Any experiences bypassing UF4007 with .01uF cap?

[mtl777]

Hi, has anybody tried bypassing a UF4007 diode with a .01uF cap? This technique is reported to work well with 1N4007's to remove RF hash in power supply circuits. I wonder if it would also work well with UF4007's which are already low noise. So I would greatly appreciate if anybody could share their experience -- whether it improved or did not make any difference or even degraded the performance.

Thanks!

[nigelwright7557]

A diode can help but too large a value will push through AC on to the DC.

I tend to put a 100nf across the supply rather than across the diodes.

[whitelabrat]

What I've done is to use two UF4007's stacked and then bypassed both with a 0.01 uF. Cleaned up switching noise nicely.

[mtl777]

Originally Posted by nigelwright7557

A diode can help but too large a value will push through AC on to the DC.

I tend to put a 100nf across the supply rather than across the diodes.

You mean a bypass cap can help. For a typical bridge circuit that has 4 diodes, what would be the recommended value of the bypass cap to be placed across each diode? Is .01uF (10nF) a good value or is it too large?

Originally Posted by whitelabrat

What I've done is to use two UF4007's stacked and then bypassed both with a 0.01 uF. Cleaned up switching noise nicely.

Thanks for sharing! Was the noise clean-up only visible through a scope or did you also notice a change in the sound of your gear? How did the sound change after putting in the bypass cap?

[Bonsai]

See http://www.hagtech.com/pdf/snubber.pdf for an excellent discussion on snubbers and how to select the values.

Take heed on " . . . too large a value will push through AC on to the DC." because you can feed AC line hash onto your DC line.

[mtl777]

Originally Posted by Bonsai

See http://www.hagtech.com/pdf/snubber.pdf for an excellent discussion on snubbers and how to select the values.

Take heed on " . . . too large a value will push through AC on to the DC." because you can feed AC line hash onto your DC line.

Thanks, that's a great article and good info! Unfortunately, I don't have a scope to make measurements with, so will have to rely on people's advice hoping that their recommended value is in the right ballpark. It also gets more complicated with the RC snubber, so I'm hoping a simple C will be good enough even if it's not as good as RC. If I take the simple C approach, what C value would you suggest as a bypass to be placed across each UF4007 diode in a typical bridge rectifier circuit that has 4 such diodes? No need to be super accurate, I just want a value that would work for most circuits.

Thanks!

[infinia]

I just want a value that would work for most circuits

no such thing. It depends wholly upon impedance's and current levels. Higher voltages and lower currents use lower values. Some folks just use a snubber or 2 across the secondary and HF bypasses at the bulk caps as well. With just caps (low loss snubbers) you end up pushing the noise somewhere else.
Post a circuit and you'll get better opinions.

[mtl777]

Originally Posted by infinia

no such thing. It depends wholly upon impedance's and current levels. Higher voltages and lower currents use lower values. Some folks just use a snubber or 2 across the secondary and HF bypasses at the bulk caps as well. With just caps (low loss snubbers) you end up pushing the noise somewhere else.
Post a circuit and you'll get better opinions.

Please see below. It employs a voltage doubler circuit that is used to generate +15V and -15V, and two half-wave rectifiers to generate the +5V rails. The 9VAC input is from a wall wart rated for 3.5A that converts the mains 110VAC to 9VAC. I'd like to know what C value (or better, R and C values) I should put across each of the diodes D1, D2, D3, D4, D5, and D12. The diodes are currently 1N4001's but I'll replace them with UF4007's which are what I have available.

BTW, it is the power supply of the Delta 1010 sound card.

Thanks!

[Bonsai]

MTL777, its easy to obsess about snubber values, but at the end of the day if you are 90% there, then you have probably solved the noise problem. For small diodes (i.e. up to about 2 or 4 amps) and bridges, I use 1nF and 100 Ohms. Try that and it should do the trick.

[prairiemystic]

Are you talking low-voltage ~30VAC or high-voltage ~650VAC?
I always put 0.01uF's across each transformer secondary winding. I checked with a 100MHz scope and the UF4007 switches very clean at low voltages.
When you get up past 300VAC, in tube gear, then you have to use snubbers.