Last night a breadboarded this pretty basic circuit. I've got a 12v wall wart running into an LM7809 voltage regulator with a transistor bypass. This part all seems to work just right. My goal is actually to get around 6 volts (to run a speaker) and 5 volts (to run an MP3 player).

So from the 9volt output of the regulator & transistor, I ran it into 4 1N4001 daisy chained and that gave me about 6.3 volts (according to my multimeter) and then chained 2 more off of it to get my roughly 5 volts (again, from the multimeter.) At the end, I ran a 1M resistor to ground. So I was getting about 0.7 volt drop per diode.

Last night I breadboarded it, everything worked beautifully. I left it running for a couple of hours, no problems at all.

So today I soldered it all together and the output from the voltage regulator is 9v, but the diodes aren't dropping nearly as much as they were last night. Now I'm getting about 0.3 volt drop per diode.

I don't have a lot of diodes to choose from. I have the 1N4001s, 1N4739s, 1N4148, 1N5817, and 1N914s. The 1N4001s seemed like they'd be the best choice, but maybe not.

How could the voltage drop have changed so much, though (there's no detectable heat to the touch from them).

Thanks for any enlightenment anyone can provide.

 

Do you have them forward biased sufficiently?
1meg is on the high side.
Max.

 

Last night a breadboarded this pretty basic circuit. I've got a 12v wall wart running into an LM7809 voltage regulator with a transistor bypass. This part all seems to work just right. My goal is actually to get around 6 volts (to run a speaker) and 5 volts (to run an MP3 player).

So from the 9volt output of the regulator & transistor, I ran it into 4 1N4001 daisy chained and that gave me about 6.3 volts (according to my multimeter) and then chained 2 more off of it to get my roughly 5 volts (again, from the multimeter.) At the end, I ran a 1M resistor to ground. So I was getting about 0.7 volt drop per diode.

Last night I breadboarded it, everything worked beautifully. I left it running for a couple of hours, no problems at all.

So today I soldered it all together and the output from the voltage regulator is 9v, but the diodes aren't dropping nearly as much as they were last night. Now I'm getting about 0.3 volt drop per diode.

I don't have a lot of diodes to choose from. I have the 1N4001s, 1N4739s, 1N4148, 1N5817, and 1N914s. The 1N4001s seemed like they'd be the best choice, but maybe not.

How could the voltage drop have changed so much, though (there's no detectable heat to the touch from them).

Thanks for any enlightenment anyone can provide.

The 5817s are Shottky barrier diodes with a forward drop of about 0.3V, can't remember off the cuff whether the 4739 is a fast diode, but the fast ones have a very slightly lower drop than bog standard mains frequency types. The 4148s & 914s won't handle the sort of current it sounds like you're aiming for.

 

The 1N4000 series is represented as dropping .6 volts at 10 ma. When you attach your "real" load, it will be even more.

 

Changing out the resistor fixed the issue, but the speaker may be running a bit lower than I'd like. I'll check the amperage on it in the morning (right now it's powering the MP3 player & speaker playing rain and crickets while my 3yo daughter sleeps).

I didn't think to look for a forward voltage graph. I just looked at the value in the Electrical Characteristics and of course, it didn't reconcile with what I was seeing which makes perfect sense now.

I really appreciate the help. I've been tinkering with electronics for a bit, but I really have a lot to learn. You'll probably be seeing more posts from me real soon (maybe even tonight).

 

Changing out the resistor fixed the issue, but the speaker may be running a bit lower than I'd like. I'll check the amperage on it in the morning (right now it's powering the MP3 player & speaker playing rain and crickets while my 3yo daughter sleeps).

I didn't think to look for a forward voltage graph. I just looked at the value in the Electrical Characteristics and of course, it didn't reconcile with what I was seeing which makes perfect sense now.

I really appreciate the help. I've been tinkering with electronics for a bit, but I really have a lot to learn. You'll probably be seeing more posts from me real soon (maybe even tonight).

IIRC, you can actually get 7806 regulators, if not you could put a couple of series diodes in the ground of a 7805. It is also possible to bias the ground from a tapping across the regulated output - this is not really any different to how you set the voltage on an LM317.

Either way, the Vf of diodes does vary a bit with current, so one of the possible regulator options will give a fair bit more accurate regulation.

 

Yeah, I was just trying to work with what I had. What I had was some diodes and an LM7809, so I went with that. I could just get a 7806 and a 7805 and call it good. I may actually end up doing that. For one thing I think I'm going to need some ferrite beads because I'm getting interference from a cable modem and wireless router that are just a few inches away. So if I'm going to have to stock up on more stuff, I may as well get the 7806 and 7805 (or the 7806 and a couple of diodes).