trobbins said:
DIY design is a balance of practicality imho. We are all engineers on a diy forum.
I think it is a lot riskier for diyers to add parallel capacitors and resistors to UF4007 when applying diodes in series to add some margin - such as with V-0-V windings that exceed 250-300Vac. For that example, 2 diodes in series from the same batch would be a reasonable up to circa 400Vac, and then going to three in series. The risk with parallel resistors is that a diyer is unlikely to know a parts voltage rating, unless they are particularly careful with their parts stocking practise, and then they would get a rude shock to realise they may need to series up a few resistors to achieve a nominal 1kV rating, or realise that the total room for the rectifier starts to balloon out. Somewhat similarly for capacitors, a diyer is unlikely to have available a parts datasheet to check what Vac can be applied, even if the part has 1kV printed on it.
If you want to learn more, please visit our website rapid 1000a rectifier.
Click to expand...
Ahhh, we finally get a well-reasoned and cogent electrical engineering response which clearly and succinctly identifies why you do not favor the load-balancing components. Yay! Engineering!
I completely agree with you that selecting such resistors (steady-state) and capacitors (transients) requires understanding of the components and the ability to do some basic math. No argument there. Diode strings are like any other aspect of electronic design: if you can't stand the heat, don't go into the kitchen. (I routinely see design practices and safety practices discussed on AK which give me the willies.) I don't think the calculations are difficult, but, as you noted, engineers may have greater facility with math. Which is why generic and unqualified recommendations like, "just use a string" are not viable solutions.
The conclusion we should draw from your analysis is simple and straightforward: one should NEVER put diodes in a series string if one does not know how to properly do it. No argument there. But this is really a non-issue. No good reason exists to string diodes together for a tube amplifier when one can easily purchase diodes at the proper voltage rating for a reasonable price. (Nobody here is building Tesla coil supplies. Not with that level of engineering skills, I hope. Such efforts will likely result in a funeral, "death by misadventure", plus an ad on Craigslist for a "Estate Sale: slightly used Tesla coil, open to offers, sold As-Is".)
Amateur stringing of diodes is risky. Will it fail? Maybe, probably not. Is it a good idea? No. Most decidedly not. Deliberately or ignorantly creating situations where components can fail is a poor design practice. While exceeding capacitor voltage ratings (typically by failing to account for RMS or de-rating for heating) is a similar bad practice, most get away with it, until they don't. The manufacturers warn that placing components in series (diodes and capacitors) results in situations where load-sharing imbalances will exceed voltage ratings. If one component fails the next-weakest typically fails and then the entire block fails.
Components fail and people should understand why. Otherwise it is a great mystery.
I don't have any problem with your response, except for the conclusion that it is more or less ok. I agree to the extent that it is a tricky and risky issue, and should be avoided. I don't agree with the conclusion that omitting safety components of minimal expense is better than including them.
But it would have been better to say that up front as it is an engineering argument, instead of attacking my statements as inaccurate, or understanding as lacking, which they are not.
trobbins said:
Wrt to Qrr, are you referencing power supplies for solid state amps? This is a Tube Audio sub-forum. It is almost impossible for a typical diyer to have a B+ power supply that will exceed the current rating of a 1N4007, even with silly levels of capacitance and ss only rectifiers.
Click to expand...
I know this is for a tube amplifier, but the statements about diode strings were not qualified as only for this purpose. Even there I disagree, but the risk is lesser.
The issue is, again, conduction angle. It explains so much about power supplies that we need to worship more at that alter. I remember the day I understood Schade's chart about conduction angle and peak vs. average current. The flash of understanding was for why tube rectifier diodes have such low input capacitance limits. It finally all made sense!
The practice of using large capacitor banks can briefly exceed diode current. That gradually damages the junction. Given the hours of operation it likely will not fail. But that doesn't mean the practice is a good one. Lots of consumer equipment is poorly designed. Ever see the teardowns of brick chargers and the corners cut? I will not purchase generics for this reason. I don't need my gear fried, or a house fire, when one blows. Risk is small, but it is avoidable.
The charging current, BTW, is not a sinusoid; it is more akin to a narrow sawtooth or low duty-cycle squarewave. This complicates back-of-the-envelope analysis.
The big issues in selecting a diode are the voltage and current ratings:
(1) Turn-On: uncharged filter capacitors draw high current as a spike. Repeated overloads can damage a junction through overheating, particularly in smaller packages.
