I don't know the exact part number but my PFC module is a "brick" module from astec. basically an all in one unit. just add power. 
input is 100-240VAC, output is 380VDC at 950W at 240V in and 550W at 120V in.
edit: found it in datasheetarchive.com
APA100-103
NOTE: 1333.07KB file size.
input is 100-240VAC, output is 380VDC at 950W at 240V in and 550W at 120V in.
edit: found it in datasheetarchive.com
APA100-103
NOTE: 1333.07KB file size.
Yes, #77 would be fine, at least for stuff below 30Mhz. Just take care of winding capacitances (use single layer).
BTW: If he goes half bridge, then 50 turns would do, and it would be easier to wind. Also, for a regulated converter you don't have to aim for 100mT at maximum duty cycle, you can go for 100mT at normal operating duty cycle instead, which may be substantially lower than 100%. Stronger fux densities would be reached during transient conditions, but since most power materials don't saturate until 300 or 400mT there is no problem with that.
For example, the half bridge transformer could be designed to operate at 75% duty cycle for +-195V input (390V) thus providing some margin to accomodate duty cycle limitations and PFC output ripple and voltage drops due to its inherently slow dynamic response. Note that the crossover frequency of the voltage control loop of a PFC is usually chosen to be between 10Hz and 20Hz in order to achieve low THD and high PF (loop gain at 100Hz has to be in the -36dB range), thus the output capacitors are let alone dealing with dynamic load changes at higher frequencies).
BTW: If he goes half bridge, then 50 turns would do, and it would be easier to wind. Also, for a regulated converter you don't have to aim for 100mT at maximum duty cycle, you can go for 100mT at normal operating duty cycle instead, which may be substantially lower than 100%. Stronger fux densities would be reached during transient conditions, but since most power materials don't saturate until 300 or 400mT there is no problem with that.
For example, the half bridge transformer could be designed to operate at 75% duty cycle for +-195V input (390V) thus providing some margin to accomodate duty cycle limitations and PFC output ripple and voltage drops due to its inherently slow dynamic response. Note that the crossover frequency of the voltage control loop of a PFC is usually chosen to be between 10Hz and 20Hz in order to achieve low THD and high PF (loop gain at 100Hz has to be in the -36dB range), thus the output capacitors are let alone dealing with dynamic load changes at higher frequencies).
Yes, but you have to be careful because if you put them directly in parallel with the output of the PFC they are likely to form a tank circuit together with wiring inductance and internal PFC capacitors, which will produce RF ringing on each switching transient. A small series resistor may tame the ringing.
djQUAN said:I don't know the exact part number but my PFC module is a "brick" module from astec. basically an all in one unit. just add power.
input is 100-240VAC, output is 380VDC at 950W at 240V in and 550W at 120V in.
edit: found it in datasheetarchive.com
APA100-103
NOTE: 1333.07KB file size.
DJ, Looking at the datasheet for this PFC module, it appears that you can run its internal clock as either the master or the slave (clk in & clk out pins) to your PWM stage. Nice unit. ASTEC does make some pretty good stuff.
the frequency is a bit on the high side though. at 1Mhz, a freq divider and integrator could work to get a lower freq triangle wave for the PWM stage?
but there is still work to do. I still haven't tested the unit if it works since I have no idea of the proper way to do so........
so it's still at home acting as a nice looking paperweight.
but there is still work to do. I still haven't tested the unit if it works since I have no idea of the proper way to do so........
so it's still at home acting as a nice looking paperweight.
fearing that I may have remembered it wrong, I had another look at the datasheet and yep, it says:
"clock output (int. clock): 4.5Vp-p @1Mhz +/-0.5%"
so I did remember correctly. I'm thinking about following Eva's advice on going half bridge so that I would need only half the required primary turns. I'm still considering on the caps to use for the other side of the primary used to split the rail at 1/2.
I already have lots of 200V caps over here which would be a shame to waste but fear that two 200V caps in series might be pushing the limits for use on a 380V supply. I'm planning two electros in series both bypassed with a 1uF 400V film cap each with 220k resistors to balance the voltage (which may or may not be necessary???)
"clock output (int. clock): 4.5Vp-p @1Mhz +/-0.5%"
so I did remember correctly. I'm thinking about following Eva's advice on going half bridge so that I would need only half the required primary turns. I'm still considering on the caps to use for the other side of the primary used to split the rail at 1/2.
I already have lots of 200V caps over here which would be a shame to waste but fear that two 200V caps in series might be pushing the limits for use on a 380V supply. I'm planning two electros in series both bypassed with a 1uF 400V film cap each with 220k resistors to balance the voltage (which may or may not be necessary???)
DJ-
Pity on the 200V Caps. Do you have any 250V electrolytics? Two of these in series, paralleled with your two 220kW balancing resistors should be OK. For your coupling cap, is a film cap you mention the same as the coupling cap for the main power transformer in an AT or ATX supply? Can a film cap handle he ripple current?
