I want to use a fast recovery diode for a pre amp B+ supply... Disregarding price, are there any differences between these three choices one should be aware of that might change the design from using one choice to the next? They are all quality diodes from what I see but I want to know I at least evaluated them all as a possibility.
1 - UF4007
2 - A selection from the Vishay HEXFRED series
3 - Cree makes high voltage Schottky diodes now, I always thought Schottky were for lower voltages but here they are:
C6D06065A Wolfspeed / Cree | Mouser
1 - UF4007
2 - A selection from the Vishay HEXFRED series
3 - Cree makes high voltage Schottky diodes now, I always thought Schottky were for lower voltages but here they are:
C6D06065A Wolfspeed / Cree | Mouser
The UF4007, or even the 1N4007, have been used for decades in power supplies with no detrimental effects, and outstanding reliablity.
That is, of course, if you're using them in a "standard" AC transformer supply, not an SMPS.
An in my opinion, a preamp should not be powered with an SMPS anyway.
That is, of course, if you're using them in a "standard" AC transformer supply, not an SMPS.
An in my opinion, a preamp should not be powered with an SMPS anyway.
The original 1N4007 is surely going to need a snubber at each diode to get rid of the 120Hz buzz caused by the reverse recovery current pinging the secondary winding.
Those Cree diodes have a high price to match their high reverse voltage, though for four in a full wave bridge and a single supply, it's still a small fraction of a total power supply cost. The data sheet is adamant and clear about "Zero Reverse Recovery Current" and "Zero Forward Recovery Voltage" - I'd go ahead and try them.
Those Cree diodes have a high price to match their high reverse voltage, though for four in a full wave bridge and a single supply, it's still a small fraction of a total power supply cost. The data sheet is adamant and clear about "Zero Reverse Recovery Current" and "Zero Forward Recovery Voltage" - I'd go ahead and try them.
benb, maybe you have tinnitus if you are hearing 120Hz buzz from pretty much every speaker connected to mains powered equipment in your house ?
No joke, as audiophiles have been in rapture about their stereo equipment for decades, and almost nobody has ever complained about 120Hz buzz from audio equipment (and when they do it is typically not because of diodes).
No joke, as audiophiles have been in rapture about their stereo equipment for decades, and almost nobody has ever complained about 120Hz buzz from audio equipment (and when they do it is typically not because of diodes).
Last edited:
All my equipment with 1n400x diodes had .01 uf disk cap snubbers from the factory. Even the cheapo dynaco stuff. One across the bridge did it.
Indeed, the amplifier that I built and shown on this site, I used a single 0.01uf cap across the AC pins of the bridge rectifier - something done with commerically manufactured products.
Dead silent the amp is with no signal, even with headphones on.
Even the 'scope shows a straight line.
Here are ceramic caps, 0.01uF at 500V, these should be suitable for the secondary of any line-level tube device. For solid state, 100V should do, and there's plenty of them.
There ARE some thru-hole parts here, just keep scrolling:
Ceramic Capacitors | Capacitors | DigiKey
In that case, you may be wondering why the OP is asking about 'fast' diodes instead of 1N400x's.
Schottky diodes normally need good heatsinking at high voltage as the leakage current rises exponentially with temperature and thermal-runaway needs to be avoided - these are SiC devices though which seem to be less susceptible to this as well as higher voltage ratings.
Although schottky diodes don't have minority carrier storage issues, they still have junction capacitance which acts to slow down switching - I'd compare devices by the VI switching waveforms in the datasheets if possible as that's the most informative data.
I would suggest using X2 rated satefy caps for this instead of ceramic.
Film Capacitors | Capacitors | DigiKey
Film capacitors can often cause ringing to be worse because of their high inductance and high Q. You need to add a resistor in series to make an effective snubber. Those class 2 ceramic discs have a high dielectric loss and act like they have a built in resistor at RF frequencies (where the ringing actually happens). The ceramic disc is really the right component. Worried about voltage safety? Get 1 or 2kV rated parts.
So why is it “120 Hz buzz” that you hear when it is an RF phenomenon? Because it cuts in and out at a 120 Hz rate, upsetting the circuit somewhat on each occurrence. And some circuits are susceptible to it and others just don’t care, operating dead silent even with those bursts of RF on the supply separated every 8.3 milliseconds. When it does happen it is definitely not in your head.
So why is it “120 Hz buzz” that you hear when it is an RF phenomenon? Because it cuts in and out at a 120 Hz rate, upsetting the circuit somewhat on each occurrence. And some circuits are susceptible to it and others just don’t care, operating dead silent even with those bursts of RF on the supply separated every 8.3 milliseconds. When it does happen it is definitely not in your head.
I haven’t seen any new equipment that has used capacitors on each diode in over twenty years, much less ceramics.
Search for snubber + diyaudio and see what you find.
I got the better results from using soft recovery diodes along with a snubber, but a decent snubber circuit will get you 90% of that with any diode.
Search for snubber + diyaudio and see what you find.
I got the better results from using soft recovery diodes along with a snubber, but a decent snubber circuit will get you 90% of that with any diode.
They don’t do it with switch mode power supplies, at least not on the mains rectifier. That would require X caps (because it’s connected to the line), and the required EMI filter would take the ringing out anyway. They normally have X caps in the filters. The Magnetics in the filters introduce the RF losses. On the secondary side fast diodes are required just due to heat issues, and usually have snubbers somewhere in the secondary circuit (usually right across the little ferrite transformer). It might be a couple of SMD parts that would not be immediately identified as such.
Good question, as there has been no response from the OP but we are up to 16 posts now.
Perhaps the OP wants to put a 'fast diodes inside' sticker on the preamp - to get a nice warm feeling inside when using the preamp - sort of like how 'Intel inside' is used for marketing.
Perhaps the OP has bought a new LISN and wants to check for any change in conducted emissions, or knows a HAM who wanted to use their kit to see if they can detect anything.
Perhaps the OP is that worried about how little they know on power supply layout and amplifier construction that they think some judicious device selection will cure all evils.
I would doubt the OP is making or modifying a smps for their B+ supply - but I could be wrong, as I presently have a small 150W smps module on the bench providing 320V B+ and -60V bias for a 20W amp, and it required a change to the rectifier/filter circuitry.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.