Yes, that was my point, I think. If the problem is actually radiation,
then maybe shielding the PSU might be both cheaper and more
efficient than using expensive special diodes? There has been
so many threads on the diode topic, but I can't remember
anyone being specific about if the problem is RFI on
the rails or radiated RFI. That is, it seems we have endlessly
discussed solutions without making clear what the problem is,
or maybe I have just missed it or forgot it?
then maybe shielding the PSU might be both cheaper and more
efficient than using expensive special diodes? There has been
so many threads on the diode topic, but I can't remember
anyone being specific about if the problem is RFI on
the rails or radiated RFI. That is, it seems we have endlessly
discussed solutions without making clear what the problem is,
or maybe I have just missed it or forgot it?
I'm completely serious. We are referring to MUR860, formerly a motorola prodoct, now produced by OnSemi, Fairchild, et. al? I tried these and they produced a lot of ringing in the power supply. Snubbers were mandatory. I had the same experience with MUR1620.
The datasheets don't even claim soft recovery for these parts. The recovery is uncharacterized.
I found the IRF HFA series and various Schottky parts are much easier to work with.
The datasheets don't even claim soft recovery for these parts. The recovery is uncharacterized.
I found the IRF HFA series and various Schottky parts are much easier to work with.
fdegrove said:
I'm Pickering why it is when I give a rule of thumb for choosing PS diodes it somehow gets over looked....
I can assure you I paid attention to it and memorized it. Howver,
although I believe it is a reasonable rule-of-thumb and I am sure
you wouldn't recommend it if it had caused you trouble, I still
think there may be cases where the rule fails. They might be
rare though, being cases with unusally small conduction angle.
Hi,
I wasn't aiming at you to paraphrase Phred...
You could probably come up with hundreds of appplications where this rule of thumb wouldn't fly...which is why it is what is
...a rule of thumb.
Still, for all intents and puproses called audio it never failed me so far. Guess you can call me lucky then.
Cheers,
I wasn't aiming at you to paraphrase Phred...
You could probably come up with hundreds of appplications where this rule of thumb wouldn't fly...which is why it is what is
...a rule of thumb. Still, for all intents and puproses called audio it never failed me so far. Guess you can call me lucky then.
Cheers,
fdegrove said:Hi,
I wasn't aiming at you to paraphrase Phred...
You could probably come up with hundreds of appplications where this rule of thumb wouldn't fly...which is why it is what is
Still, for all intents and puproses called audio it never failed me so far. Guess you can call me lucky then.
Cheers,
Nah, it is probably a good rule of thumb. It's just that I am
a scientist and used to think about worst cases.
Anyway, if in doubt, use some heavy duty diodes and check
what the charge currents are and then select diodes
depending on the outcome of this test.
Edit: Of course both you and I know enough to realize it
was just a rule of thumb, but many others with less
understanding of diodes and charge currents may not
realize it, so I thought it better to play safe and point it out.
jwb said:
This of course, this should be taken care of in board design, with good layout. Both on the amplifier, and power supply. (Think Twisted pairs for signals and returns, etc.)
It's probably overkill, but I try to deisgn my boards like I would for an RF amplifier. Lots of extra vias to reduce ground plan inductance. All of the traces (or most of them) on the bottom side, with extremely short path lengths.
And what about the transformer? Transformers can emit lots of magnetic flux (That is what your probe is looking for).
just my $.02
-Dan
fdegrove said:Hi,
Somehow I vaguely recall Peter Daniel mentioning that not all MUR860 diodes sounded alike...depending on where they were coming from.
I'm Pickering why it is when I give a rule of thumb for choosing PS diodes it somehow gets over looked....
Cheers,
Sorry fdegrove. Thnaks
OK so 3 times the RMS voltage ! How about current. Could You express the ampere rating as a function of estimated continous output power in a specific impedance using a full bridge for one channel ?
Then I suppose we got it !
BR /
Hi,
The issue for the diodes isn't the continous power as much but the inrush current of the circuit at switch-on.
That is of course dependent on the load attached to it but in general I also triple the current rating on a per bridge basis as you're using four diodes that are sharing the cycle duty.
If you want diodes to withstand higher inrush currents it can be helpful to put a pair in series if the voltage drop isn't a problem.
Cheers,
The issue for the diodes isn't the continous power as much but the inrush current of the circuit at switch-on.
That is of course dependent on the load attached to it but in general I also triple the current rating on a per bridge basis as you're using four diodes that are sharing the cycle duty.
If you want diodes to withstand higher inrush currents it can be helpful to put a pair in series if the voltage drop isn't a problem.
Cheers,
Some years ago I've made "living on the edge" experiment. A schottky dioda is only 40V max. I wanted to built +/-60V bridge. So I put 2 of these schottky diodes in series for every bridge arm. It doesn't work. The first dioda always broke.
Making small voltage dioda in series doesn't double the voltage rating?
Making small voltage dioda in series doesn't double the voltage rating?
fdegrove said:
Is it me being tired or is there something wrong here? How
would connecting diodes in series increase their current
capability? Connecting diodes in parallel may increase the
current capability, but it is not a straightforward thing to do
since nothing guarantees equal current sharing between
them.
Diodes In Series
Hi all,
noticed an interesting posting about diodes in series, yes two diodes can change property in how the act at both turn-on AND turn-off but there may be some serious issues to pay attention to.
For couple of years ago i replaced in a SMPS a single diode to at that time a new diode type from ST which internally was a doubble diode connected in series, the switching losses went down BUT the EMC meassuring was worse meaning more RF dirt in the supply voltage.
I have not investigated the issue well enough but I suspect doubble diodes might evetually be much more nasty at turn off so this should be researched.
Input from others who have researched more thoroughly in the matter are wellcome!
Cheers
Hi all,
noticed an interesting posting about diodes in series, yes two diodes can change property in how the act at both turn-on AND turn-off but there may be some serious issues to pay attention to.
For couple of years ago i replaced in a SMPS a single diode to at that time a new diode type from ST which internally was a doubble diode connected in series, the switching losses went down BUT the EMC meassuring was worse meaning more RF dirt in the supply voltage.
I have not investigated the issue well enough but I suspect doubble diodes might evetually be much more nasty at turn off so this should be researched.
Input from others who have researched more thoroughly in the matter are wellcome!
Cheers
Hi,
Not the continous current capability but the ability to withstand the inrush current better at turn-on.
Look at it this way: the series of two diodes has twice the PIV of the single diode and the higher series resistance act as an inrush current limiter.
I never said it was a cure all solve all but it can save your day...when you don't have current limiting resistors handy, when the diodes in the junkbox are rated a little too close for comfort, etc.
So basically: D1+D2 = 2*PIV and 2*FV, as you see continuous current handling has not changed one bit as is to be expected.
Cheers,
Not the continous current capability but the ability to withstand the inrush current better at turn-on.
Look at it this way: the series of two diodes has twice the PIV of the single diode and the higher series resistance act as an inrush current limiter.
I never said it was a cure all solve all but it can save your day...when you don't have current limiting resistors handy, when the diodes in the junkbox are rated a little too close for comfort, etc.
So basically: D1+D2 = 2*PIV and 2*FV, as you see continuous current handling has not changed one bit as is to be expected.
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.