Hi
And output too if you can, but you DO NOT need to do this for smps to work, but it is much faster and safer if you use minimal possible to start with.
Input current does change, I had it limited to just over 3A, and I pushed SG, buffers, and 3 pair or was it 4 of IRFZ44n to over 700kHz, current limiting was doing its job... Everything got burning hot in seconds...so for normal use you will need about 2A
I would aim for ~30kHz freq., then add filters to primary and secondary, you will get good waveforms this way
And output too if you can, but you DO NOT need to do this for smps to work, but it is much faster and safer if you use minimal possible to start with.
Input current does change, I had it limited to just over 3A, and I pushed SG, buffers, and 3 pair or was it 4 of IRFZ44n to over 700kHz, current limiting was doing its job... Everything got burning hot in seconds...so for normal use you will need about 2A
I would aim for ~30kHz freq., then add filters to primary and secondary, you will get good waveforms this way
Hi Luka,
I thing that you right Luka, I need to get a stable power stage prototype before going forward to assembeling compleate version of SMPS.
Of course I`ll use a fuses in case if I`ll get car`s battery, or either electronic protection in case if I`ll bulid mains 240-12V converter.
It`s seems to be that is more simple to operate with low Fsw... I`ll try to be in 30..50KHz range for Fsw. Just anybody can to anwer me how other projects working at 50-80 KHz ?
I thing that you right Luka, I need to get a stable power stage prototype before going forward to assembeling compleate version of SMPS.
Of course I`ll use a fuses in case if I`ll get car`s battery, or either electronic protection in case if I`ll bulid mains 240-12V converter.
It`s seems to be that is more simple to operate with low Fsw... I`ll try to be in 30..50KHz range for Fsw. Just anybody can to anwer me how other projects working at 50-80 KHz ?
Can I to use power supply below, as 12 V sourse for experiments with my SMPS project ?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=180231667182&ssPageName=STRK:MEWA:IT&ih=008
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=180231667182&ssPageName=STRK:MEWA:IT&ih=008
Hi
Why would you need it to operate at that "high" freq.? If you get nice waveforms at 30k, I think you will still have them on 80k, probably less, but still something you can live by
Hi Q
Nice to hear that. I had id at 18k I think it was, but sound from trafo was not loud, ok mybe if there was dead quite in room, but even computor makes more sound
Why would you need it to operate at that "high" freq.? If you get nice waveforms at 30k, I think you will still have them on 80k, probably less, but still something you can live by
Hi Q
Nice to hear that. I had id at 18k I think it was, but sound from trafo was not loud, ok mybe if there was dead quite in room, but even computor makes more sound
Tolik,
(Damn! I has this long-winded response to your questions, and I accidentally hit a button that unrecoverably cleared everything. UGH!)
I'll try to reconstruct what I typed.
First, you could try the eBay supply, but you might want to consider the following before doing anything:
1) Since you would essentially be feeding one SMPS from another, there could be some unpredictable interactions between the two units. This could be mitigated by placing a moderately-sized (>20Ah) battery at the output of the eBay supply. The larger the battery, the lower the effective output impedance of that supply.
2) See if the eBay supply has V(out) adjustability. 12V-only is nice, but to reliably test your SMPS under development, you will want 13.8V (minimum), also to keep the battery described in #1 charged. Adjustability below 12V is nice, 'cuz you can test your unit for regulation at low input voltages.
3) Maybe skip the switcher thingy all-together and go with a nice BIG linear. Sole advantage here is the quietness of the linear. Disadvantages: lack of adjustability, heaviness and heat generation (lack of efficiency), just to name a few.
Second, regarding the adjustable switching frequency, replacing the fixed-value R(t) with an R(t) of 90% original value, and seriesing it with a potentiometer of 20% original value, you will be able to adjust your frequency by +/-10%. In Ham radio, when we are powering our rigs from switchers, if the SMPS makes a noise on any band, adjusting its frequency effectively moves the noise off the band, solving the problem. Same principle for powering AF amps. Especially more so, beacuse you can hear the output from the amp.
As far as range of adjustability, I wouldn't go for much more than +/-20%, 'cuz your transformer turns calculations are based on frequency, and you don't want to change frequency too much and cause problems from either too high or too low of primary turns (based on frequency).
Third, regarding the grounding of the '3525's SYNC pin, this is widely unknown, and I only found out about it in the last month or two when I was reading up on synchronizing multiple PWM chips' oscillators to eliminate beat frequencies. So, please, don't feel dumb or stupid.
(Whew! I think I made it!)
