Component ID

irobot · Joined: 13 Mar 2009 Posts: 49

Having trouble trying to identify a diode.

It is marked "K3" with the numbers "78" printed in a vertical format (probably lot/date code) after the "3". It is a conventional axial thru-hole package.

I'm pretty sure this is a schottky diode, made by either Sanken or Shindengen.

Thanks,
Mike

vtech · Joined: 08 May 2006 Posts: 1264 Location: USA

Mike,
What circuit is it in and what lead you to think it is a schottky? (Does it measure like one on a diode test function?)
Reason I ask is according to two of the older reference books that I have "K3" is shown to be an equivalent to 1N34A which is a very old germaniaum diode used in RF application. Is it in some sort of an RF circuitry? Not sure if that is what you have?
Also, if you were to check a germanium diode on a diode tester, it will behave very similar to a schottky. (both check w/ low forward voltage drop)

irobot · Joined: 13 Mar 2009 Posts: 49

Vtech thanks for the response.

I should have been more specific.

This is one of several diodes in a switch-mode PS from a Tascam/Teac semi-professional sound mixer/PC interface. I have a schematic for a similar model that uses schottky/barrier diodes and with other vague similarities.

The main supply voltages (+-12V, 5V, 3.3) are clean and rock-steady.

There is a separate +48VDC ("phantom") supply used for micorphone bias that's giving me the problem. It's not stable. There is also a ~1.8VDC voltage in the same area that pulsates at a rate of about 1 Hz.

I can't locate a service manual/parts list. I tacked soldered in a IN4007 (1 amp) as a sub for the +48V rectifier (also replaced filter caps) and tack soldered in a 150V 3A schot/barrier diode for the one in question.

Voltages seem OK in this configuration, but both my subs seem to run warm.

It's possible I need something like a 2A instead of a 1A 1N4007.

As far as the schot/barr diode I'm puzzled. The original reads like it should in the forward bias direction, but when I use an ohms scale I get a pulsating reading (positive lead on the anode). So that's what led me to believe that it's leaky. I had also subbed a conventional rectifier for this, but then the tranformer gets hot.

Any thoughts/suggestions?

Thanks,
Mike

irobot · Joined: 13 Mar 2009 Posts: 49

I meant to mention the presumably schottky/barrier diode - I tested out of circuit, of course. . .

vtech · Joined: 08 May 2006 Posts: 1264 Location: USA

Ok, I see.
Assuming the 48V in question is scan drived, you must use a fast switching rectifier. The main reason behind using a schottky is to provide the lowest possible drop across the diode & less noise.
Are you saying there are 2 diodes in the 48 volt line? I am not sure if I followed you there?
I can tell you that in SMPS applications the recovery time is the single most important in transistor/rectifier selection and the typical symptom is a "warm running" diode. The 1N4xx series are generally not good candidates in switch mode application due to slow recovery time which translates in to heat. Of course you have to also count the current draw.

irobot · Joined: 13 Mar 2009 Posts: 49

Ahhhh . . Fast or ultra-fast recovery diode - that's most likely the problem. I didn't think of that!

The rectifier I subbed (IN4007) is in the secondary side, conventional half-wave config with filter cap-series choke-filter cap configuration. I didn't measure the frequency at the anode, but it must be relatively high.

The schottky (?) is in the primary side circuitry, and actually, my sub runs at a normal temp. At one time I subbed this with a conventional 1A diode, which also ran at a normal temp. (I need to add that my testing was done in a very controlled manner!).

I really appreciate the advice. Let me find something suitable and I'll post my results.

Thanks again,
Mike

irobot · Joined: 13 Mar 2009 Posts: 49

Typo in the above. Meant to say I didn't measure the frequency at the anode, but it must be relatively high.

I replaced the diodes in question with ultra-fast recovery types. That resolved the heating issue! Everything runs very cool after a several-hour burn-in test. The +48V supply is perfectly stable *but* I'm only measuring +29V.

There are two windings coming off the transformer. One feeds the +48V and the other feeds a lower supply voltage which I believe should be about +5~10 volts (judging by the 16V voltage rating on the filter cap). This voltage reads about .7V

To restate, both these supply voltages are derived from simple half-wave rectification. The 48V has a 3W, 1.8K resistor directly across it's output. The other lower supply has a 3W, 18 (eighteen) Ohm across it's output.

I'm stumped!