HI,
I am in the process of making a 400W pfc with a smps for amp...
But being a newbie it is hard for me get the calculation right...... help needed
i am in a bit of trouble........ i have calculated the details for the pfc inductor but it gives me 875microhenry and 238 turn .......... which seems absurd..... for a 400w pfc
here is my calculation....
vin(min)= 90V
vin(max)=265V
pout=400W
fsw = 100khz
Bmax = 0.25T
Ae = 0.97cm square
Core used ETD34 ferrocube...
ipk = 1.414 X pout/ 0.95 X vin(min) = 6.615A
ripple current =15% of ipk = 0.99A
Lcritical = 1.414 Xvin(min){(1-(1.414 X vin(min))/vout} / ripple current X fsw = 875 microhenry
no of turns = lcritical X ipk x 10000 / Bmax X Ae X 1000000 = 238
i have been refering to a pdf by metglass and some other online documents.....
kindly tell me what am i doing wrong .............
regards
sekhar
I am in the process of making a 400W pfc with a smps for amp...
But being a newbie it is hard for me get the calculation right...... help needed
i am in a bit of trouble........ i have calculated the details for the pfc inductor but it gives me 875microhenry and 238 turn .......... which seems absurd..... for a 400w pfc
here is my calculation....
vin(min)= 90V
vin(max)=265V
pout=400W
fsw = 100khz
Bmax = 0.25T
Ae = 0.97cm square
Core used ETD34 ferrocube...
ipk = 1.414 X pout/ 0.95 X vin(min) = 6.615A
ripple current =15% of ipk = 0.99A
Lcritical = 1.414 Xvin(min){(1-(1.414 X vin(min))/vout} / ripple current X fsw = 875 microhenry
no of turns = lcritical X ipk x 10000 / Bmax X Ae X 1000000 = 238
i have been refering to a pdf by metglass and some other online documents.....
kindly tell me what am i doing wrong .............
regards
sekhar
Hi Sekhar,
your numbers do not look absurd.
There might be some chances for optimization, but such optimization should based on real measurements. (I.e. if the circuit does not show overshoot effects in the inductor current, during ramp up and during load steps, you may allow higher flux densities, ie. 300mT. But don't move to this direction without carefull analysis of the real beast)
As a starting point your numbers should work.
If you go for continuous 400W from 90V, the inductor heat might become an issue....
You also may consider a SiC diode, because this system is doing hard switching of high voltage devices with 100kHz (Or check with Eva for her magnetic snubbers - well it's easier to choose a SiC diode....)
your numbers do not look absurd.
There might be some chances for optimization, but such optimization should based on real measurements. (I.e. if the circuit does not show overshoot effects in the inductor current, during ramp up and during load steps, you may allow higher flux densities, ie. 300mT. But don't move to this direction without carefull analysis of the real beast)
As a starting point your numbers should work.
If you go for continuous 400W from 90V, the inductor heat might become an issue....
You also may consider a SiC diode, because this system is doing hard switching of high voltage devices with 100kHz (Or check with Eva for her magnetic snubbers - well it's easier to choose a SiC diode....)
Last edited:
...well, you didn't tell us your output voltage, but for an output voltage slightly above 400V - your calculations would fit.
High number of turns:
Well, high inductance, high current, small Ae... yes, you will need such high number of turns and a large gap.
With respect to the fact that most of the flux in your application will be low frequency and just the ripple is 100kHz, you may consider other core materials like cool Mu, in order to optimize the inductor size.
But don't bother to much. If you need the SMPS for an audio amp, then the average load is much less than the peak load. Thermal design is sufficient if it can handle full power in a rythm of ON/OFF 1sec/2sec.
I get 1mH and 29t...but don't quote me on this
Edit: also, get this Transformer And Inductor Design Handbook, Third Edition
if you can't, I will try to provide you with it
you have everything that you would need to calculate PFC inductor, and that includes gap too!
if you already have core, take some no too big wire, and get L you need with core on
Edit: also, get this Transformer And Inductor Design Handbook, Third Edition
if you can't, I will try to provide you with it
you have everything that you would need to calculate PFC inductor, and that includes gap too!
if you already have core, take some no too big wire, and get L you need with core on
Last edited:
Simply recalculate 'every design' and note the differences, this will bring evidence why different designs end up in different inductors.
