Hey all,
I'm working on a new amp build - sissysit24 😉 Zen recommended a choke. I have a few questions - hoping to sort things out.
I read up on https://www.diyaudio.com/community/threads/how-do-you-calculate-choke-size-in-a-power-supply.159832/ which covers this topic but is closed so started a new one.
Assuming I did the math right on things:
I'm planning to use the W12 from @rhthatcher . In this configuration I would end up needing a total of 4 chokes - 2 rectifiers per channel. I think each rectifier would be passing .5 amp - is that right? ( each channel would draw aprox 1 a, dividing that evenly I think makes sense?)
When I go to look for a choke from Hammond the size / current I came up with above don't make sense to me. I think I'm either off on my calculations or just don't understand things well enough yet! If my understanding above is correct then something like Hammond 193NP would be a sensible choice?
Thoughts? Am I in the right ballpark here or off? If there is a thread / doc that covers this please let me know. I've googled and searched but not exhaustive. Most of what I found related to using chokes in tube amplifiers.
I'm working on a new amp build - sissysit24 😉 Zen recommended a choke. I have a few questions - hoping to sort things out.
I read up on https://www.diyaudio.com/community/threads/how-do-you-calculate-choke-size-in-a-power-supply.159832/ which covers this topic but is closed so started a new one.
Assuming I did the math right on things:
- I think the amp is 25w, need to verify that
- I think the current draw, based on the 24v 25w would be approx 1 amp ish. ( per channel )
- Based on the formula I think I would need a 2H choke
- (24 / .01)/1200 == 2
- I'm not sure I'm doing this right!
I'm planning to use the W12 from @rhthatcher . In this configuration I would end up needing a total of 4 chokes - 2 rectifiers per channel. I think each rectifier would be passing .5 amp - is that right? ( each channel would draw aprox 1 a, dividing that evenly I think makes sense?)
When I go to look for a choke from Hammond the size / current I came up with above don't make sense to me. I think I'm either off on my calculations or just don't understand things well enough yet! If my understanding above is correct then something like Hammond 193NP would be a sensible choice?
Thoughts? Am I in the right ballpark here or off? If there is a thread / doc that covers this please let me know. I've googled and searched but not exhaustive. Most of what I found related to using chokes in tube amplifiers.
Probably the calculation you used is for a choke input power supply. For a CLC filter, you don't need high inductance but want low DC resistance due to the high current flow (reduce voltage drop). A choke in the tens of millihenries range with resistance under 1 ohm would work. Something like a Hammond 156ZJ.
You can easily model a CLC power supply using PSUD2.
You can easily model a CLC power supply using PSUD2.
Thoughts on how to come up with the sizing or is that what PSUD2 is for?
RE PSUD2 - I need to get that installed and working on my windows machine 😉 Do I need to draw in the entire power supply and amp for this or can I get away with just the power supply side?
The rectified AC has ripple, which the C, LC, and/or RC components filter out. You decide how much ripple in the DC that you can live with and design the filters to achieve this.
PSUD2 software can save you lots of calculations. You just need to enter the power supply (transformer, rectifier, filters, load current). PSUD2 will calculate the output of the power supply including graphics showing the voltage vs. time.
PSUD2 software can save you lots of calculations. You just need to enter the power supply (transformer, rectifier, filters, load current). PSUD2 will calculate the output of the power supply including graphics showing the voltage vs. time.
You want a high enough inductance that the resonant frequency is well below 50Hz. Otherwise what you have is not a filter but a resonator. I encountered a PSU a few years ago where 0.3H had been used instead of 3H and the resonance was at 72Hz. Not great.
I'm guessing to do this simulation properly I need to draw in the W12 - at least the capacitor topology.
I looked over the psud2 simulator and the rectifiers they had. Its not clear to me how to add the one I linked earlier. I don't know how to translate the data sheet into the rectifier field values. I'm going to give it a closer look tomorrow.
I looked over the psud2 simulator and the rectifiers they had. Its not clear to me how to add the one I linked earlier. I don't know how to translate the data sheet into the rectifier field values. I'm going to give it a closer look tomorrow.
In terms of the inductor's resonant frequency, a 10mH choke probably has quite low capacitance so the resonant frequency would likely be in the ultrasonic range.
I have used low inductance, high current Hammond inductors in CLC filters in several solid state amps and have not encountered any issues.
Here is an example of a PSUD2 analysis. I don't believe the choice of bridge rectifier is a critical factor for the purpose of examining the filter response.
I have used low inductance, high current Hammond inductors in CLC filters in several solid state amps and have not encountered any issues.
Here is an example of a PSUD2 analysis. I don't believe the choice of bridge rectifier is a critical factor for the purpose of examining the filter response.
