MR. MZZJ
Mr. mzzj wrote:
"I wouldn't use that bridge rectifier for SMPS, most likely its slow 50-hz variety and causing heating and EMI problems with higher freq."
Be relax, ok? The bridge rectifier is for input tension 115/230VAC 60Hz. In output will use MUR or UF.
Thanks by comments.
Mr. mzzj wrote:
"I wouldn't use that bridge rectifier for SMPS, most likely its slow 50-hz variety and causing heating and EMI problems with higher freq."
Be relax, ok? The bridge rectifier is for input tension 115/230VAC 60Hz. In output will use MUR or UF.
Thanks by comments.
CHACALPOWERS,
Looks good. What are the values for your output caps? For your feedback section, do you sample the (-) negative output too, or just the positive?
What is the osc/switch freq? Oh, never mind- I just saw it: the Rt/Ct combo (7.5kW & 0.001 mF cap. This gives a clock freq of ~190kHz, right?
I agree w/Luka: it looks simple & straightforward.
Best of luck with the rest of the project.
Steve
Looks good. What are the values for your output caps? For your feedback section, do you sample the (-) negative output too, or just the positive?
What is the osc/switch freq? Oh, never mind- I just saw it: the Rt/Ct combo (7.5kW & 0.001 mF cap. This gives a clock freq of ~190kHz, right?
I agree w/Luka: it looks simple & straightforward.
Best of luck with the rest of the project.
Steve
Gate Drive & other things
Luka and Steve
I would have thought you would recognize the design, there is no need to monitor the negative rail it should track the positive rail.The gate drive is straight forward and works perfect if you pay attention to the number of windings and the way you wind the transformer. C12 and C13 along with the resistor are filter's for the IGBT's. The reason there are no values for the feedback is that they will depend on your design and the calculations are in many references from TI and others and then there will be some cut and try. All the material needed to design and build this supply is in this thread however I think a two switch forward converter is a better design for amps and easier to build for DIY.
chas1
Luka and Steve
I would have thought you would recognize the design, there is no need to monitor the negative rail it should track the positive rail.The gate drive is straight forward and works perfect if you pay attention to the number of windings and the way you wind the transformer. C12 and C13 along with the resistor are filter's for the IGBT's. The reason there are no values for the feedback is that they will depend on your design and the calculations are in many references from TI and others and then there will be some cut and try. All the material needed to design and build this supply is in this thread however I think a two switch forward converter is a better design for amps and easier to build for DIY.
chas1
Hi Chas
About feedback, I sense only positive, as negative should be the same as you said. I didn't had any time yet to try gate trafo, but will, since in full bridge I will need 2x IR (instead I will go for gate trafo), and I want to try everything. But instead of MOS-fets or igbt's I will use standard transistor like MJE 13009 along with circuit for fast turn off of base.
How is you two switch forward converter? Have you stoped, or you complited it allready?
About feedback, I sense only positive, as negative should be the same as you said. I didn't had any time yet to try gate trafo, but will, since in full bridge I will need 2x IR (instead I will go for gate trafo), and I want to try everything. But instead of MOS-fets or igbt's I will use standard transistor like MJE 13009 along with circuit for fast turn off of base.
How is you two switch forward converter? Have you stoped, or you complited it allready?
About the gate trafo have no costs for me. I can´t to buy IR2110 because it costs around US$20,00 in Brazil.
Do you understand me my projects limitations?
Regards
Do you understand me my projects limitations?
Regards
Originally posted by luka: But instead of MOS-fets or igbt's I will use standard transistor like MJE 13009 along with circuit for fast turn off of base.
Luka (et al.),
EVA did a full-bridge supply about a year back. I don't remember the thread, or even if it was earlier in this thread, but she said she got up to 1.8kW with only four MJE13009s. Just FYI.
Regrding the negative-tracking-positive thing, sure it will follow it because it is mutually coupled to the (+) V(out), but I respectfully disagree that it "tracks" it as closely as a symmetric sensing of both (+) & (-) V(out)s does. I realize the mutual coupling concept is used in all AT & ATX pc supplies with good results, but each output is also sensed, albeit, with somewhat less regulation than the +5V bus, but nevertheless, all outputs are sensed as well as mutually coupled. I think the same principle would apply here. I am not discrediting the idea of sensing only the (+) V(out), and I believe it works well, it's just my opinion that bipolar sensing affords a greater degree of control.
🙂
Steve
Luka (et al.),
EVA did a full-bridge supply about a year back. I don't remember the thread, or even if it was earlier in this thread, but she said she got up to 1.8kW with only four MJE13009s. Just FYI.
Regrding the negative-tracking-positive thing, sure it will follow it because it is mutually coupled to the (+) V(out), but I respectfully disagree that it "tracks" it as closely as a symmetric sensing of both (+) & (-) V(out)s does. I realize the mutual coupling concept is used in all AT & ATX pc supplies with good results, but each output is also sensed, albeit, with somewhat less regulation than the +5V bus, but nevertheless, all outputs are sensed as well as mutually coupled. I think the same principle would apply here. I am not discrediting the idea of sensing only the (+) V(out), and I believe it works well, it's just my opinion that bipolar sensing affords a greater degree of control.