(2) Operating with Small Conduction Angle: A large capacitor filter bank reverse biases the diode for much of the charging cycle, leading to a very narrow region for charging of the capacitors. This pulls amperes for a few milliSeconds. Even when the diode is rated for the current, this causes the transformer to ring. So does the Qrr. People report that diode Qrr affects tube gear. Do I really need to find those anecdotal reports?
(3) Turn-Off: The collapsing inductance in the transformer spikes the voltage far past the normal peaks. Flyback diodes can eliminate this issue, of course, but most power supplies do not use them. (Ignorance, cost saving, or both.)
The company is the world’s best cabinet huawei supplier. We are your one-stop shop for all needs. Our staff are highly-specialized and will help you find the product you need.
Diode selection matters, and the voltage ratings can be exceeded.
Before people start diode rolling (i.e. using series strings or parallel configurations) the ramifications should be understood. Otherwise a safe rule of thumb would be to use a single diode, not a stack, of, what, back of the envelope, at least, say, five times peak voltage and average current. Sound about right? Remember, case (3) may be very high without flyback diodes.
Engineering can deliver a rule of thumb for safety factors so people who can't do math or select components won't have to do either.
Ahhh, we finally get a well-reasoned and cogent electrical engineering response which clearly and succinctly identifies why you do not favor the load-balancing components. Yay!I completely agree with you that selecting such resistors (steady-state) and capacitors (transients) requires understanding of the components and the ability to do some basic math. No argument there. Diode strings are like any other aspect of electronic design: if you can't stand the heat, don't go into the kitchen. (I routinely see design practices and safety practices discussed on AK which give me the willies.) I don't think the calculations are difficult, but, as you noted, engineers may have greater facility with math. Which is why generic and unqualified recommendations like, "just use a string" are not viable solutions.The conclusion we should draw from your analysis is simple and straightforward:. No argument there. But this is really a non-issue. No good reason exists to string diodes together for a tube amplifier when one can easily purchase diodes at the proper voltage rating for a reasonable price. (Nobody here is building Tesla coil supplies. Not with that level of engineering skills, I hope. Such efforts will likely result in a funeral, "death by misadventure", plus an ad on Craigslist for a "Estate Sale: slightly used Tesla coil, open to offers, sold As-Is".)Amateur stringing of diodes is risky. Will it fail? Maybe, probably not. Is it a good idea?Most decidedly. Deliberately or ignorantly creating situations where components can fail is a poor design practice. While exceeding capacitor voltage ratings (typically by failing to account for RMS or de-rating for heating) is a similar bad practice, most get away with it, until they don't. The manufacturers warn that placing components in series (diodes and capacitors) results in situations where load-sharing imbalances will exceed voltage ratings. If one component fails the next-weakest typically fails and then the entire block fails.Components fail and people should understand why. Otherwise it is a great mystery.I don't have any problem with your response, except for the conclusion that it is more or less ok. I agree to the extent that it is a tricky and risky issue, and should be avoided. I don't agree with the conclusion that omitting safety components of minimal expense is better than including them.But it would have been better to say that up front as it is an engineering argument, instead of attacking my statements as inaccurate, or understanding as lacking, which they are not.I know this is for a tube amplifier, but the statements about diode strings were not qualified as only for this purpose. Even there I disagree, but the risk is lesser.The issue is, again, conduction angle. It explains so much about power supplies that we need to worship more at that alter. I remember the day I understood Schade's chart about conduction angle and peak vs. average current. The flash of understanding was for why tube rectifier diodes have such low input capacitance limits. It finally all made sense!The practice of using large capacitor banks can briefly exceed diode current. That gradually damages the junction. Given the hours of operation it likely will not fail. But that doesn't mean the practice is a good one. Lots of consumer equipment is poorly designed. Ever see the teardowns of brick chargers and the corners cut? I will not purchase generics for this reason. I don't need my gear fried, or a house fire, when one blows. Risk is small, but it is avoidable.The charging current, BTW, is not a sinusoid; it is more akin to a narrow sawtooth or low duty-cycle squarewave. This complicates back-of-the-envelope analysis.The big issues in selecting a diode are the voltage and current ratings:Diode selection matters, and the voltage ratings can be exceeded.Before people start diode rolling (i.e. using series strings or parallel configurations) the ramifications should be understood. Otherwise a safe rule of thumb would be to use a single diode, not a stack, of, what, back of the envelope, at least, say, five times peak voltage and average current. Sound about right? Remember, case (3) may be very high without flyback diodes.Engineering can deliver a rule of thumb for safety factors so people who can't do math or select components won't have to do either.
