky post so that others will be spared this painful learning.

[BrianE]

Cool, glad to know it really works! I like the cheapness and simplicity of this... maybe I should build one of these things too.