I've always used tube rectifiers, and am particularly fond of mesh plate AZ1 and AZ11. I've also used hybrid bridges with UF4007 diodes in the negative supply. Preamp HTs are up to around 50mA at 250v. Amps up to 200mA at 300-350v. The first capacitor won't be more than 47uF polypropylene, more like 10uF or less. Supply would be CLCRC. Second cap would be 47uF or 100uF polypropylene.
I'd like to try SIC Schottky diodes in a full wave bridge and I'm looking at something good sounding and above all safe - don't want any explosions. So what's safe for a preamp and amp as above? I've been looking at these:
C4D02120A | Wolfspeed 1200V 10A, SiC Schottky Diode, 2-Pin TO-220 C4D02120A | RS Components Wolfspeed Schottky 1200v, 10A £7.14 5 pack +Vat
I do have a lot of these:
C3D02060F | Wolfspeed 600V 4A, SiC Schottky Diode, 2-Pin TO-220 C3D02060F | RS Components
Wolfspeed Schottky 600v, 4A £8.90 10 pack +Vat
Could these be used as is for a preamp as above, or should I double them up, and if so what resistors and capacitors to use? Is 4A enough for an amp of 200mA?
Basically I want to find out everything possible about using these SIC diodes. I also want to know what sound quality I might expect, if anyone has directly compared tubes and these high voltage SICs.
.
I'd like to try SIC Schottky diodes in a full wave bridge and I'm looking at something good sounding and above all safe - don't want any explosions. So what's safe for a preamp and amp as above? I've been looking at these:
C4D02120A | Wolfspeed 1200V 10A, SiC Schottky Diode, 2-Pin TO-220 C4D02120A | RS Components Wolfspeed Schottky 1200v, 10A £7.14 5 pack +Vat
I do have a lot of these:
C3D02060F | Wolfspeed 600V 4A, SiC Schottky Diode, 2-Pin TO-220 C3D02060F | RS Components
Wolfspeed Schottky 600v, 4A £8.90 10 pack +Vat
Could these be used as is for a preamp as above, or should I double them up, and if so what resistors and capacitors to use? Is 4A enough for an amp of 200mA?
Basically I want to find out everything possible about using these SIC diodes. I also want to know what sound quality I might expect, if anyone has directly compared tubes and these high voltage SICs.
.
Last edited:
Andy, at the risk of “offending the theologically compelled”, but there really isn't a 'sound' to solid-state rectifiers, so long as
The reason valve rectification has such markedly different 'sound' when implemented with various brands-and-types of rectifiers has a whole lot to do with VF measuring anywhere from 15 to over 50 volts, nominal. That's a big drop. And the variability between various rectifier types insures that you WILL develop opinions about 'i like this' or 'i hate that'. Not so with solid state.
Oh, for the religiously resolute, there will likely be resounding rebuttals about how little The Goat knows about anything rectification-and-filtering oriented in power supplies. Whatever! My life's experiences tell otherwise.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
1. They're working right,
2. They're not being run into avalance breakdown and
3. They don't have egregious (“selenium”) VF forward voltage drops
Most solid-state rectifiers of the silicon, silicon carbide and Schottky types have VF between 0.5 to 2.5 volts. On a 350 to 450 volt C supply, that's what, between 0.3% to 0.7%? Or twice that if using FWB configuration?2. They're not being run into avalance breakdown and
3. They don't have egregious (“selenium”) VF forward voltage drops
The reason valve rectification has such markedly different 'sound' when implemented with various brands-and-types of rectifiers has a whole lot to do with VF measuring anywhere from 15 to over 50 volts, nominal. That's a big drop. And the variability between various rectifier types insures that you WILL develop opinions about 'i like this' or 'i hate that'. Not so with solid state.
Oh, for the religiously resolute, there will likely be resounding rebuttals about how little The Goat knows about anything rectification-and-filtering oriented in power supplies. Whatever! My life's experiences tell otherwise.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Hello Goatguy - I have no problem at all with the question of how rectifiers of various kinds influence the sound of amps. You're most welcome to your individual beliefs on this, though if you have anything to report on comparative tests I'd be interested to hear it.
What most concerns me is safety - being well within any limits so I don't have to worry at all. I've seen several reports of currents and PIVs being larger than expected at start-up and I'm no expert at all on these devices. So what I'd really like to know most of all is what are the safe limits in a real-world application of the 600v 4A parts or the 1200v 10A parts. I'd assume the latter are good for an amp of 200mA and 350v HT. I'm not sure about the 600v 4A devices, though.
What most concerns me is safety - being well within any limits so I don't have to worry at all. I've seen several reports of currents and PIVs being larger than expected at start-up and I'm no expert at all on these devices. So what I'd really like to know most of all is what are the safe limits in a real-world application of the 600v 4A parts or the 1200v 10A parts. I'd assume the latter are good for an amp of 200mA and 350v HT. I'm not sure about the 600v 4A devices, though.
Yes, audiowise there is no difference. SIC will give a higher voltage and less sag. This may be audible, instead of the rectifier itself.
I have used a component audio sniffer, allowing me to “listen” to for example diodes. With SIC diodes you will hear a rattling 100 Hz. (and other spurious) and when listening to a zener you will literally hear the zener noise. But this is a different story and has little to do with how your B+ will sound. It’s only DC!