Attachments
So without a choke vs with - what is the choke doing just smoothing out the dc a bit in tandem/before the filter caps? How does the lack of or improperly sized choke manifest?
You don't and it makes no difference. I don't know for certain but suspect that rectifier parameters in PSUD2 are there just to generate warnings when exceeded. If for some obscure reason you really need to use realistic models, use a spice program, not PSUD2.
Acquiring some basic electronic knowledge may be a good idea prior to using a simulator.
bilsch, maybe it's better if you follow an existing power supply and amp design, especially if you have little to no experience with design. Design experience can grow from following an existing design and you don't run the high risk of not being aware of important issues that could easily lead you to being disheartened with the outcome, or causing collateral damage along the way. Many people 'cut their teeth' on kit amps, or magazine amps, or nowadays a forum related amp.
C’mon, it’s not that hard. CLC is a pi filter whose resonance is at 1/(2 pi sqrt(LC)) where C is half the value of the actual capacitors, assuming they are equal. Put this frequency a couple of octaves below 100Hz and you have -24dB of 100Hz ripple ref. the original level.
I'd guess the OP has no idea of what general ripple is wanted/needed, and whether that's for idle or full pelt or ..., let alone get to choosing and buying adequately rated parts that may or may not suit an as yet undefined amp. And then there is constructing, and testing and mains safety. It's a long list of aspects to start from scratch with, so it'd be sad if something went wrong on that long journey and the OP gave up in frustration because of a myriad of forum snippets that aren't presented in an understandable manner for the OP.
I am reading that he is asking how to size the choke, and I’m trying to answer that, instead of just waffling about how it’s all too hard.
So thanks for the responses all. I am indeed trying to branch out and learn a little while having fun building another amplifier 😉
The recommendation from Zen was that I should explore a choke in the power supply so thats what I'm trying to do 😉 I have not yet bought said power supply parts or transformer yet though like I mentioned the W12 is maybe what I would end up with? Zen mentioned a dual mono configuration and thats what I have in my m2x - of which the sissysit iirc is a derivative of.
Where I am at the moment is trying to find time to play with the power supply tool and, if I understand what I need to do correctly, model the power supply circuit without the choke. Once I have that I can see what the ripple would be - probably playing with different resistor/capacitor configurations? Then also playing with different choke sizing to reduce the ripple to some threshold - I'll have to re-read the forum a bit for that. I know others have mentioned a few things.
My understanding may be naive - thats part of why I came to the forum 😉 Hoping to learn! I had a similar question on the M2x with the transformer sizing and went down an interesting rabbit hole there! This is just another. I don't want "use X" I want to understand how/why X and learn something.
The recommendation from Zen was that I should explore a choke in the power supply so thats what I'm trying to do 😉 I have not yet bought said power supply parts or transformer yet though like I mentioned the W12 is maybe what I would end up with? Zen mentioned a dual mono configuration and thats what I have in my m2x - of which the sissysit iirc is a derivative of.
Where I am at the moment is trying to find time to play with the power supply tool and, if I understand what I need to do correctly, model the power supply circuit without the choke. Once I have that I can see what the ripple would be - probably playing with different resistor/capacitor configurations? Then also playing with different choke sizing to reduce the ripple to some threshold - I'll have to re-read the forum a bit for that. I know others have mentioned a few things.
My understanding may be naive - thats part of why I came to the forum 😉 Hoping to learn! I had a similar question on the M2x with the transformer sizing and went down an interesting rabbit hole there! This is just another. I don't want "use X" I want to understand how/why X and learn something.
Hi Bilsch,
The Radiotron Designer handbook (4th edition) will give you some formulas page 1182-1185.
Basically, you start by calculating the load resistance RL.
Then divide it by 6 * pi * mains frequency.
In your case, RL = 24 ohms (approx.) => Lc = 24 / (6*pi*50) = 25.5mH. This is the minimum value. In practice, this value is doubled for good measure.
So, you need 50mH.
Now, do not forget that the DC voltage after filter will be about 0.9 * RMS transformer voltage. you need to also take into account the voltage drop in the diodes and the choke.
Check Morgan Jones book "Building Valve Amplifiers" on the subject. and this link perhaps:
The Radiotron Designer handbook (4th edition) will give you some formulas page 1182-1185.
Basically, you start by calculating the load resistance RL.
Then divide it by 6 * pi * mains frequency.
In your case, RL = 24 ohms (approx.) => Lc = 24 / (6*pi*50) = 25.5mH. This is the minimum value. In practice, this value is doubled for good measure.
So, you need 50mH.
Now, do not forget that the DC voltage after filter will be about 0.9 * RMS transformer voltage. you need to also take into account the voltage drop in the diodes and the choke.
Check Morgan Jones book "Building Valve Amplifiers" on the subject. and this link perhaps: