Diodes and Noise

[tubeswell]

You guys think UF4007 need snubbers?

Probably not - but 1N4007 definitely benefit from snubbers in my experience.

[iknowjohnny]

You guys think UF4007 need snubbers?

Probably not - but 1N4007 definitely benefit from snubbers in my experience.

I don't hear any noise, either buzzing when not playing or anything buzzy on top of the notes. But do you think it's possible that using snubbers may just improve the tone by eliminating something i'm not noticing but would cause a noticeable improvement if it were gone?

[tubeswell]

You guys think UF4007 need snubbers?

Probably not - but 1N4007 definitely benefit from snubbers in my experience.

I don't hear any noise, either buzzing when not playing or anything buzzy on top of the notes. But do you think it's possible that using snubbers may just improve the tone by eliminating something i'm not noticing but would cause a noticeable improvement if it were gone?

As I said earlier, I noticed that the snubbers (when used with the 4 x 1N4007s in the FW rectifier I used in my 6G15 style unit) eliminated an audible buzz. Mind you, I reckon that was at least partly because the 6G15 is a reverb unit and it is plugged into the front of another amp, and any buzz that gets into the (sensitive) reverb unit signal path is further amplified in the subsequent amp. So in this case I would say that using snubber caps is definitely beneficial.

Now when I say 'audible buzz', it was not a buzz that I heard when playing notes. Rather it was a buzz that was there when the amp was merely switched on. Grounding aside (and I know this because I got eventually got the grounding in the 6G15 so perfect that ground hum was all but eliminated) there was still an audible noise floor buzz (that originated in the 6G15 unit) without the snubbers. The conclusion I came to about all of this (which I have certainly not (to-date) disproved in my experiment with removing them and putting them back again), was that the snubbers actually eliminated this buzz. I should add that I conducted that experiment after I had got the ground hum eliminated.

Are my ears sensitive? possibly. Did I imagine all of this? Definitely not. Is diode switching noise a phenomenon that is only experienced by me? I don't think so. As I noted earlier, Merlin Blencowe notes it (in the articles on power supply design on his website) and no doubt others are probably also aware that it could be potentially problematic.

There probably is some buzz from switching noise with those type of diodes in any amp. But it might well be so tiny/faint as to seem inaudible if the amp doesn't contain any sensitive signal path (like a reverb circuit), which otherwise could accentuate it.

FWIW I have also built a tweed bassman style amp with reverb using SS rectification (with 6 x 1N4007 and snubber caps) - and the noise floor was very quiet/silent. However I must also add that in that case, I decided to use snubber caps in that amp after my positive experience with the 6G15 unit, and I haven't tried swapping the snubbers in and out on that other amp to see if they make any difference.

[PlinytheWelder]

The switching spikes may be in the MHz range and inaudible, but the IM products may be audible.

[Structo]

I recommend the Fairchild Stealth dual diode hexfred that is 8A per side, 1200V rating, you only need one of them, Mouser charges about $1.00 each, you cannot beat them with two sticks.

No noise, soft recovery, one simple TO-220 package does it all. In audio applications they are barely ticking over.

Separate anodes, common cathode, just what you need, good for 400-500 Vp with no nasties.

Mouser P/N 512-ISL9K8120P3. They are not ROHS compliant.

WOW!
That looks like a great solution for a SS rectifier in a guitar amp.
$1.24 each.

Making note of that part in my amp notes right now!

BTW, the part number you listed is obsolete, the new number is
512-ISL9K8120P3_Q and it is not ROHS

[C Moore]

I recommend the Fairchild Stealth dual diode hexfred that is 8A per side, 1200V rating, you only need one of them, Mouser charges about $1.00 each, you cannot beat them with two sticks.

No noise, soft recovery, one simple TO-220 package does it all. In audio applications they are barely ticking over.

Separate anodes, common cathode, just what you need, good for 400-500 Vp with no nasties.

Mouser P/N 512-ISL9K8120P3. They are not ROHS compliant.

WOW!
That looks like a great solution for a SS rectifier in a guitar amp.
$1.24 each.

Making note of that part in my amp notes right now!

BTW, the part number you listed is obsolete, the new number is
512-ISL9K8120P3_Q and it is not ROHS

In a "typical" tube, guitar amp, can the heat sink on this diode be ignored?
Thanks

[David Root]

Mea culpa, Tom, I missed the _Q on the P/N.

I bought two of these about 3 years ago or so, and I swear they cost three times as much each then as what I just paid for ten more.

Recently I put the second one of the original two in my 50W #13 ODS in the Princeton chassis, absolutely dead quiet, AC on the preamp tubes' heaters, (I normally used 12VDC) AND I used carbon comps in the OD section too, as per Tony's inspired recipe. Cranked, there is a little hiss from them, but zero hum.

I recently found a single diode for a Music Man 700V supply. Also a 220 package, made by IXYS. 1800 V at 2.3 A. Two of those would be all you need, and noooo snubbers need apply. Mouser P/N 747-DSA1-18D.

[David Root]

Need a heat sink? Probably not, but I use a small one anyway as I do not like to put high voltage TO-220 packages on the chassis because the metal back of the thing is connected directly to the cathode.

I know there is the little insulation kit with the mica and the nylon shoulder washer, but then you need the thermal grease, which is a PITA, and if you pinch the shoulder washer slightly screwing the whole thing down to the chassis, you can lose the insulation to the 4-40 screw and whammo! there goes your AC fuse! Ask me how I know this.

I use a small heat sink designed for a single TO-220 package, that I got a dozen of from Apex Jr. for about 15 cents each I think. They are black anodized aluminum and stamped THM 6079.

[Structo]

So you are using the heatsink on the diode, with the sink just hanging in the air?
Not bolted to the chassis?

If so I know what you mean, it's always made me a bit nervous depending on a paper thin mica insulator and a tiny nylon shoulder washer to keep high potential from ground.

[Structo]

On the data sheet for the Stealth diode is says:

The soft recovery minimizes ringing,
expanding the range of conditions under which the diode may
be operated without the use of additional snubber circuitry.

So is the noise associated with SS diodes more properly called ringing?

[Ken Moon]

The proper technical term is "schmutz"

[20to20]

On the data sheet for the Stealth diode is says:

The soft recovery minimizes ringing,
expanding the range of conditions under which the diode may
be operated without the use of additional snubber circuitry.

So is the noise associated with SS diodes more properly called ringing?

Yes, but don't lump the term "noise" as in audible artifacts with "electronic noise".

The recovery time for this diode is similar to other fast recovery rectifiers.
"Fast Recovery. . . . . . trr < 25ns"

The "noise" is Mhz. and is relevant to high freqeuncy circuits. RF type.

[David Root]

Tom, the heatsink is on the HV PS/bias board, not exactly "hanging in the air" but no not fixed to the amp's chassis.

OK from a dissipation point of view it might as well be hanging in the air.

[VacuumVoodoo]

The switching spikes may be in the MHz range and inaudible, but the IM products may be audible.

Switching spikes are at the switching frequency i.e. full wave rectification of 60Hz ac voltage will produce spikes at 120Hz. As the spikes get shorter their content of high frequency components increases in both amplitude and bandwidth. They contain components in the high MHz range but their fundamental frequency is still double the AC frequency.
The buzz is the 120Hz fundamental and it's first 2 to 7 overtones, the over 50kHz and MHz portion of the HF soup will screw up your radio communications. That's when FCC starts chasing you.

[M Fowler]

Does the type of diodes used in an amp have anything to do with slow amp or fast amp?

My rocket octal power section using IN4007 diodes verses Henry's Red Plate Blackverb (Dumble) style circuit the Red Plate is fast and easier to play compared to my Rocket both are 50-60 watts range. Or is that more in the filter stages and choice of choke etc?