Pity on the 200V Caps. Do you have any 250V electrolytics? Two of these in series, paralleled with your two 220kW balancing resistors should be OK. For your coupling cap, is a film cap you mention the same as the coupling cap for the main power transformer in an AT or ATX supply? Can a film cap handle he ripple current?
I will search around the house if I can find those 250V caps at about 680uF (yes, they are on the big side) and will use those. been a while since I have seen them.......
I don't know if they are similar in the PC PSU but I just got them bare and thought they would work as nice speaker crossovers. I doubt they would have good ripple ratings so I guess they could go bye bye.
would it be a bad idea to connect the primary directly to the caps' center tap?
as a site note: I have worked on quite a few dc-dc SMPS so now I'm looking for more experience in offline ones. hence the many questions.
I don't know if they are similar in the PC PSU but I just got them bare and thought they would work as nice speaker crossovers. I doubt they would have good ripple ratings so I guess they could go bye bye.
would it be a bad idea to connect the primary directly to the caps' center tap?
as a site note: I have worked on quite a few dc-dc SMPS so now I'm looking for more experience in offline ones.
hence the many questions.
DJ-
Connecting the other end of the primary directly to the capacitor centerpoint can be done, but unless you're using current-mode control, you will run the danger of core saturation.
And, No, your questions are too many or bad ones. Actually, you are asking very good ones.
About the 250V 680mF caps: you want to make sure they are 105 Centigrade, and rated for use in SMPSs.
What kind and type are they? Panasonic FC? Elna? Rubycon? Nichicon? Sprague 80D?
Connecting the other end of the primary directly to the capacitor centerpoint can be done, but unless you're using current-mode control, you will run the danger of core saturation.
And, No, your questions are too many or bad ones. Actually, you are asking very good ones.
About the 250V 680mF caps: you want to make sure they are 105 Centigrade, and rated for use in SMPSs.
What kind and type are they? Panasonic FC? Elna? Rubycon? Nichicon? Sprague 80D?
they are Sprague. the light blue ones. not sure which model though. they were taken off an offline SMPS and I think are 105°C rated. the bad news is, I have no idea where I left them......
there are no current mode IC's available here as far as I know so they will be run in voltage mode with TL494. unless you could recommend an IC that does current mode control and has been around for many years........then I'll try and look for one.
there are no current mode IC's available here as far as I know so they will be run in voltage mode with TL494. unless you could recommend an IC that does current mode control and has been around for many years........then I'll try and look for one.
Current-mode ICs: From the Unitrode division of Texas Instruments there are the: UC1846/2846/3846 (a.k.a. SG1846/2846/3846), UCC3806, UCC3808, and UC3825. From ONSemi, there are the MC33025 & MC34025. Main difference between the 33- and 34- is the temp range: -40C to 125C -v- 0C to 70C.
Both the TL494 and the SG3524 can be set up for current-limiting, but they are NOT true current-mode, which is cycle-by-cycle current limiting, based in inductor current on the primary side.
The UC1846, '2846 , & '3846 have been around for almost 20 years.
Both the TL494 and the SG3524 can be set up for current-limiting, but they are NOT true current-mode, which is cycle-by-cycle current limiting, based in inductor current on the primary side.
The UC1846, '2846 , & '3846 have been around for almost 20 years.
TOROIDS WINDING
Hi Eva and smps design friend
I have one 1.5 inches out. diam. toroidal transformer and other 2 inches out. diam. toroidal transf. Tell me: in a half-bridge mode and with 127VAC, how many turns is necessary for primary winding? (1.5 and 2 in. toroids). My TL494 is running in 35-40KHz, I think. 6K8 resistor (Rt), 2n2 cap. (Ct), is it correct?
Please help me.
Thank you very much
Hi Eva and smps design friend
I have one 1.5 inches out. diam. toroidal transformer and other 2 inches out. diam. toroidal transf. Tell me: in a half-bridge mode and with 127VAC, how many turns is necessary for primary winding? (1.5 and 2 in. toroids). My TL494 is running in 35-40KHz, I think. 6K8 resistor (Rt), 2n2 cap. (Ct), is it correct?
Please help me.
Thank you very much
Re: TOROIDS WINDING
You need to know cross-sectional area of your cores. And core material, ie ferrite like 3C90 or iron-powder core or molypermalloy-powder core or.. or ...
PLEASE stay away from line-operated power supplies as you don't seem to have required knowledge to do it safely. Do you know how to debug and construct line-operated smps? do you have safety isolation transformer? have a scope?
sachserjuris said:
Based on the information you have given I would say between 1 and 100000turns.sachserjuris said:
You need to know cross-sectional area of your cores. And core material, ie ferrite like 3C90 or iron-powder core or molypermalloy-powder core or.. or ...
PLEASE stay away from line-operated power supplies as you don't seem to have required knowledge to do it safely. Do you know how to debug and construct line-operated smps? do you have safety isolation transformer? have a scope?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.