During your project you will probably struggle much more with parasitic effects of the layout and the components, rather than with the basic calculation.
-An ETD34 with 1mH (edit:875uH) and 238 turns might suffer from parasitic winding capacitance. ==> Check for winding constructions with small capacitance like multi section bobbins or wild windings.
- Or consider a core with larger Ae. ==> less turns.
- Or consider a higher ripple current ==> lower L ==> Less turns, but more demanding regarding HF-effects inside the winding (i.e. eddy currents due to fringing field and may be also headache with the proximity effect, might require multi stranded wires or HF-litz).
- Reverse recovery charge of the boost diode?
- Gate charge of the Switch?
- Parasitic layout inductances?
- Undesired fast ringing during/after the switching events?
- Safety: Fusing, clearances, creepages, isolation materials, temperatures, flamability, fault conditions..... !! Check for the relevant standards in your country.
-An ETD34 with 1mH (edit:875uH) and 238 turns might suffer from parasitic winding capacitance. ==> Check for winding constructions with small capacitance like multi section bobbins or wild windings.
- Or consider a core with larger Ae. ==> less turns.
- Or consider a higher ripple current ==> lower L ==> Less turns, but more demanding regarding HF-effects inside the winding (i.e. eddy currents due to fringing field and may be also headache with the proximity effect, might require multi stranded wires or HF-litz).
- Reverse recovery charge of the boost diode?
- Gate charge of the Switch?
- Parasitic layout inductances?
- Undesired fast ringing during/after the switching events?
- Safety: Fusing, clearances, creepages, isolation materials, temperatures, flamability, fault conditions..... !! Check for the relevant standards in your country.
Last edited:
I doubt that the etd34 will be sufficient for a PFC 90V/AC==>400V/DC delivering continuous 400W.
But before heading for a bold design, check if you need it really continuously.
***
Edit:
And of course check, if you need 90V-264V.
Luka's hint might push you forward a huge step.
If you calculate it for input voltages around 200V-240V only, then things should relax a lot.
***
Let's hope that more people will jump in this thread.
I cannot talk for Luka, but I will not be able to give you support all the time. Running out of time, so I can give you hints just once and while.
..and it would be a wonder if everything would immediately work alright.
You are entering a difficult and dangerous playground.
It can exciting, but it can also be very frustrating.
In the beginning it usually does not save money vs buying a commercial product.
If it becomes passion, never give up. And always consider safety first, before powering up.
But before heading for a bold design, check if you need it really continuously.
***
Edit:
And of course check, if you need 90V-264V.
Luka's hint might push you forward a huge step.
If you calculate it for input voltages around 200V-240V only, then things should relax a lot.
***
Let's hope that more people will jump in this thread.
I cannot talk for Luka, but I will not be able to give you support all the time. Running out of time, so I can give you hints just once and while.
..and it would be a wonder if everything would immediately work alright.
You are entering a difficult and dangerous playground.
It can exciting, but it can also be very frustrating.
In the beginning it usually does not save money vs buying a commercial product.
If it becomes passion, never give up. And always consider safety first, before powering up.
Last edited:
I do massively disagree regarding the opinion that it would not matter, which topology you use.
But AN6086 is a good paper to be compared vs the UC3854-related papers and study the differences and understand why it does matter which topology and core types+shapes you are going use.
Don't think that you would get everything for free in the Fairchild solution.
(Even if I am loving it, and it has proved to be pretty good up to 1kW at 230V.)
But:
You need two inductors.
You have 100% ripple in the inductor current ==> HF-loss-effects in the winding are a key factor for the real choke design!
But AN6086 is a good paper to be compared vs the UC3854-related papers and study the differences and understand why it does matter which topology and core types+shapes you are going use.
Don't think that you would get everything for free in the Fairchild solution.
(Even if I am loving it, and it has proved to be pretty good up to 1kW at 230V.)
But:
You need two inductors.
You have 100% ripple in the inductor current ==> HF-loss-effects in the winding are a key factor for the real choke design!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.