Diode Trouble

[Greek Acrobat]

Diodes aren't supposed to heat up when they're in use, are they? If not, then I've got some hassle and don't know what's up.

I've been trying to hook up this circuit:



but using two op amps before the diodes. Two of the diodes are heating up a hell of a lot. I think it's all wired up ok, any ideas what's happening/not happening? The diodes are all 1N914s. Am I just overdoing things?

I'm also getting a very mushy, staticky high end. Has this got to do with the diode problem? Or is there a cap I can stick somewhere to roll off the high end? Or would you expect that from this layout anyway? Or am I still overdoing it?

Cheers.

[Zero the hero]

I can't see the picture...

[Greek Acrobat]

Hmm, I can.
The URL is http://www.geocities.com/austenfantanio/simplehate.gif

Don't know if that'll help though.

[smoguzbenjamin]

I guess you could test if it's the diodes creating the high-end mush but simply taking em out. It it's still bad... it's not the diodes...  

BTW aren't you supposed to place caps on the output/input anyway to block stray DC current?

[Ansil]

Diodes aren't supposed to heat up when they're in use, are they? If not, then I've got some hassle and don't know what's up.

I've been trying to hook up this circuit:



but using two op amps before the diodes. Two of the diodes are heating up a hell of a lot. I think it's all wired up ok, any ideas what's happening/not happening? The diodes are all 1N914s. Am I just overdoing things?

I'm also getting a very mushy, staticky high end. Has this got to do with the diode problem? Or is there a cap I can stick somewhere to roll off the high end? Or would you expect that from this layout anyway? Or am I still overdoing it?

Cheers.

when you say tow opamps,, are you refering to say like a tl072 using both halves of it.. or are you talking about two lm386's..   cause the circuit will work with either way but basically only designed for one 386 not two..sorry if i was assuming.

k first of all. the gain produced by this circuit is simply that.. i left out all tone shaping out of the circuit.  if you are not using any fx before this pedal then you should be fine without input and output caps to block dc, but then again you don't benefit from the tone tailoring of it also.. i ran this into a semi clean amp to act as a different channel on my amp rather than making it just a pedal to push my clean channel with its own sound.  in other words it is very interactive.   the 386 is bridged and set for a gain of 200.   the four diodes are just your basic silicon diodes they sell at radioshack for small signal..  i do believe they are 1n914's..  just like the schematic shows.. you will need to make basically a loop where the band of the first diode will not be touching the band of the second one.  

then you need to put one side of it  to ground and the other attach it to your input then tie it to a pot.

i guess since the layout is pretty easy, it would behoove you to check the orientation of the diodes and make sure they are going the correct way.  also i can't stress enough if you are using two 386's you can melt the diodes if they are in the wrong way.  

plus the circuit also was designed around teh 386..(did i say that before) cause it pushes the output harder than a regular opamp.

feel free to add your own tone shaping into the scheme. i usually use a .047 input and a .1 output cap then hit the diodes.

http://www.geocities.com/austenfantanio/simplehate

ps you have to go to the link and go in the address bar and take off the http:// and it will show the picture.
or you can just go to the actual page it is on  
http://www.geocities.com/austenfantanio/snips

[Greek Acrobat]

Hehe, I could look like a real twit here....

Does the loop of diodes need to be orientated the right way, ie clockwise/anticlockwise? I just assumed that wouldn't make a difference. (I'm new - never used diodes before)

I think I'm just gonna put that embarrased smiley in my sig. It'll save me sticking it in every other post.  :roll:

Yes, I used 2 386s.

Cheers.

[Ansil]

yep if you use two 386'sthe diodes can get hot.. unless you use like 1n4001's or better, even then  they will probally get warm.  the 386 puts out alot of juice..  that is why it can eat batteries.. but i think it is worth it  considering i dont' havet to spend extra money on a voltage bias ie two resistors and another cap.  and having to wire up all that nonsense..   but anyway  using one lm386 with a .047 cap after the 10k gain pot. and using a .1uf before the diodes will give a nice round tone.  with enough highs cutting through  you can do two 386's in to each other but you need to seriously think about heatsinking them if you run them wide open.  and i usally put a pot in there about 500k to 1M in series with the output of the first and the input of the second.  thats how my little amp is and it works nicely, that way i can get the saturation onf the opamp working hard but not the un nesecary volume.

email me if you need too.

[Greek Acrobat]

Right, I'm really confused now...

I thought I'd check everything from the start. I hooked up the whole circuit again but this time only one 386 and only 2 (new) diodes in parallel.

