Why the MUR860 diode?

[Eva]

The truth is that most people is using MUR860 just because they saw other people do so and because the TO-220 shiny cases with heatsinks look fancier in their PCBs, but without even understanding ultrafast diodes.

From my point of view it's quite funny to see that happen, because I usually do SMPS stuff and I routinely use most of these diodes in demanding circuits where a standard recovery diode would explode or cause something to explode (that's why ultrafast ones were invented), but for slow 50/60Hz rectification I still use conventional diodes with RC damping networks, because objectively speaking they give very good switching performance and lower dissipation.

 

[Peter Daniel]

Re: Re: Why the MUR860 diode?

the MUR820 has better capacitance and forward voltage drop than 840 and 860, yet, like you, I too have yet to understand why 860 is the 'popular' choice around here..

This is how it all started (note that I specify them as schottky following DigiKey description at that time). This was more than 3 years ago, Eva wasn't even here yet

http://www.diyaudio.com/forums/showthread.php?postid=134303#post134303
http://www.diyaudio.com/forums/showthread.php?s=&postid=146704&highlight=#post146704
http://www.diyaudio.com/forums/showthread.php?s=&postid=136530&highlight=#post136530
http://www.diyaudio.com/forums/showthread.php?s=&postid=137191&highlight=#post137191

 

[N-Channel]

Re: The sun sets on ON's MUR860?

And Carlos likes the MBR16100. Has anyone tried the MBR10100, or compared them? I have some coming from ON Semi.

I have used both the Motorola and newer OnSemi MBR10100s in many SMPSs for Audio Applications. I can find no discernible difference between the two. I do like their lower forward voltage drop as compared to their ultrafast counterparts.

Steve

 

[cpemma]

The truth is that most people is using MUR860 just because they saw other people do so and because the TO-220 shiny cases with heatsinks look fancier in their PCBs, but without even understanding ultrafast diodes.

From my point of view it's quite funny to see that happen...

My usual supplier sells the fast-recovery BYW80-200 at £0.54; googling for a datasheet I spotted the same diode at £2.35 at an audiophile shop. I can't call snake-oil, but it does show how ready the specialist shops are to cash in on a trendy hyper-fi item.

 

[wrenchone]

Even "normal" recovery diodes can cause a lot of broadband RFI when used with a capacitive input filter. The diodes are forced to recover into a voltage source (the charged filter cap), so there can be a substantial reverse recovery transient. I see the effects of this all the time when performing EMI testing on switching power supplies. The signature of this noise is a broadband continuum that rises with decreasing frequency, and is especially prominent down at 150-200kHz. The noise can be substantial all the way down to 10kHz. The reverse recovery current pulses can also interact with stray inductance and capacitance in the current path and cause ringing, enhancing noise production. Changing the normal recovery diodes to glass passivated diodes with a guaranteed 2us trr brings down the noise quite a bit. This is the rationale behind using an ultrafast or Schottky rectifier in a 60Hz circuit. However, if the diodes are not properly snubbed, the problem area just gets shifted to a higher frequency. At the same time, though, the reverse recovery spike for an ultrafast diode will be much smaller and faster than with a normal recovery diode, and hence more easily damped with a snubber.
I generally use whatever ultrafast diodes I have on hand, and damp them with a combination of ferrite beads and RC snubbers.