Steve
(Damn! I has this long-winded response to your questions, and I accidentally hit a button that unrecoverably cleared everything. UGH!)
I'll try to reconstruct what I typed.First, you could try the eBay supply, but you might want to consider the following before doing anything:
1) Since you would essentially be feeding one SMPS from another, there could be some unpredictable interactions between the two units. This could be mitigated by placing a moderately-sized (>20Ah) battery at the output of the eBay supply. The larger the battery, the lower the effective output impedance of that supply.
2) See if the eBay supply has V(out) adjustability. 12V-only is nice, but to reliably test your SMPS under development, you will want 13.8V (minimum), also to keep the battery described in #1 charged. Adjustability below 12V is nice, 'cuz you can test your unit for regulation at low input voltages.
3) Maybe skip the switcher thingy all-together and go with a nice BIG linear. Sole advantage here is the quietness of the linear. Disadvantages: lack of adjustability, heaviness and heat generation (lack of efficiency), just to name a few.
Second, regarding the adjustable switching frequency, replacing the fixed-value R(t) with an R(t) of 90% original value, and seriesing it with a potentiometer of 20% original value, you will be able to adjust your frequency by +/-10%. In Ham radio, when we are powering our rigs from switchers, if the SMPS makes a noise on any band, adjusting its frequency effectively moves the noise off the band, solving the problem. Same principle for powering AF amps. Especially more so, beacuse you can hear the output from the amp.
As far as range of adjustability, I wouldn't go for much more than +/-20%, 'cuz your transformer turns calculations are based on frequency, and you don't want to change frequency too much and cause problems from either too high or too low of primary turns (based on frequency).
Third, regarding the grounding of the '3525's SYNC pin, this is widely unknown, and I only found out about it in the last month or two when I was reading up on synchronizing multiple PWM chips' oscillators to eliminate beat frequencies. So, please, don't feel dumb or stupid.
(Whew! I think I made it!)
Steve
Ok guys, now I have 3 good deals: 400W 13.8V 29A SMPS , 400W 12V 33A SMPS, car lead-acid battery. Both SMPS have +/-10% regulation of output voltage. What will be better to you optinyon ?
Steve,
I thinked about linear - it`s expensive($120..150 just trafo only), & it`s a pice of work to build 30..40A regulated module.
Steve,
I thinked about linear - it`s expensive($120..150 just trafo only), & it`s a pice of work to build 30..40A regulated module.
Hi
Switching is way better for you, and my vote goes for 13.8v one(-/+ 10% = 12.42v-15.18v), coz you can use it as charger for real car battery if you are going to be doing some really big tests. I mean having 10.8v is not usefull voltage in car that much, I would like to keep it above 11v + I think you will still get a bit more regulation then 10%, maybe 1-2% more...
Switching is way better for you, and my vote goes for 13.8v one(-/+ 10% = 12.42v-15.18v), coz you can use it as charger for real car battery if you are going to be doing some really big tests. I mean having 10.8v is not usefull voltage in car that much, I would like to keep it above 11v + I think you will still get a bit more regulation then 10%, maybe 1-2% more...
Hi Luka,
Generally I`m also considered this SMPS as best choise, so now I`m going for it. I just wonted to hear you suggestions. As Steve notised - may be I`ll be need additional reduction of effective output impedance from SMPS, so I think to use 2 47000uf 80V 10mR 10 nH Mundorf caps for this purpose.
Generally I`m also considered this SMPS as best choise, so now I`m going for it. I just wonted to hear you suggestions. As Steve notised - may be I`ll be need additional reduction of effective output impedance from SMPS, so I think to use 2 47000uf 80V 10mR 10 nH Mundorf caps for this purpose.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=160229177402&ssPageName=STRK:MEWN:IT&ih=006
400W 13.8V on the way . Just damn delay with my work on this project will be 10-14 days due to shipping time.