The diode with the cathode to ground is heating up just as before. It's really quite hot - you know, burning hot. The other is fine - I don't think there's any trouble there.

I'm stuck here, there's nothing that I can think of that's wrong with the circuit. Any ideas?

[aron]

I can't see the picture either, but do you have a capacitor before the diodes to block DC?

Using your meter set to DC, measure the diode that's heating up. Put the red tip on the diode/signal junction and the black on ground.

See what the voltage is on that junction. If there is significant voltage, you need a blocking capacitor before the diode.

Of course I am assuming these are clipping diodes; standard pair from signal to ground.

[Greek Acrobat]

Yes, you assume right - plain old clipping diodes.

How much V is significant? The voltage is 1.09V across the diode - that too much?

[aron]

If you are measuring DC, yes, that's too much.

Put a capacitor before the diodes. You can put a 1uF or larger capacitor before the diodes positive side facing away from the diodes.

Or you can put any film or ceramic type capacitor .1uF or larger before the diodes.

[Ansil]

i don't really have a clue, i never measured it.. basically i did the diodes like this.
from the chip output  and in between the volume pot lies:  

                     +     +
                     d      e
                     i       d
                     o      o
                     d       i
                     e      d
                     +     +      
                     d      e
                     i       d
                     o      o
                     d       i
                     e      d
                     +***+
                         *
                         *
                     *ground*

assuming that the word diode is from looking at the diode with the band on the bottom.  then you would have two diodes connected in series tied to ground..  the the other two diodes would be upside down..  does that help.

[Greek Acrobat]

Thanks aron, that's great. I'll go see what caps I have and give that a go.

Out of curiosity, how much is too much (obviously 1V is)? What's the rule of thumb on this kind of thing? Oh, and what is it that the cap does to help?

Sorry if the questions are a little simple - I'm still learning.

Thanks again.  

[Ansil]

http://www.geocities.com/austenfantanio//386fun.htm


i never had any problems with this.

[Ansil]

the cap blocks dc

[Peter Snowberg]

Out of curiosity, how much is too much (obviously 1V is)? What's the rule of thumb on this kind of thing? Oh, and what is it that the cap does to help?

Sorry if the questions are a little simple - I'm still learning.

I finally saw the schematic and yes.... you NEED a DC blocking capacitor in there to stop DC from flowing to ground. Almost any "large" value will work, but smaller ones will limit the bass response. Try something in the 100 to 470uF range. As you get larger you will get more bass, but the highs may be dimished so you can make up for that by adding a small film cap (say 0.1uF) in parallel.

The LM386 automatically biases the output to 1/2 of the Vcc voltage so the cap stops a constant drain of power that would be there otherwise.

It's inside the speaker so you can't feel it, but your voice coil is subjected to the same force that's heating the diode. It's not good for either one, or for the 386.

With signal lines, you should never see any DC unless the point you measure is being held (is biased) at a particular DC level. I would have to say that the rule of thumb is to have exactly zero volts across the diodes when there is no input. If you see more than the forward voltage of the diode you're using (~.3 volts for Ge and ~.5 - .7 for Si), you will get heating.

The DC blocking cap solves these issues and reduces the battery drain by a HUGE amount.

One last note is that the + side of the cap should face the 386.

Please don't ever worry about any questions; simple or complex. We're ALL still learning (hopefully for life!).

Take care,
-Peter

[Greek Acrobat]

Aha! I tried the cap and it works just the treat. The lowest value I tried was .1uF which worked but I found the best was the highest I had (47uF) which falls in line with that higher value point you made. I'll get a hold of some larger ones and play around with them.

Eek, I didn't realise it could be damaging the speaker too. I suppose it makes sense. Hope there's nothing permanent there.

Thanks a lot Peter, that was really helpful!

Cheers guys  

[Ansil]

unless you are directly running a speaker off of this, it is not damaging your speaker, cuase there are coupling caps somehwere in the amp before you get to the speaker..  that is why i never worried with it when i first made it. cause it went   like this.. guitar>>>> pedal>>>>amp.   and according to my amp schematic there is a capacitor right after the input. ahhhhhhh  see the wisdom in my madness...  yes young skywalker.

of course i never had any heat from diodes either. of coure the second version had a 5mm green Led in there so it would light up when you played effectively absorbing some of the dc..  still no output though.. and the led is in series  with the signal not going to ground.

jsut in case anyone wnated to try it.  similar to the cross over distortion taht is present in the boss heavy metal pedal  hm-2

have fun