| edl |
Hi,
PCB in PDF format. |
|
|
| edl |
| Parts on PCB in PDF format. |
|
|
| luka |
| Have you builded and tested it? |
|
|
| mitwrong |
Hey, What's the argument???
Good Luck! |
|
|
| luka |
Now I see the pictures finally. Seems to be working :D , Great work. What is its max power,even if some changes were needed?
Feedback is doing its job, could I see some output scope picture under heavy load? |
|
|
| edl |
| It performs well with AT/ATX PC-PSU's main and driver transformer. They are cheap and available. |
|
|
| luka |
| How is main trafo winded, turn-ratio? |
|
|
| mitwrong |
| quote: | Originally posted by luka
How is main trafo winded, turn-ratio? |
Luka:
The calculation to wind a pwr transformer is complicated, you have to know the lamination magnetic flux, otherwise take this simple equation.
volt per turn = winding cross section X K.
winding cross section = the centre E of the lamination W X L.
K is constant, depends on the quality of the lamination, usually is 9.00. for general quality. If higher, use 8 or 7. Try to flip a piece of lamination, see how many flips till it breaks, if 3 times, take K=7, if 5-6 flips, take K=8, the rest take K=9.
eg. pri = 220V, sec = 24v2A
Therefore: pri = 220 X 9 = 1980T
sec = 24 X 9 = 216T
220V at 48 watts + 2% eff lost = 50W.
Check from wire gauge table on 220v = 0.218A = ? AWG
24V = 2A = ? AWG
The input + loss make it .22A
Select the wire from the gauge stated, and start your winding.
Good Luck. |
|
|
| luka |
| Are you talking about normal 50Hz transformer? I did want to know for SMPS, where would be pointless to put 2k turns on primary. For normal(50Hz) I know how to calc. coz I did build one myself. |
|
|
| luka |
| And can someone put pictures up again, pls? I does not display them to me, looks like they were deleted. |
|
|
| megajocke |
Mains connected traces and output gnd only some millimeters from each other??? :hot:
(in the corner, near the fuse) |
|
|
| luka |
| so? Air will not conduct up to let say 1.5kV/mm, so there should be enough space. |
|
|
| N-Channel |
Mega- I noticed that, too.
For most countries, it is 4mm. Ever stricter is the 6mm requirement. And, yes luka, even the best fiberglass pc boards can conduct. Maybe not 4mm, but this is for safety and a comfortable error margin.
Other than that, nice work and pics, too!
Steve |
|
|
| luka |
| Hm.. I never look at that problem to closely.Fiberglass board conducting..., but still there must be much more than 330v even for this board to conduct. At 1mm there shouldn't be any worry, I think :) |
|
|
| megajocke |
| But there are spikes on the mains too. Isolation tests are done with 2.5kV AC RMS or something. If an arc would start due to a spike, with mains voltage over only 1mm it would probably continue too... And the PCB carbonizing won't make it better. |
|
|
| luka |
Well having 2.5kV AC spikes on the mains is not normal for me, so...
Use mains filter in order to prevent getting spikes after it.And because you have big bulk cap on primary, that spike would be added to DC bus voltage which wouldn't increase due to very low power of spike, so there is nothing to be woried. But if you are, the how can you explain To-220 FET housing has leads so close next to each other if there would be risk of arcs?;) |
|
|
| N-Channel |
| I didn't want to sound preachy (at least I hope I didn't come across as sounding like that), but the reason I mentioned the 4mm gap is because here across the Pond, the NEC (Nat'l Electric Code) calls for a 4mm gap between the primary and secondary sides of SMPSs. I believe it is the same for many regulatory agencies around the world, too. This distance was chosen, not for the 330VDC but for the multi-kV spikes that megajocke mentions. Anywho, I think it is best to err on the side of caution. :) |
|
|
| luka |
hi
Probably it is over here to, but from my prospective (wich has an error :)) that is to much, since you don't build things for hospital use but personal.You can always put protective layer over tracks and prevent of anything happening. But that is only my opinion. |
|
|
| N-Channel |
| I've always been accused of overbuilding things. :D |
|
|
| luka |
| That can't be bad for ones personal health :D, it is never bad to over-protect things, just in case of any kind of error. |
|
|
| dhanukak |
Hello Gev
ur circui t idea is great.Im going to try this.
May i know CURRENT SENSE TRANSFORMER construction details?such as winding ratios. wire AWG & toride diameter.?
Thanks
Regards
Dhanuka |
|
|
| dhanukak |
Is this circuit has short circuit protection?
Wht will happen when output short?(operation)?
Thanks
Dhanuka |
|
|
| mitwrong |
| quote: | Originally posted by luka
Are you talking about normal 50Hz transformer? I did want to know for SMPS, where would be pointless to put 2k turns on primary. For normal(50Hz) I know how to calc. coz I did build one myself. |
Yes, U are right, it's for 50/60 Hz. If for smps, 30KHz switching freq., the primary standing voltage for 220 rectified DC, the core area 10mm square, the turn should be 0.8 X DC supplied voltage, and the secondary is in terms of winding ratio from primary to secondary.
But one point must bear in mind, for 30KHz freq., the tranny is very easily to arching between coils, good insulation is a must, otherwise arching, sparking will takes place, never reuse copper wire for switching transformer. |
|
|
| dhanukak |
Thanks Mitwrong
Ur idea is very useful to me.If u can pls let me know the DRIVER TRANSFORMER construction details?Becoz Current sense section(windings) included in Driver Transformer in this circuit.(As i understood).If u have any Idea about that pls let me know
Thanks
Regards
DHanuka |
|
|
| mitwrong |
Dhanukak:
The info. I provided was the past experience when I was worked in a SMPS manufacturer. The unit was employed PWM by TL494. The cct was designed in the States, I was the one wants to know everything, That's why I counted the turns inside the ferrite.
The output voltage was controlled via an optic coupler back to TL494 from one of the secondary winding. |
|
|
|
