So, given that one had a 40uF PP cap following the FET in a pass regulator with an output of 500V, and one wanted to stay in the SOA, how would one calculate the series resistance necessary between the FET and the filter cap?
See previous linked doc.
Could one presume starting at the intersection of the 10ms and 500V, dropping back to 13V at 30A, then the max current is 30A and the resistance needed would be (500V/30A = ) 16.667Ohms? {ignoring the 13V}
This yields a time constant of 40e-6*16.667 = 0.66ms.
Three time constants would yield 2ms, so the impulse would decay form 30A to 1A in 2ms.
Is this still within SOA?
See previous linked doc.
Could one presume starting at the intersection of the 10ms and 500V, dropping back to 13V at 30A, then the max current is 30A and the resistance needed would be (500V/30A = ) 16.667Ohms? {ignoring the 13V}
This yields a time constant of 40e-6*16.667 = 0.66ms.
Three time constants would yield 2ms, so the impulse would decay form 30A to 1A in 2ms.
Is this still within SOA?
Last edited:
afaik, SOA is calculated on the basis of voltage input minus voltage output in a pass mosfet....
the 500 volt Vds spec must never be exceeded...
surely you will not want to operate the mosfets with 500vdc across its drain and source terminals, substantially less than you can get away with...
the 500 volt Vds spec must never be exceeded...
surely you will not want to operate the mosfets with 500vdc across its drain and source terminals, substantially less than you can get away with...
Not only are you wildly OT, but it's becoming quite routine of you to diss other people's experiences which may be significantly better than your own.
Mosfet discussions have zilch to do with the OP, which is why you should remain a Emeritus bla bla and not have any rights to intervene.
I say POT KETTLE BLACK.
Oh.
For the person who claimed the claimed PSU with 150 000 farads was crap...
It's routine now to be able buy 10 000 farad super capacitors, even off Aliexpress nowadays.
You only need 15 of them. 😡
I have no idea why, you would want to, but that's for somewhere else! 🙄
ie.
You don't know what you don't know,-
ie. including what consitutes high quality recording, & high quality reproduction, vital for a production environment in a major opera house or concert hall.
I work by blind test and consensus, not "opiniated" anything.
I see what plainly and obviously WORKS, and what despite acres of bla bla superior this and that, & forum BS plainly DOES NOT.
I regularly make blind tests on equipment & recordings.
They are particularly revealing, inc showing deep rooted prejudices,arrogance, hearing defects and being blinded by science.
It's particularly interesting to see women (who seem to be almost totally unrepresented here), very often can spot "nasties" in sound which ego tripping men can't see, hear, go into denial, or are too dumb to want to realise.
ie. (There's nothing wrong with valve rectifiers in the right places, idem simple silicon diodes).
The highest capacity of a double layer capacitor available from Digi-key is 6000F. It's 2.5V chassis mount.
That means you need 25 of them at that's only good for 2.5V My tube amps use 300V.
I would need 3000 of them for my amp to have a 150000F capacity in the PSU. Not practical whatsoever. Oh and they want 460$ each... Roughly 1.4 MILLION dollars...
🙄
That means you need 25 of them at that's only good for 2.5V My tube amps use 300V.
I would need 3000 of them for my amp to have a 150000F capacity in the PSU. Not practical whatsoever. Oh and they want 460$ each... Roughly 1.4 MILLION dollars...
🙄
nice to have your facts wrong eh?
Did I say they got it from Digikey?
Did I actually SAY it was anything other than a SOLID STATE amp that had that PSU in it??
Oh, and btw if you really wanted all those totally unneccessary Joules of energy at 450V in your PSU, what would anyone sensible do?
Do it with capacitors??? !!
Anyone would have to be crazy.
What you think they put in the aircraft, eg. Dreamliner?
BATTERIES!
Aircraft are the architype of what ever technology is best, because they have to LIFT their own fuel supply.
so,-
It's completely BANAL to make a powerful battery power supply for 400, 500 or 600V, you just buy a load of electric bike batteries.
As far as I can see, there's loads of really powerful 🙄 inputs into this thread.
Mostly completely crazy, talking about Mosfets, high speed diodes SMPS PSU, everything totally irrelevant and mostly stupid too, except a perfectly quiet PSU which I actually use btw.
Lithium batteries or LIPO if you really want to have some punch.
You can even push it thru an invertor if you want from 12V up.
One simple power pack I have will power my dual 50W per channel valve amp for 30 mins, and the invertor costs about 20 EURO.
duhh!
I power an entire tube system using 12V ATX supplies into boost converters, nine of them actually. I can also run the system from a car battery if I want to. Never bothered.
Since this is a Tubes forum, I assumed it was a tube amplifier you were talking about. Either way, 150000F at any voltage is insane for an amplifier. You also said "you only need 15 of them" which implies a very low B+ in said amplifer.
I was also wrong in needing 3000 of them in my example. Since they are all connected in series instead of parallel, I would need 10s of thousands of them for 150000F@300V I wouldn't want to short that LOL
I used Digikey as an example. If I was buying expensive super caps, I wouldn't buy them on Aliexpress.
Whether or not you think MOSFETS or SMPS etc are "crazy" doesn't matter to me though. I use whatever does the job the best/cheapest way and that never includes wasting time, money, space, and power on vacuum diodes or general "audiophoolery", but it has included ultrafast diodes and/or SMPS and DC-DC boost converters. The schematics I posted earlier in this thread are examples of "perfectly quiet" linear PSUs, too.
Since this is a Tubes forum, I assumed it was a tube amplifier you were talking about. Either way, 150000F at any voltage is insane for an amplifier. You also said "you only need 15 of them" which implies a very low B+ in said amplifer.
I was also wrong in needing 3000 of them in my example. Since they are all connected in series instead of parallel, I would need 10s of thousands of them for 150000F@300V I wouldn't want to short that LOL
I used Digikey as an example. If I was buying expensive super caps, I wouldn't buy them on Aliexpress.
Whether or not you think MOSFETS or SMPS etc are "crazy" doesn't matter to me though. I use whatever does the job the best/cheapest way and that never includes wasting time, money, space, and power on vacuum diodes or general "audiophoolery", but it has included ultrafast diodes and/or SMPS and DC-DC boost converters. The schematics I posted earlier in this thread are examples of "perfectly quiet" linear PSUs, too.
6-volt-heater … for all the railing you're emitting about how so many of the other post's posters are full of beans, I have to say lad, you've outdone yourself in the last 5 posts. Completely.
We were once talking about rectifier sound; naturally, as with any “what does it sound like, and why?” conversation, we geeksters angled off to most-of-the-compass-points of a circle surrounding “the issue”. You, amongst us.
With a kind of inevitability that is vexing, at some point the SMPS friendly DIYers piped in, with their fondness of SMPS drop-in-and-use-it civility. NOT vexing. But what is vexing is that immediately come rebuttals from those-who-have-been-stung-by-RF-noise due to many SMPS designs mind-boggling-lack-of-RF-mitigation. Sad. Yet, given good SMPS, they're pretty awesome abstractions of heavy transformers, endless arguable choices in rectifiers, PI networks, and the like.
Still, the solutions — tho' contentious — keep underpinning a common goal, don't they?
We desire of our power supplies essentially 5 things:
№ 1 - they're quiet (very low ripple)
№ 2 - and offer low demand sag (low impedance)
№ 3 - and they're fast to deliver varying current demand
№ 4 - and shouldn't be squirrelly, stinky, toxic, huge, fragile, …
№ 5 - are basically cost-effective and trouble-free for years of use.
Right?
If we agree on that list, then when we go back to the OP's posted question, it becomes obvious why some of the (early) posters posted what they did.
Well, obvious to me.
If a properly designed and implemented power supply, for the purposes of delivering DC to the amplifier's voltage and current amplifying components is idealized, it must embody the above 5 points. Plus a 6th! … an independence-from-power-line-variance.
