Semiconductor diodes

[Leevi]

I have met quite often problems with some diodes. The last example was BY299
which was suddenly burned in a circuit where the current and voltage were much lower than what the diode specifications state.
As a consequence of a burn rectifier diode there is a risk that the PT will burn as well.
 
http://www.datasheetcatalog.org/datasheet/panjit/BY298.pdf

I have also been disappointed with 1N400x series diodes.
http://www.datasheetcatalog.org/datasheets/37/132863_DS.pdf

What do you think which one will normally cause the burning: current or voltage?

Can you recommend some good diodes.

I have used successfully the following which is actually a bridge but
can be used as single diodes.
http://www.datasheetcatalog.org/datasheet/good-ark/KBPC3502.pdf

/Leevi

[Iannone]

Without knowing more about the power supply you put these diodes into its hard to be sure, but I suspect that the BY299 is too little diode for most guitar amps.  The PIV rating of 800 V is going to be inadequate for most amps.  Here's why:

During the positive part of the cycle, the cap connected to the diode's cathode is going to charge to whatever the B+ voltage is, lets say 425V (for a BF-style 6V6 amp).  During the negative portion of the cycle, the cap stays at 425V while the the anode of the diode is at -425 V.  The voltage across the the diode is now 850 V.  Thus, your diode needs a PIV rqating is AT LEAST (and preferably a bit more) TWICE your expected B+ voltage.

[FYL]

I have met quite often problems with some diodes.

A large number of active components are fakes, mainly those sold retail in small quantities on the net or elsewhere. You should only source components from trusted sources, and even then, brace yourself for a few bad surprises.

[PRR]

> cause the burning: current or voltage?

Yes.

> current and voltage were much lower than what the diode specifications state.

Do you know how to estimate what actually happens to the DIODE? It is not the same as what comes out as DC. As Iannone says, the peak voltage is often twice the DC voltage, plus all surges and transients on your wall-power. The start-up current is over ten times the transformer rating, the RMS is at least three times the DC current.

The 800V diode is at its voltage limit in a small Champ.

I built an over-size Champ with doubler supply, and those 2A diodes would be good.

1N4007 is not reliable in bigger four-6L6 amps. 500V DC supply means 1,000V on diodes. 250mA/pair current means 1.5A diode current.

The KBPC3502 used as Bridge would be ample for less than 1,000V DC with any practical number of bottles. Used as single diodes (internally series-parallel) with the "2 diode CT" rectifier plan it is only safe to less than 500V. It is 20 times more expensive than 1N4007; as you've discovered, under-buying rectifiers is more expensive than over-buying.

[Leevi]

OK, the BY299 was working on its maximum limits and was probably burned
by some peak voltage when the amp was switched on.

It's quite difficult to trust on 1N4007 because of its small physical size and just believe
that 1000V is OK. I think 20 times higher price for more robust diode is just saving.

/Leevi

[38Super]

It's quite difficult to trust on 1N4007 because of its small physical size and just believe
that 1000V is OK.

Mouser has an end of life on Fairchild 1N4007_Q for $0.02.  Less if you buy 25 or more.  Good practice is to place two in series with snubbing caps across each individual diode.  that will give you 2000V reverse voltage capability.

These diodes are rated at 1.1A continuous with 30A surge.  To draw more than 1.0 amp continuous off your HT winding is not likely.

cheers,

rob

[PRR]

> To draw more than 1.0 amp continuous off your HT winding is not likely.

The spike-wave through a rectifier is MUCH more brutal than the DC current.

Rectifiers must be rated two or three times the DC current.

Also I bet the 1.1Amp number may be de-rated in a hot chassis.

So say 0.3A or 300mA. A couple 6L6 or EL34 in 4KCT load can touch that.

I doubt 1N4007 is "end of life". I think what that note means is even Fairchild can't break-even on commodity stuff like rectifiers, too many other players have the vats and tools to make rectifiers.

[Leevi]

I have used  KBPC2506 bridge (600V_RRM/420V_RMS) in many amps without problems even with higher voltages
than what has been given in the specs. I think the strong construction makes it more robust.
/Leevi

[Tone Junkie]

Why not just use a 5408 diode and not worry about it anymore. if my memory serves me right unfortunatly it lets me down often, its rated at 1000v at 3 amp instead of 1000v 1amp for 4007.
Bill

[Tiny_Daddy]

You can install 2 diodes in series for higher voltage, instead of just one diode. No balancing resistors required at amp voltage levels. BTW Doug sells the 3 amp diodes. Also you can buy a 1.5 amp bridge rectifier and cut off the AC pins and install it as one diode. That's what I've got in my stereo supplying 750 volts to the regulator tubes.

[Leevi]

Why not just use a 5408 diode and not worry about it anymore. if my memory serves me right unfortunatly it lets me down often, its rated at 1000v at 3 amp instead of 1000v 1amp for 4007.

I ordered couple of days ago a set of 1N5408 and hope I don't need to worry anymore. But that I'll know after I have used them in  my amps.
I have also ordered KBPC3510 bridges i.e. 1000V/700V,  35A!

http://pdf1.alldatasheet.net/datasheet-pdf/view/61416/FUJI/KBPC3510.html

/Leevi