I'm sorry, but I have to disagree with this, at least for the benefit of other readers.
The problem with MV rectifiers is that if one accidentally breaks while hot, no matter how rarely that may happen, the consequences are much greater than if I broke just any old tube in my collection (external anode ceramic tubes might be an exception, if not nearly as problematic...)
The issue is long term exposure to Mercury vapors, particularly for young children. In practice there would be no way to clean up after a MV rectifier breaks. You would
literally have to discard all the loose items in the room, and physically remove the outhermost layer of walls, ceilings and floor. The microscopic droplets of condensed Mercury from a leak of Mercury vapor will not be cleaned off by any means, short of physical removal of the substrate (floorboards, furniture, paint, wallpaper, etc.)
In reality the contamination may spread to the whole house, which pretty much means it could be easier to just crush the whole thing, contents included, and start anew. Hardly the issue if you broke the odd EL34 in operation...
The first time you break a hot MVR, you are screwed. Simple as that, and people need to be made aware of this. Pardon my French.
The reference to the haphazard treatment of Mercury spills in the past is not a valid excuse. I can point to an incident with another hazardous chemical, which took place in a stairwell at the Institute for Physics and Astronomy, at the University of Aarhus, Denmark, in the late sixties. Already then the scientists knew they had a long term problem, even if the acute danger was virtually zero. Eventually they realized the only reasonable thing to do was to break out the hammers and chisels, and physically remove the outermost layer of brickwork, ceramic floor tiles and replace the steel railings. So they did.
For those, who may not be aware: Long term exposure to Mercury vapors in the environment plus young children equals dumb teenagers with poor motor skills.
The real issue here are all the superfluous warnings we find on just about any item we can buy or acquire. They make it difficult for people to differentiate the real dangers among the noise from the Nanny state.
Finally, why don't you just use 3B28s instead? Same electric properties if you like the sonics of MV rectifiers, with none of the potential problems. There is just no excuse for recommending the use of MV rectifiers.
They need to go extinct.
I'm sorry, but I have to disagree with this, at least for the benefit of other readers.The problem with MV rectifiers is that if one accidentally breaks while hot, no matter how rarely that may happen, the consequences are much greater than if I broke just any old tube in my collection (external anode ceramic tubes might be an exception, if not nearly as problematic...)The issue is long term exposure to Mercury vapors, particularly for young children. In practice there would be no way to clean up after a MV rectifier breaks. You wouldhave to discard all the loose items in the room, and physically remove the outhermost layer of walls, ceilings and floor. The microscopic droplets of condensed Mercury from a leak of Mercury vapor will not be cleaned off by any means, short of physical removal of the substrate (floorboards, furniture, paint, wallpaper, etc.)In reality the contamination may spread to the whole house, which pretty much means it could be easier to just crush the whole thing, contents included, and start anew. Hardly the issue if you broke the odd EL34 in operation...The first time you break a hot MVR, you are screwed. Simple as that, and people need to be made aware of this. Pardon my French.The reference to the haphazard treatment of Mercury spills in the past is not a valid excuse. I can point to an incident with another hazardous chemical, which took place in a stairwell at the Institute for Physics and Astronomy, at the University of Aarhus, Denmark, in the late sixties. Already then the scientists knew they had a long term problem, even if the acute danger was virtually zero. Eventually they realized the only reasonable thing to do was to break out the hammers and chisels, and physically remove the outermost layer of brickwork, ceramic floor tiles and replace the steel railings. So they did.For those, who may not be aware: Long term exposure to Mercury vapors in the environment plus young children equals dumb teenagers with poor motor skills.The real issue here are all the superfluous warnings we find on just about any item we can buy or acquire. They make it difficult for people to differentiate the real dangers among the noise from the Nanny state.Finally, why don't you just use 3B28s instead? Same electric properties if you like the sonics of MV rectifiers, with none of the potential problems. There is just no excuse for recommending the use of MV rectifiers.They need to go extinct.
If you are looking for more details, kindly visit eltek default password.