Hi Okmog,
Spotted this implementation many years ago, the diodes and inductors form an interesting combination for passive components. Current goes through the inductor and choke I suspect at slightly different rates and each has influence on the other and as a whole, cost is negligible suggest you try. All of my thoughts are set to achieve improvement in particularly audio circuits. See if in your circuit noise or audio performance is better.
Cheers / Chris
Spotted this implementation many years ago, the diodes and inductors form an interesting combination for passive components. Current goes through the inductor and choke I suspect at slightly different rates and each has influence on the other and as a whole, cost is negligible suggest you try. All of my thoughts are set to achieve improvement in particularly audio circuits. See if in your circuit noise or audio performance is better.
Cheers / Chris
The current is going through the diode AND the inductor in parallel, so it's not bypassed, it's just an optional route.
since the inductor is there only to fill in gaps in the current acting as a filter, not a primary power conduit
This design has several advantages including transient performance and it allows you to use cheaper low power inductors
The diodes are complete madness: they will mostly cancel the effect of the inductors.
What's the purpose of using big expensive chokes, only to ruin their effect with parallel diodes?
Am I missing something?
What's the purpose of using big expensive chokes, only to ruin their effect with parallel diodes?
Am I missing something?
Wow, a long thread with many wild suggestions.
If you look at the OP, the xfmr has 2 separate secondaries (assuming the OP drew it correctly), so all you have to do is run 2 separate bridges and positive regulators, then siamese the output.
If you look at the OP, the xfmr has 2 separate secondaries (assuming the OP drew it correctly), so all you have to do is run 2 separate bridges and positive regulators, then siamese the output.
I first saw this implementation in a Krell schematic. labeling it as buck conversion or considering the function of a diode as a separate component is like saying that road is wrong because it leads in a unwanted direction. Mach5 is right it does have advantages, give it a try.
Cheers / Chris
Cheers / Chris
Mmm...no, Chris, I don't believe so. What is the DC resistance of your 1H chokes, what is the input voltage to the choke and what is the load current?
Hi dgta
The original schematic was provided in response to a question for voltage reduction, at that time the author required 184v Dc dropped to 100v a quite difficult task. I answered with reducing current followed by zener regulation and LM317.and LM337 for the negative rail. The response then changed to a dual unregulated supply at close to 100v DC. I then supplied schematic showing two transformers for each polarity, each with series connected 35 vac windings being a commonly available voltage, into a bridge rectifier for each polarity, followed by passive filter involving the now argued use of a diode and inductor combination.
So to answer your question the input voltage is 100v, the resistance of a typical inductor of this value is 0.1 ohm load current was as i understand originally quite low, but then changed within the response. Hence an inductor diode combination would not offer any appreciable load itself to the required use.
Cheers / Chris
The original schematic was provided in response to a question for voltage reduction, at that time the author required 184v Dc dropped to 100v a quite difficult task. I answered with reducing current followed by zener regulation and LM317.and LM337 for the negative rail. The response then changed to a dual unregulated supply at close to 100v DC. I then supplied schematic showing two transformers for each polarity, each with series connected 35 vac windings being a commonly available voltage, into a bridge rectifier for each polarity, followed by passive filter involving the now argued use of a diode and inductor combination.
So to answer your question the input voltage is 100v, the resistance of a typical inductor of this value is 0.1 ohm load current was as i understand originally quite low, but then changed within the response. Hence an inductor diode combination would not offer any appreciable load itself to the required use.
Cheers / Chris
Last edited:
OK, let's see what that means. Let's arbitrarily assume a current of 100mA for simplicity. The choke will have an IR drop of .01V or 10 mV. That means the diode has a forward bias of 10mV. Look at the data sheet for 1N4148. At 10mV bias, it passes ZERO current. It's an open circuit, it does nothing. Not until about 650mV at room temp.
Let's now crank up the current and assume that "quite low but then changed" current means the limit of your inductors, 5A. Now we have an IR drop of 500mV across the inductor. Look at the 1N41248 sheet again, ZERO current at 500mV.
One more time. Let's say the inductor can be overloaded and we go to 6A, the rating of your diodes. At 6A load current you have a 600 mV forward bias on the 1N4148. Data sheet, current still ZERO. Open circuit. See where this is going?