Damn I have small budget for this my project, otherwrise I was taking - http://cgi.ebay.com/1000W-13-8V-72A...NameZWD2VQQcmdZViewItemQQ_trksidZp1638Q2em122
400W 13.8V on the way
. Just damn delay with my work on this project will be 10-14 days due to shipping time.Damn I have small budget for this my project, otherwrise I was taking - http://cgi.ebay.com/1000W-13-8V-72A...NameZWD2VQQcmdZViewItemQQ_trksidZp1638Q2em122
Hi
LOL that last one is a beast, I don't think you will/would need it
about caps, wait with them, I think it will be ok as it is, if not, make your own board, say 5x 1000-2200uF 25v caps ||, one side goes to smps, other to your project...+you can add fuse, something like 60A so that caps won't do something you wouldn't like if something goes wrong
LOL that last one is a beast, I don't think you will/would need it
about caps, wait with them, I think it will be ok as it is, if not, make your own board, say 5x 1000-2200uF 25v caps ||, one side goes to smps, other to your project...+you can add fuse, something like 60A so that caps won't do something you wouldn't like if something goes wrong
Hi Luka,
Yeah, 72A is brute
I can to use Nichicon HZ or Sanyo WG 1500uf 16V. I like small motherboard caps due to size & very low ESR/ESL
BTW, Very intresting toroidal cores at very good price-
http://cgi.ebay.com/Toroid-Cores-Po...ihZ005QQcategoryZ7288QQtcZphotoQQcmdZViewItem
http://cgi.ebay.com/Large-Toroid-Co...photoQQcmdZViewItemQQ_trksidZp1742.m153.l1262
Yeah, 72A is brute
I can to use Nichicon HZ or Sanyo WG 1500uf 16V. I like small motherboard caps due to size & very low ESR/ESL
BTW, Very intresting toroidal cores at very good price-
http://cgi.ebay.com/Toroid-Cores-Po...ihZ005QQcategoryZ7288QQtcZphotoQQcmdZViewItem
http://cgi.ebay.com/Large-Toroid-Co...photoQQcmdZViewItemQQ_trksidZp1742.m153.l1262
Hello guys,
At least my new power source arrived, & I can to continue my work I tested the unit, & found that it working just fine 
To my setup I added: UCC27324 - to properly drive MOSFET`s, & 8x 1500uF Nichicon HZ motherboard caps at input bank.
I tried to remoove output inductor to see what will happen - the wave forms on primary become just wroster, so I placed it back.
I`ll try to make bigger inductor & check it out. I`ts intresting, because not a lot of designs are contain output inductors...
Generally I noticed that at 25 KHz SMPS working better than at 50KHz. But I calculated my transformer to work at 50KHz @ 2000 Gauss. From my calculations, I should to make 4.6 turns per each primary. Is I was wrong here ???
Is I should to use vlue of B multiplied by 2 ?
Steve,
Due to winters in my country are warm, I desided to go with SMPS 
At least my new power source arrived, & I can to continue my work
I tested the unit, & found that it working just fine To my setup I added: UCC27324 - to properly drive MOSFET`s, & 8x 1500uF Nichicon HZ motherboard caps at input bank.
I tried to remoove output inductor to see what will happen - the wave forms on primary become just wroster, so I placed it back.
I`ll try to make bigger inductor & check it out. I`ts intresting, because not a lot of designs are contain output inductors...
Generally I noticed that at 25 KHz SMPS working better than at 50KHz. But I calculated my transformer to work at 50KHz @ 2000 Gauss.
From my calculations, I should to make 4.6 turns per each primary. Is I was wrong here ???Is I should to use vlue of B multiplied by 2 ?
Steve,
Due to winters in my country are warm, I desided to go with SMPS
Tolik-
Waveforms look pretty good. Yeah, many manufacturers skip the output inductors to save a few pennies, hoping that the output caps are enough to achieve regulation. Not so, because an inductor is essential in the buck regulator to provide the necessary di/dt for regulation. I say "buck" regulator, because if you remove the transformer and output diode bridge, the remaining components are arranged in a buck topology. In lieu of regulation, they simply run the PWM IC in "wide-open-throttle" mode and let the output voltage vary with output load and input voltage. Cheap - Cheap! (Did I hear a birdie somewhere?!?!)
At 30-50kHz, a 4T + 4T primary is usually OK. For your B value of 2000G, is this your B(max) or B(sat)? B(max) is the number you want to work with for calculating under normal operating, and B(sat) is the never-to-exceed value (this is essentially the point of core saturation).
Steve
Waveforms look pretty good. Yeah, many manufacturers skip the output inductors to save a few pennies, hoping that the output caps are enough to achieve regulation. Not so, because an inductor is essential in the buck regulator to provide the necessary di/dt for regulation. I say "buck" regulator, because if you remove the transformer and output diode bridge, the remaining components are arranged in a buck topology. In lieu of regulation, they simply run the PWM IC in "wide-open-throttle" mode and let the output voltage vary with output load and input voltage. Cheap - Cheap! (Did I hear a birdie somewhere?!?!)
At 30-50kHz, a 4T + 4T primary is usually OK. For your B value of 2000G, is this your B(max) or B(sat)? B(max) is the number you want to work with for calculating under normal operating, and B(sat) is the never-to-exceed value (this is essentially the point of core saturation).
Steve
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.