SMPS
hmmm
very good question indeed.
Light weight may be?
High peak Currents?
And may be cheaper!
hmmm
very good question indeed.
Light weight may be?
High peak Currents?
And may be cheaper!
Simply will supply you with high currents, So its very easy to drive 1000W amplifier with this SMPS (heavy use) and maybe 1500W amplifier for music usage. also low loads are not a problem, either
Since 60 mv ripple at output, I think this is a rare thing into this SMPS.
This SMPS came up after several protos, and its ROBUST, and its WORKING!
TURNS
11 turns 11 Turns for primary
8 turns 8 Turns for Secondary for +-50VDC
You will need to use litz wire if you can, or strand 0.2mm wires or 0.3mm wires
in total of 1.5 ~ 1.6mm for all together.
Example, 4x 0.4MM = 1.6MM
use that for primary, and for secondary
Hope this helps
11 turns 11 Turns for primary
8 turns 8 Turns for Secondary for +-50VDC
You will need to use litz wire if you can, or strand 0.2mm wires or 0.3mm wires
in total of 1.5 ~ 1.6mm for all together.
Example, 4x 0.4MM = 1.6MM
use that for primary, and for secondary
Hope this helps
I'm sorry, but it looks all wrong to me.
First: where's the power incuctor? All SMPS need an inductor for energy storage and halfbridge transformer isn't one, shouldn't be at least.
The MUR460s from switching nod to rails is a waste of diodes and these 18k resitors to the caps midpoint is a waste of resistors and watts.
2mF to form a midpoint is way too much, especialy if you use another series little cap. Keep in mind that the midpoint is self balancing itself because of secondary rectifier.
Also these 4x1mF capacitors don't decouple switching currents, they conduct them, so you really need another capacitor across the rails to close the fast current loops before the mains rectifier.
First: where's the power incuctor? All SMPS need an inductor for energy storage and halfbridge transformer isn't one, shouldn't be at least.
The MUR460s from switching nod to rails is a waste of diodes and these 18k resitors to the caps midpoint is a waste of resistors and watts.
2mF to form a midpoint is way too much, especialy if you use another series little cap. Keep in mind that the midpoint is self balancing itself because of secondary rectifier.
Also these 4x1mF capacitors don't decouple switching currents, they conduct them, so you really need another capacitor across the rails to close the fast current loops before the mains rectifier.
PCB guides:
The resisitors standing upright the PCB aren't any good here, they are too close to each other's and other components' leads. Have o look at R4 and Q2 gate for example. Put them further away from anything they could touch (they will vibrate with 1000W music with some bass, get close enough to spark and BOOM).
Secondly the gate drive signals are critical for any switching circuits, the upper side gate drive track experiences high dv/dt and they both experience high di/dt, especialy with powerful driver like ir2110. What you have done is laid them very very long and with close proximity to each other, so that both capacitive and inductive interference between them will occur.
mur810 rectifier diodes are on their limit or very close to for 1kW +-50V SMPS.
The resisitors standing upright the PCB aren't any good here, they are too close to each other's and other components' leads. Have o look at R4 and Q2 gate for example. Put them further away from anything they could touch (they will vibrate with 1000W music with some bass, get close enough to spark and BOOM).
Secondly the gate drive signals are critical for any switching circuits, the upper side gate drive track experiences high dv/dt and they both experience high di/dt, especialy with powerful driver like ir2110. What you have done is laid them very very long and with close proximity to each other, so that both capacitive and inductive interference between them will occur.
mur810 rectifier diodes are on their limit or very close to for 1kW +-50V SMPS.
1- MUR460 is not installed in real SMPS, it was there JUST in case some one used an IGBT without diode.
2- 18K resistors in your mind are needed, but I think they are VERY important to bleed caacitors when you disconnect the mains, better than getting shocked by the lethal charge into the capacitor. so I prefer to keep them even if thats wrong, while its Not.
ALL manufacturers using same thing, they cant be wrong.
3- What power inductor you are talking about exactly? if you mean ETD49, It ok
4- Not sure about witch capacitor you are talking about, the 1UF
5- that 1uf across high rails is there in real application
Thanks
If the midpoint is moved a bit in one way for some reason, one secondary pwm pulse will have greater amplitude than the other. Through the secondary rectifier for the higher pulse will flow higher current, discharging the capacitor with higher voltage across it, bringing the balance back. Same should work for transformer flux, with some help of Rdson.
The speed of mur460s and bodies isn't that important for the simple reason: for a good low leakage inductance, well snubbed transformer no current flow there anyway. never.
1- MUR460 is not installed in real SMPS, it was there JUST in case some one used an IGBT without diode.
OK
2- 18K resistors in your mind are needed, but I think they are VERY important to bleed caacitors when you disconnect the mains, better than getting shocked by the lethal charge into the capacitor. so I prefer to keep them even if thats wrong, while its Not.
ALL manufacturers using same thing, they cant be wrong.
Don't know ALL, but I believe MANY use startup resistors with self-powered scheme, so that startup resistors bleed the capacitance without dissipating 2 or 3 watts all the time. Also NO manufacturer allows the customer to touch the guts of the circuit with lethal voltage across. Connected to mains or not.
3- What power inductor you are talking about exactly? if you mean ETD49, It ok
The one that is missing. The transformer isn't the inductor except for some small leakage inductance, which is minimized. You have to store the energy in magnetic field somewhere.
4- Not sure about witch capacitor you are talking about, the 1UF
Where? I'm talking about all of them
5- that 1uf across high rails is there in real application
Good!
Hope you find my comments some helpful.
Regards
Adam
OK
2- 18K resistors in your mind are needed, but I think they are VERY important to bleed caacitors when you disconnect the mains, better than getting shocked by the lethal charge into the capacitor. so I prefer to keep them even if thats wrong, while its Not.
ALL manufacturers using same thing, they cant be wrong.
Don't know ALL, but I believe MANY use startup resistors with self-powered scheme, so that startup resistors bleed the capacitance without dissipating 2 or 3 watts all the time. Also NO manufacturer allows the customer to touch the guts of the circuit with lethal voltage across. Connected to mains or not.
3- What power inductor you are talking about exactly? if you mean ETD49, It ok
The one that is missing. The transformer isn't the inductor except for some small leakage inductance, which is minimized. You have to store the energy in magnetic field somewhere.
4- Not sure about witch capacitor you are talking about, the 1UF
Where? I'm talking about all of them
5- that 1uf across high rails is there in real application
Good!
Hope you find my comments some helpful.
Regards
Adam
Last edited:
Hi,
I admire their efforts to achieve a good smps, perhaps the road is still long.
Meanwhile, 4x0.4 = 1.6 (only in mathematics) is not on the wires.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.