It will take 6.5A of load current before the 1N4148 does anything. That exceeds the ratings of both your rectifiers and chokes and requires 650W++ transformers 🙂
Now you tell me how you envision the operation of these diodes, what they do and how they do it.
Let's now crank up the current and assume that "quite low but then changed" current means the limit of your inductors, 5A. Now we have an IR drop of 500mV across the inductor. Look at the 1N41248 sheet again, ZERO current at 500mV.
One more time. Let's say the inductor can be overloaded and we go to 6A, the rating of your diodes. At 6A load current you have a 600 mV forward bias on the 1N4148. Data sheet, current still ZERO. Open circuit. See where this is going?
It will take 6.5A of load current before the 1N4148 does anything. That exceeds the ratings of both your rectifiers and chokes and requires 650W++ transformers 🙂
Now you tell me how you envision the operation of these diodes, what they do and how they do it.
The diode is installed so that it does not conduct under normal conditions. When current to the inductive load is rapidly interrupted, (such as turning on and off ), a large voltage spike may be produced in the reverse direction (as the inductor attempts to keep current flowing in the circuit). Yes I believe its a snubber circuit, and this is the intention Krell had using it.
No, that can't happen either.
First and foremost, you don't use a signal diode as a snubber. 1N4148 can only carry very small currents. Any inductive spike of any size will blow it sky-high.
Yes, in some circuits, when you turn the current off, an inductor can produce a voltage spike at the downstream side. For a diode to snub that spike, it would have to be connected backwards from what you show.
In this circuit, there will most likely be NO spike. See if you can figure out why.
Can you post the Krell schematic so we can see what you saw?
First and foremost, you don't use a signal diode as a snubber. 1N4148 can only carry very small currents. Any inductive spike of any size will blow it sky-high.
Yes, in some circuits, when you turn the current off, an inductor can produce a voltage spike at the downstream side. For a diode to snub that spike, it would have to be connected backwards from what you show.
In this circuit, there will most likely be NO spike. See if you can figure out why.
Can you post the Krell schematic so we can see what you saw?
Hi
The diode quite correctly should be of a higher rating for snubber purpose, my schematic drawing program "oregano" sadly classes all diodes to be 1n4148,
The diagram of the Krell PSU containing this implementation I saw was over 20 years ago, so I cannot provide a schematic. That is how it was implemented and in my use which is in audio i have found it to be an improvement, which suggest frequency.
If its not a snubber can you suggest what it is ? and have you tried it ? what are the characteristics of a parallel diode inductor combination connected this way? Perhaps someone with measuring gear can kindly try it at all frequencies particularly relate to psu and audio frequency. Krell found an advantage, and I think that deserves some interest from the DIY community.
Cheers / Chris
The diode quite correctly should be of a higher rating for snubber purpose, my schematic drawing program "oregano" sadly classes all diodes to be 1n4148,
The diagram of the Krell PSU containing this implementation I saw was over 20 years ago, so I cannot provide a schematic. That is how it was implemented and in my use which is in audio i have found it to be an improvement, which suggest frequency.
If its not a snubber can you suggest what it is ? and have you tried it ? what are the characteristics of a parallel diode inductor combination connected this way? Perhaps someone with measuring gear can kindly try it at all frequencies particularly relate to psu and audio frequency. Krell found an advantage, and I think that deserves some interest from the DIY community.
Cheers / Chris
Chris, with all due respect...
If you don't understand how a circuit works, and you won't, or can't, work your way through it, then you really shouldn't take bits and pieces out of context and claim some sort of magical qualities.
If you saw it 20 years ago and you can't even remember the Krell model, are you sure you memorized the circuit correctly? I looked at several older Krell PSU schematics, like the KSA line, and don't see anything even remotely like this. Just garden variety regulators.
I think it's a bit much to ask someone to build something that doesn't make sense, and test it for you, in order to verify some contradictory claims you make and can't explain.
You say you found an improvement. OK, what exactly was the improvement? Did it sound better?
If you don't understand how a circuit works, and you won't, or can't, work your way through it, then you really shouldn't take bits and pieces out of context and claim some sort of magical qualities.