That 6th one is important: it allows one's amplifiers, preamplifiers, and all the rest to deliver (hopefully) almost exactly the same results within the ±10% line voltage variation which is given to most nations' standards. In some areas, the variance can be well over ±22%, I have personally measured. (Farm, in Kansas, mid-summer, at the end of a LONG rural 'route', powered by a skinny 3-wire 12,000 V overhead pole, along with about 20 other farms along the way, each with huge silage silos, running 24–7 giant drying fans, or dairy feeders, enormously improbable HVAC units for industrial hog production. –22% sag was normal)
I'm just saying: if substituting different rectifiers results in real and incontrovertible sound quality differences, then there clearly is something fundamentally amiss with the design and/or rendering of the power supply using them.
... Respectfully ...
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
We were once talking about rectifier sound; naturally, as with any “what does it sound like, and why?” conversation, we geeksters angled off to most-of-the-compass-points of a circle surrounding “the issue”. You, amongst us.
With a kind of inevitability that is vexing, at some point the SMPS friendly DIYers piped in, with their fondness of SMPS drop-in-and-use-it civility. NOT vexing. But what is vexing is that immediately come rebuttals from those-who-have-been-stung-by-RF-noise due to many SMPS designs mind-boggling-lack-of-RF-mitigation. Sad. Yet, given good SMPS, they're pretty awesome abstractions of heavy transformers, endless arguable choices in rectifiers, PI networks, and the like.
Still, the solutions — tho' contentious — keep underpinning a common goal, don't they?
We desire of our power supplies essentially 5 things:
№ 1 - they're quiet (very low ripple)
№ 2 - and offer low demand sag (low impedance)
№ 3 - and they're fast to deliver varying current demand
№ 4 - and shouldn't be squirrelly, stinky, toxic, huge, fragile, …
№ 5 - are basically cost-effective and trouble-free for years of use.
Right?
If we agree on that list, then when we go back to the OP's posted question, it becomes obvious why some of the (early) posters posted what they did.
Well, obvious to me.
If a properly designed and implemented power supply, for the purposes of delivering DC to the amplifier's voltage and current amplifying components is idealized, it must embody the above 5 points. Plus a 6th! … an independence-from-power-line-variance.
That 6th one is important: it allows one's amplifiers, preamplifiers, and all the rest to deliver (hopefully) almost exactly the same results within the ±10% line voltage variation which is given to most nations' standards. In some areas, the variance can be well over ±22%, I have personally measured. (Farm, in Kansas, mid-summer, at the end of a LONG rural 'route', powered by a skinny 3-wire 12,000 V overhead pole, along with about 20 other farms along the way, each with huge silage silos, running 24–7 giant drying fans, or dairy feeders, enormously improbable HVAC units for industrial hog production. –22% sag was normal)
I'm just saying: if substituting different rectifiers results in real and incontrovertible sound quality differences, then there clearly is something fundamentally amiss with the design and/or rendering of the power supply using them.
... Respectfully ...
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Last edited:
Yah. 150,000 F at 300 V
E = ½CV²
E = ½ 150,000 × 300²
E = 6,750,000,000 J (joules) … divide by 3.6×10⁶ J/kWh
E = 1,875 kWh.
Which would be almost 3× my monthly household power budget. E = ½ 150,000 × 300²
E = 6,750,000,000 J (joules) … divide by 3.6×10⁶ J/kWh
E = 1,875 kWh.
E = 6.75×10⁹ J … divide by 4.186×10⁹ J/ton TNT
E = 1.6 TONS of TNT
You most certainly wouldn't want to accidentally short that out. It'd not just blow you to smithereens, but pretty much the city block, too. E = 1.6 TONS of TNT
________________________________________
BUT… let's give the '150 kF' bit its own likelihood. Just 15 ea., of the 10 kF, 2.5 V doohickeys one might purchase from AliExpress.
