My Dyna/VTA 70s have a generic 1N4007 diode in the power supply circuit. I just recently saw a post that the Schottky diodes are quieter, so I wondered if it's worth pursuing an upgrade for? If so, which Schottky is a good sub for the 1N4007? Maybe a C3D06060F? TIA!
Completely unnecessary in such a low current drag application (bias).
If you really want to try it, an SB1100 would be fine. https://www.diodes.com/assets/Datasheets/ds30116.pdf
Personally, my bias supply usually uses HER207 because they come installed in the DC-DC convert I use. https://www.mouser.com/datasheet/2/395/HER201 SERIES_E11-249064.pdf
If you really want to try it, an SB1100 would be fine. https://www.diodes.com/assets/Datasheets/ds30116.pdf
Personally, my bias supply usually uses HER207 because they come installed in the DC-DC convert I use. https://www.mouser.com/datasheet/2/395/HER201 SERIES_E11-249064.pdf
Last edited:
If you go for the UF4007, you get a quieter supply. The UF series are equivalent wrt voltage and current but switch faster and softer for less HF junk.
Jan
Jan
before you expend the effort of buying and changing rectifiers, you might want to search for info from Mark Johnson about resistor-capacitor networks that address the root cause of the problem that the high speed rectifiers attempt to reduce.
I think it is interesting we are worried about a little switching noise from the rectifier in a half wave power supply.
It is the difference between a snap-recovery (or hard-recovery or step-recovery when the behavior is intentional) and a soft-recovery diode.
Nowadays, most of the diodes, even the slowest, cheapest ones have a ~relatively soft recovery.
In the early days of silicon rectifiers, this was not the case.
A fast diode will have a short reverse recovery time. A soft recovery diode will allow the reverse recovery charge to recombine softly (i.e., gradually over time) rather than abruptly (like a step function). I write about this at some length here: Taming the LM3886 Chip Amplifier: Rectification and Snubbers – Neurochrome
Tom
Well often not a good idea. Schottky's have high reverse leakage current that increases exponentially with temperature. Usually this means you either use them at low voltage, or you have to heatsink them to prevent thermal runaway (when the voltages are high and the diode gets hot the reverse leakage can become 10's of milliamps and lead to thermal runaway). Standard pn-junction diodes struggle to leak more than a few microamps even under duress, which is never a thermal dissipation issue.
If you don't need to rectify at 10's or100's of kHz (as in a SMPS), and the voltages are medium or large, standard silicon rectifiers are the more sensible option, with fast soft recovery types being the best option, and less heatsinking requirements (heatsinking hardware gets expensive and takes up space and weight).
1N4007 are Rs. 3000 - $40 - for 10,000 in reels, reputed brand, bulk price, in India.
0.4 cents US each.
If the slightly higher cost of the options above would have been offset by the reduced cost of snubbers, the manufacturers of SMPS would have done it long back.
You also have to meet certification standards like UL and others if you sell them, so hazards have to be kept in mind.
Among the reasons 1N4007 are still used.
We spent much more than a few cents in discussing this...
0.4 cents US each.
If the slightly higher cost of the options above would have been offset by the reduced cost of snubbers, the manufacturers of SMPS would have done it long back.
You also have to meet certification standards like UL and others if you sell them, so hazards have to be kept in mind.
Among the reasons 1N4007 are still used.
We spent much more than a few cents in discussing this...
Snubbering doesn’t help switch mode supplies. Power burnt up in the resistor is CV^2*f for square wave drive. Ok at 120Hz, not so ok at 120k. The only way to avoid the *power loss* is speed up the diode’s reverse recovery and eliminate (or minimize) the snubber.
If you can buy the faster rectifier for less than the cost of the cap and resistor it may be an option for line frequency supplies. If it actually eliminates the ringing and not just push it higher in frequency where it’s even harder to get rid of. And in the original circuit, the current draw is probably too low to induce ringing anyway. If it aint broke don’t fix it.
If you can buy the faster rectifier for less than the cost of the cap and resistor it may be an option for line frequency supplies. If it actually eliminates the ringing and not just push it higher in frequency where it’s even harder to get rid of. And in the original circuit, the current draw is probably too low to induce ringing anyway. If it aint broke don’t fix it.
Hi,
The latter show alot higher leakage currents .... >100ma are possible ... and the upper temp limit will also be lower ... sometimes not even 100°C.
Check the Datasheets.
jauu
Calvin
There are two types of Schottkys, one with higher Vf and quite low Ileak and the other with lower Vf and high Ileak.
The latter show alot higher leakage currents .... >100ma are possible ... and the upper temp limit will also be lower ... sometimes not even 100°C.
Check the Datasheets.
jauu
Calvin
If you really have a bug up your butt about high voltage schottkys, there is always silicon carbide. If you like paying too much for a rectifier, go for it.
People always seem to look for the most expensive or complicated solution, don’t they? $100 says the original circuit in question wasn’t even ringing to begin with.
People always seem to look for the most expensive or complicated solution, don’t they? $100 says the original circuit in question wasn’t even ringing to begin with.
Sure, but the OP's schematic doesn't need anywhere near a 1kV diode - the input is 55VAC.
Therefore, I suggest HER102G or HER103G. Frankly, a 1N4148 would work in that position.
Therefore, I suggest HER102G or HER103G. Frankly, a 1N4148 would work in that position.
A 100 volt schottky would be leaky, so you use a 1200 volt to solve the leakage, right? That kind of thinking is all over the place, as is the idea of using a 1 or 1.5kVA trafo and 60 amp high speed diodes to power LM3886’s putting out all of 40 watts. With an unsolvable ground loop between the two separate PCBs for the two channels.
4148 might have trouble with the turn on surge charging up the cap. Might get unlucky and blow it, unless you put some R in series. I thought HER10x had gone obsolete in favor of the UF400x. I’d probably grab a UF4007 out of the bin on a new build, but if a 1N was in there already and not causing issues I’d just leave it.
4148 might have trouble with the turn on surge charging up the cap. Might get unlucky and blow it, unless you put some R in series. I thought HER10x had gone obsolete in favor of the UF400x. I’d probably grab a UF4007 out of the bin on a new build, but if a 1N was in there already and not causing issues I’d just leave it.