If you saw it 20 years ago and you can't even remember the Krell model, are you sure you memorized the circuit correctly? I looked at several older Krell PSU schematics, like the KSA line, and don't see anything even remotely like this. Just garden variety regulators.
I think it's a bit much to ask someone to build something that doesn't make sense, and test it for you, in order to verify some contradictory claims you make and can't explain.
You say you found an improvement. OK, what exactly was the improvement? Did it sound better?
Hi
It works, it is a combination of components principally its ability is as an inductor which is fairly obvious. It was also implemented in a Krell schematic. You are right that in a schematic components rely on each other, and its generally not good to take them out of context. Subjective conversation of improvement audio based is of course down to the end user to try, in my opinion it does offer an improvement in audio performance which is probably why Krell used it. Their schematic sadly as I say was last seen but studied by myself over 20 years ago.
Publishing ideas for discussion is what DIY audio is all about... it is as we know very physics based and is reliant on well this works and this does not, I believe there is opportunity despite being surrounded by tradition. Its in everyones interest to find new ideas and ways of implementing L C and R components that do indeed work and if such things sound good and don't exceed voltage or current requirements, thats good for everyone.
Cheers / Chris
It works, it is a combination of components principally its ability is as an inductor which is fairly obvious. It was also implemented in a Krell schematic. You are right that in a schematic components rely on each other, and its generally not good to take them out of context. Subjective conversation of improvement audio based is of course down to the end user to try, in my opinion it does offer an improvement in audio performance which is probably why Krell used it. Their schematic sadly as I say was last seen but studied by myself over 20 years ago.
Publishing ideas for discussion is what DIY audio is all about... it is as we know very physics based and is reliant on well this works and this does not, I believe there is opportunity despite being surrounded by tradition. Its in everyones interest to find new ideas and ways of implementing L C and R components that do indeed work and if such things sound good and don't exceed voltage or current requirements, thats good for everyone.
Cheers / Chris
I judge things on their merits, not reputation.
This scheme is not only worthless, it is detrimental, because it removes any benefit brought by the inductor, and the fact that it might have been used by Krell or anybody else doesn't change that reality:
Implemented that way, and in this context, it is a non-sense.
This scheme is not only worthless, it is detrimental, because it removes any benefit brought by the inductor, and the fact that it might have been used by Krell or anybody else doesn't change that reality:
Implemented that way, and in this context, it is a non-sense.
First trick is to drop half your rectifier diodes (go to halfwave rectification). That may bring you down to usable voltages without need for regulators. Just a thought.
Doc
Doc
love these semi truth table discussions
So if the diode does nothing and is now being described as nonsense, how can it also remove any benefit brought by the inductor, so it is therefore must be doing something to remove a benefit, but a few posts back it was doing nothing at all
Your post invites that it combines with the inductor in a negative way. does it for instance reduce the value of L ? does it add to the value of L ?
Given that most power supplies contain spectra that is unwanted, is a diode in parallel with an inductor really as bad as you suggest.
Have you have tried it ? er... as I thought...
Cheers / Chris
I am not interested in byzantine arguments about the meaning of words.
I said from the beginning the scheme is worthless because it removes the effect of the inductor.
For a perfect inductor and a perfect diode, the diode will completely cancel the effect of the inductor for all conditions.
With real components, having a forward drop and a resistance, the effect will depend on the current drawn and could vary from none at low currents, conduction during part of the cycle at intermediate currents, and finally continuous conduction at higher currents.
If the purpose is to increase the unwanted spectral energy, that's perfectly OK
And you? Did you try it?
But you guessed correctly: I didn't try it, for the same reason I do not need to push a fork into a mains outlet to know it will blow the fuse.
However, there is a good substitute for putting flaky ideas to the test: simulation. It takes 5 minutes and destroys nothing:
Attachments
Last edited:
Yes I have tried it many times, and please lets keep this to a friendly discussion. I recall about the same time as seeing a inductor diode provided this way in a krell schematic, circuit ideas with an inductor paralleled with a capacitor prior to the bridge rectifier, apparently greatly aiding efficiency.
Cheers / Chris
Cheers / Chris
- Status
- Not open for further replies.