🙂
Steve
Hi Steve
I know all about Eva's full bridge, it is actually 36Khz,2KW, two E42/21/20, she told me all about it and that when time will be right she will post it here on diyaudio. She has so much knowlage about this stuff that is mind boggling for me.
I aslo think would be better if sensed from (+) to (-), but sensing (+) and (-) seperate would be ultimate sulution, but not needed for many designs as my is.
I know all about Eva's full bridge, it is actually 36Khz,2KW, two E42/21/20, she told me all about it and that when time will be right she will post it here on diyaudio. She has so much knowlage about this stuff that is mind boggling for me.
I aslo think would be better if sensed from (+) to (-), but sensing (+) and (-) seperate would be ultimate sulution, but not needed for many designs as my is.
Re: Gate Drive & other things
Gate drive dc voltage level in Chacalpower's schematic is ill-defined and even smallest gate-drain leakage current (PCB or igbt internal) will eventually cause explosion. It may work OK for a second, or a hour, or a day but neverheless its a bad desing.
Are you saying that schematic "CHACALPOWERS" presented with that brain-dead gate drive is actually your idea? 😱chas1 said:
Gate drive dc voltage level in Chacalpower's schematic is ill-defined and even smallest gate-drain leakage current (PCB or igbt internal) will eventually cause explosion. It may work OK for a second, or a hour, or a day but neverheless its a bad desing.
luka said:
Sorry for the confusion, I am not suggesting sensing both busses separately, with separate TL431-opto combos, although that would be pretty cool. Rather, I suggest sensing both (+) &(-) busses with the same TL431-opto feedback network. Simply lift the connections of said network that would normally go to secondary-side ground, and tie them to the (-) bus, adjusting component values for the now-doubled voltage being sensed. No additional parts necessary, just different values and slightly different connection. Since primary- and secondary-side grounds are isolated (but AC-coupled), there should be no issue of a ground shift by making these changes. The idea for (+) to(-) sensing came from a DC-DC converter schematic from a two-part article in the 04-89 & 01-90 issues of what was then the Audio Amateur magazine. The article was on a 75W+75W MOSFET car amplifier, authored by Mr. Randy Vikan. There are a couple of good tricks he uses here for the SG3525 chip o make the circuit work. If you do a websearch for it, you might come across the schematic.
Hope this clears things up a bit. Let me know if it does not.
Steve
almost forgot, love your avatar. I use the Motorola one to lament their getting out of the discrete semiconductor business.
Hi
🙂 thanks, made it myself while I was bored. Motorola is not making semiconductor anymore? Since when?
🙂 thanks, made it myself while I was bored. Motorola is not making semiconductor anymore? Since when?
They pretty much do finished products now, and some mprocessors & specialty chips, but no more discretes nor any of the integrated circuits we're all familiar with. That sector was spun off in 2001 as ONSemi.
gate drv (bad design)???
If you can do better post your design; Crown uses this in their Pulse Series (2 X 650), (2 X 1100) check it out and then you can direct your comments to them. A good DIY doesn't reinvent the wheel. I never said the design was mine but included it in my supply of over 1kW which has served me well for over a year. You might also make reference to Topic 6 by Roger Adair of TI (a 300 KHz Current - Mode HB converter) and then design of a FB converter in Electronics Now oct 1995 by Reinhard Metz who happens to be a member of this forum. Crow comes in pie form I think. A word to the wise " if you don't understand what's in front of you don't put your hands on it")
chas1
If you can do better post your design; Crown uses this in their Pulse Series (2 X 650), (2 X 1100) check it out and then you can direct your comments to them. A good DIY doesn't reinvent the wheel. I never said the design was mine but included it in my supply of over 1kW which has served me well for over a year. You might also make reference to Topic 6 by Roger Adair of TI (a 300 KHz Current - Mode HB converter) and then design of a FB converter in Electronics Now oct 1995 by Reinhard Metz who happens to be a member of this forum. Crow comes in pie form I think. A word to the wise " if you don't understand what's in front of you don't put your hands on it")
chas1
monitoring both rails
Steve
Your idea is a good one and if I had to do over I might use it but I use a uP to monitor all faults and shutdown supply if any rail is out of spec's.
chas1
Steve
Your idea is a good one and if I had to do over I might use it but I use a uP to monitor all faults and shutdown supply if any rail is out of spec's.
chas1
Re: gate drv (bad design)???
ARE WE talking about same thing?
http://www.diyaudio.com/forums/attachment.php?s=&postid=1250579&stamp=1183807698
Crown definetely doesn't use gate drive like that in Pulse series, and TI's focus.ti.com/lit/ml/slup083/slup083.pdf paper doesnt show it either. In Chacal's schematic move C12 to the left side of R7 so that it is in series with R6 and its fine.
😡
chas1 said:
ARE WE talking about same thing?
http://www.diyaudio.com/forums/attachment.php?s=&postid=1250579&stamp=1183807698
Crown definetely doesn't use gate drive like that in Pulse series, and TI's focus.ti.com/lit/ml/slup083/slup083.pdf paper doesnt show it either. In Chacal's schematic move C12 to the left side of R7 so that it is in series with R6 and its fine.
😡
Ouch... You definately want to move C12 and C13 out of there. It should not be in series with the FET drive. You will have a BOOM on your hands.