Regards, Gerrit
I have used a component audio sniffer, allowing me to “listen” to for example diodes. With SIC diodes you will hear a rattling 100 Hz. (and other spurious) and when listening to a zener you will literally hear the zener noise. But this is a different story and has little to do with how your B+ will sound. It’s only DC!
Regards, Gerrit
You might want to check out the WS3A020120K. It has a lower Vf, higher power handling, and both diodes are in one package. https://datasheet.lcsc.com/szlcsc/2004090032_CETC-WS3A020120K_C510829.pdf
There's also the STPSC5H12. I think I might try these... https://datasheet.lcsc.com/szlcsc/1810010334_STMicroelectronics-STPSC5H12B-TR1_C191034.pdf
Personally, I just use HER308. It's 100 times less expensive and works well. https://datasheet.lcsc.com/szlcsc/2012311738_GOODWORK-HER308_C908706.pdf
There's also the STPSC5H12. I think I might try these... https://datasheet.lcsc.com/szlcsc/1810010334_STMicroelectronics-STPSC5H12B-TR1_C191034.pdf
Personally, I just use HER308. It's 100 times less expensive and works well. https://datasheet.lcsc.com/szlcsc/2012311738_GOODWORK-HER308_C908706.pdf
Last edited:
That would be too easy
and also I prefer fast or no recovery so I can use them to rectify 75kHz from an SMPS. The higher current rating means you won't destroy it even if you shorted the output (fuse would blow instead) and a lower Vf for those triodes that like every volt.High PIV Schottky diodes, being majority carrier only, do not exhibit a reverse recovery spike, AKA switching noise. They are every bit as quiet as vacuum diodes.
Inductive kick back spikes definitely can destroy SS diodes. However, they are very easily managed. A very high WVDC 0.01 μF cap. or an appropriately rated metal oxide varistor (MOV) in what would be the 1st position of a Π section filter suppresses the spikes, without impacting on critical current behavior.
The only remaining advantage to vacuum rectification is the automatic "soft" start associated with types that contain cathode sleeves. Directly heated vacuum rectifiers start conducting almost as quickly as SS diodes do. A carefully selected negative temperature coefficient (NTC) inrush current limiting thermistor is a good, if imperfect, solution to "softening" the start of both SS and directly heated vacuum rectifiers.
https://www.mouser.co.uk/ProductDetail/Wolfspeed-Cree/CSD01060A/?qs=%2Fha2pyFaduhwlfwI1zQp8rAm0165xyYV4XIv1Eqqde2WhAFwg78ihg==
I always used the CSD01060A unless i needed higher reverse voltage, but for most of my builds i used those.
On big advantage of tube rectification is the built in start up delay. I usually implement some sort of delay on the HV in my amps to give the filaments at head start.
I always used the CSD01060A unless i needed higher reverse voltage, but for most of my builds i used those.
On big advantage of tube rectification is the built in start up delay. I usually implement some sort of delay on the HV in my amps to give the filaments at head start.
Or you can simply put a reverse diode across the choke, like we do on relay coils.
Switching noise originates across the power transformer's secondary winding, and should be dealt with there with a Zobel (RC damping). Vacuum rectifiers have very low capacitance, which minimizes direct coupling to signal circuits, but the most important thing to do is to damp the power transformer's spiky ringing.
YOS,
Chris
YOS,
Chris
Not too sure about that. Certainly if the turn-off conditions can force a pn diode to enter a reverse-recovery condition, then yes the Schottky won't exhibit the transient recovery, but imho those conditions are not typical of B+ power supplies unless a significant level of secondary winding leakage inductance occurs with a very peaky diode current (low winding resistance coupled with very large first filter capacitance and an ss diode turn-off voltage span). The soft turn-off of the vacuum diode is very much related to its ability to increase on-resistance from the time it passes peak current to when voltage drop falls through 0V (rather than any ss diode where the turn-off voltage span is smaller, eg. 1-2V through to 0.3-0.6V).
As it has been said, hard cut-off occurs because solid state diodes have non-zero forward voltage. This kind of hard cut-off is not the same as reverse recovery spike.
Snubbing spikes is only a partial solution. Snubbers are designed for fundamental spike frequency, but the spike is broadband.
Snubbing spikes is only a partial solution. Snubbers are designed for fundamental spike frequency, but the spike is broadband.
They make 1700V diodes now: 1700V SiC Schottky Diode | Wolfspeed.
In my case, I just used 3 UF5408 in series for my choke input supply.
In my case, I just used 3 UF5408 in series for my choke input supply.
SS diodes can show up a transient disturbance at the time of turn-off, whether the diode is pn or schottky. The pn can cause a particular disturbance, due to reverse recovery, if the dI/dt is substantial. For normal mains rectification for B+ supplies the dI/dt is so slow (compared to switchmode situations) that pn reverse recovery won't be a contributor to the disturbance. The disturbance happens due to the change in dI/dt when current hits zero, when there is sufficient leakage inductance to have enough energy to exhibit itself outside of the winding as a disturbance. If the diode junction has sufficient capacitance (charge) at the time that the current gets to zero, then the energy from the disturbance can find a way through the diode's capacitance - a Schottky may be worse than a pn for that aspect.
Wrt to SiC diodes, and as indicated in another thread where SiC was brought up as an option, the current capability of a TO-220 type device is very dependent on heat-sinking - without heatsinking, the part is unlikely to be capable of dissipating more than about 1W, and then the deratings start to become quite restrictive, even for a '10A' part.