E = ½CV²
E = 0.5 × 150,000 × 2.5²
E = 469,000 J (divide by 3.6×10⁶ J/kWh)
E = 0.13 kWh or about ⅐ kWh.
That's much more reasonable. Dunno why '15' of them is a magic number (I would have thought, for a solid power supply for a silicon-output amplifier one really ought to be shooting for a symmetric + … 0 … – center tapped supply, so that either 12, 14, 16, 18 of the kilofarad wunderkind would be 'the number'), but its still the same math. E = 0.5 × 150,000 × 2.5²
E = 469,000 J (divide by 3.6×10⁶ J/kWh)
E = 0.13 kWh or about ⅐ kWh.
The problem with even ⅐ kWh of power storage is that it takes more than ⅐ of a kWh of line juice to charge-em-up!
Specifically (and it really is a fun exercise for a math-head like us), it takes a minimum of 2× the stored power, when charging thru an ohmic current-limit clamp, be it a big-fat-wirewound resistor, or a heat-sink-and-constant-current MOSFET, BJT or anything else. We can get away with less when a switched inductor is used. And its own independent flyback recovery fast response diode. One can actually get close to 5% waste-to-heat.
Anyway.
The best “medicine” for Big Claims bbb-bb-b-bûllsnot is Math.
LOL
⋅-=≡ GoatGuy ✓ ≡=-⋅
Last edited:
ie. Apart from one small significant thing, the total obsolete character of the power distribution network of one the world's most advanced economies, - the U S of A.
Years ago the USSR (hardly the epitome of modernity it is true) decided to wire all their networks at 230V AC 50hz.
(but lots stayed at 120V then sensibly decided to wire 3 phase into many people's kitchen cookers at sensible voltages...)
Luckily doing that they also wired the whole of Eastern Europe the same way which is quite large,- the only downside they wired the majority of private apartments with aluminium wire, which today presents some disadvantages it may be said... 🙄
(it certainly gives me headaches when moving a hi powered amp from a alloy wired apartment to a pro environment wired with copper and high power stability, stage lights with multi KW of 3 phase etc)
Put a small fine point on it.
What you have in the USA, is the 110V equivalent of those USSR aluminium wired houses then claim that 60hz is smart compared with 50...
That is not my fault, but it may easily explain why western Europe, especially the UK was one of the most advanced audio economies in the world until really quite recently, including the invention of the world's largest production microprocessor (the acorn which became the now ubitiquitous ARM), and of course some of the best valve based rock music amplifier amps in the world..
Even Hendrix wanted them.
What problems you have with power losses, bad supply regulation and a host of other problems is absolutely NOT my fault.
Excuse this ancient nonsense from the past how you like...
They are self inflicted injuries, solved apparently by a forest of so called "hi tech" and loads of string, sellotape and a sea of goobledeegook.
It didn't even save, said power lines from setting off forest fires which levelled whole towns in a weekend...
We might have other worries but we don't have such stuff. 😀😀
(least of all in a professional environment).
Last edited:
Tabarnak!!! Here I am nervous of shorting a 3300uF cap LOL (although in the real world the wire or wrench or whatever would have resistance that would limit the discharge, right?)
If it's smaller than 100uF, I short it to discharge it using a cheap test lead gator clip wire... It acts as a resistance since the wire gets hot when I do it 😀 I think it's like #26 or #28...
Last edited:
again, you are very good at projection, you are aptly describing yourself with your posts...i can tolerate you.....
Last edited:
one more justification for the smps case, my 24v/4 amp smps can constantly supply an ACA with 24v at 2.2+ Amps and not worry about the ac line going from 90 volts to 240 volts ac....how is that?
If somebody is looking for russian diodes besides 6d22s there are many of them.
Here is a sortable list, data and datasheets:
audio-wiki.ru/doku.php/1tubes/table-rectifier
Here is a sortable list, data and datasheets:
audio-wiki.ru/doku.php/1tubes/table-rectifier
Last edited: