right,
Qg in nc.
iwas wondering how Qg and Drive amp relate. I have not found good info on this.
i isolated my pwm controller and seem to have cured the inerference to the chip.
now i need to do what you are doing and work on the voltage feedback filters to stabilize the less that 45% pulse width regulation.
i wanted to do this without the opto since i have transformer isolated the control section of the supply
have you ever heard of interference to the gate transformers?
Qg in nc.
iwas wondering how Qg and Drive amp relate. I have not found good info on this.
i isolated my pwm controller and seem to have cured the inerference to the chip.
now i need to do what you are doing and work on the voltage feedback filters to stabilize the less that 45% pulse width regulation.
i wanted to do this without the opto since i have transformer isolated the control section of the supply
have you ever heard of interference to the gate transformers?
Well I don't have any exp. with gate trafos, so I can't hepl you there. But a lot of people do, maybe will help you the on thi part
eva et al
well i finally got all of my parts and the holidays are over so. i have rewound an etd-59, with 25 turns of multistrand #28 and 2 turns of .015" foil. just fits on the former.
8strands-1 foil-8 strands-1foil-6 strands, each separated by .006" fiberglass motor insulation
all of the #28 are paralleled and the two foils are parallelled.
i have started putting together a pcb with no ground loops and ground planes on both sides.
i will run this at about 150khz @ 800 Bmax.
i was going to use 247 packages because of the currents, with ferrite beads. should these beads be on each leg???????😕
i thought the 220 packages might be too close for 370 v. every body was tking about creepage and since 0.1" spacing is 2.54mm minus the pad diameters the spacing is 1mm or less....😕 😕
i found that with my breadboard mess that the emi was interfering with the 33025 and creating noise on the osc pin and transfering that to hi freq pulses on the outputs, causing plumes of PLASMA to exit the fets and evaporate other parts and a hole in the pcb.
neat, my friend says he just lets the smoke out, I LET THE FIRE OUT TOO, anyone know cpr?
well i finally got all of my parts and the holidays are over so. i have rewound an etd-59, with 25 turns of multistrand #28 and 2 turns of .015" foil. just fits on the former.
8strands-1 foil-8 strands-1foil-6 strands, each separated by .006" fiberglass motor insulation
all of the #28 are paralleled and the two foils are parallelled.
i have started putting together a pcb with no ground loops and ground planes on both sides.
i will run this at about 150khz @ 800 Bmax.
i was going to use 247 packages because of the currents, with ferrite beads. should these beads be on each leg???????😕
i thought the 220 packages might be too close for 370 v. every body was tking about creepage and since 0.1" spacing is 2.54mm minus the pad diameters the spacing is 1mm or less....😕 😕
i found that with my breadboard mess that the emi was interfering with the 33025 and creating noise on the osc pin and transfering that to hi freq pulses on the outputs, causing plumes of PLASMA to exit the fets and evaporate other parts and a hole in the pcb.
neat, my friend says he just lets the smoke out, I LET THE FIRE OUT TOO, anyone know cpr?
whay would you use ferrite beads on each leg or just one. That is not necessery.And why so high freq, you willl have a lat of emi problems and large skin effect.
The 220 style cases should be fine for 370VDC, just look inside all those PC power supplies... The creepage requirement is between live parts and ground. Spikes large enough to bridge the leads will probably already have blown the transistors. The short from either the blown transistor or arc between the leads will blow the fuse and save most of the other parts in the supply.
transistor cross
hi eva et al,
i am wondering if the stpsa92 would be a suitable replacement for the bc327 for the gate drive turn off?
http://www.st.com/stonline/products/literature/ds/8842.pdf
hi eva et al,
i am wondering if the stpsa92 would be a suitable replacement for the bc327 for the gate drive turn off?
http://www.st.com/stonline/products/literature/ds/8842.pdf
Turn it Off!
Jimbo,
I would go with something a little more robust, since the peak drive capability of the IR211X chips is already +/-2A, putting a turn-off transistor with an I(c) of only 600-700mA would not help things any. Instead, I would go with an MJE243/MJE253 PNP-NPN combo, and do an emitter-follower ckt.
Have ONSemi sample you some. These have an I(c) of 4A, and a V(ce) of 100V. Just my 2 cents' worth.
Steve
EDIT: Or better yet, email me and I will send you some.
Jimbo,
I would go with something a little more robust, since the peak drive capability of the IR211X chips is already +/-2A, putting a turn-off transistor with an I(c) of only 600-700mA would not help things any. Instead, I would go with an MJE243/MJE253 PNP-NPN combo, and do an emitter-follower ckt.
Have ONSemi sample you some. These have an I(c) of 4A, and a V(ce) of 100V. Just my 2 cents' worth.
Steve
EDIT: Or better yet, email me and I will send you some.
Some standard transistors that may come very handy as gate buffers are BD139 and BD140 (or BC639 and BC640), or BC337 and BC327, however there is little point in buffering things such as a +/-2A driver IC (others like IR2112 are weaker and benefit from buffering).
Another nice pair of bipolar transistors for switching and gate buffer applications are ZTX851 and ZTX951, they come into a tiny package yet they have useful gain and bandwidth above 5A. Their drawback is that they are somewhat harder to find (I got mine from RS-Amidata).
Another nice pair of bipolar transistors for switching and gate buffer applications are ZTX851 and ZTX951, they come into a tiny package yet they have useful gain and bandwidth above 5A. Their drawback is that they are somewhat harder to find (I got mine from RS-Amidata).
compensation
i was looking at the compensation loop, if you use an isolated feedback loop then it looks like you need compensation, if you use direct feedback then the compensation in the, say, 33025 for instance, the compensation capacitance should be enough.
is that correct?
i was looking at the compensation loop, if you use an isolated feedback loop then it looks like you need compensation, if you use direct feedback then the compensation in the, say, 33025 for instance, the compensation capacitance should be enough.
is that correct?
well well well.....
so i've been trying to reduce some ringing with no luck, the ringing is on the input to the main trans on the gates, everywhere, so i looked at the dc to the swtiching transistor, and it was on the b+ so i put a 1uf on and it is nearly non existant and has been reduced everywhere else that the snubbers do their job.
the bulk capacitor/doubler bank is 8 inches away connected by #12 wire.
so i gues i will have to rearange this............
so i've been trying to reduce some ringing with no luck, the ringing is on the input to the main trans on the gates, everywhere, so i looked at the dc to the swtiching transistor, and it was on the b+ so i put a 1uf on and it is nearly non existant and has been reduced everywhere else that the snubbers do their job.
the bulk capacitor/doubler bank is 8 inches away connected by #12 wire.
so i gues i will have to rearange this............
voltage compensation loop questions
okay!!!!!!!! yaehhhhh.
i have rewound the transformer instituted eva's gate drive/turnoff circuit and steve's transistor selection.
i found that schottkys work much better than 1n4148's and that the mje253's have much better response thatn tip42's and have better current capabilities than stpsa92's.
the transformer has almost no spikes or major ringing, so the e cores work better with large step-up/down ratios than toroids. the multi-layer interleaved winding technique lets the power be absorbed from primary to secondary better.
i have set up the over current control so that it tripss on the cycle by cycle and resets the slow start cap just fine.
NOW............. the biggy ,,,,,,,,,,,,,,when i get close to the voltage feedback starting to bias the error amp the thing gets unstable and oscillates. soooooooo. i need to make the voltage feedback circuit now.
i have looked at browns book and some other posts on the diy. i am seeing that everyone is using tl431's and optos.
my control is alraedy isolated form the input. WHAT IS THE BEST way to proceed without a opto, or is that the best????????????????
okay!!!!!!!! yaehhhhh.
i have rewound the transformer instituted eva's gate drive/turnoff circuit and steve's transistor selection.
i found that schottkys work much better than 1n4148's and that the mje253's have much better response thatn tip42's and have better current capabilities than stpsa92's.
the transformer has almost no spikes or major ringing, so the e cores work better with large step-up/down ratios than toroids. the multi-layer interleaved winding technique lets the power be absorbed from primary to secondary better.
i have set up the over current control so that it tripss on the cycle by cycle and resets the slow start cap just fine.
NOW............. the biggy ,,,,,,,,,,,,,,when i get close to the voltage feedback starting to bias the error amp the thing gets unstable and oscillates. soooooooo. i need to make the voltage feedback circuit now.
i have looked at browns book and some other posts on the diy. i am seeing that everyone is using tl431's and optos.
my control is alraedy isolated form the input. WHAT IS THE BEST way to proceed without a opto, or is that the best????????????????
output inductor
well things are coming along i think i have figured the non-isolated voltage feedback loop and will try it out in the next day or two.
Now th out put inductor, What is the consequence of having to much inductance?
i think i need about 500nh and i think i have about 30000nh.
power supply, Vin 270 min 320 max, 18 volt out at 50 amps, at 150khz.
out put will vary to 6volts
well things are coming along i think i have figured the non-isolated voltage feedback loop and will try it out in the next day or two.
Now th out put inductor, What is the consequence of having to much inductance?
i think i need about 500nh and i think i have about 30000nh.
power supply, Vin 270 min 320 max, 18 volt out at 50 amps, at 150khz.
out put will vary to 6volts
Well, are non if you don't go too much over. The first thing that comes to my mind is reaction time, when output changes a lot, from light power to high and supply can't detect the change fast, so output voltage falls. But I am NOT an expert at this so I could be way off.
Originally Posted by jamesrnz:
NOW............. the biggy ,,,,,,,,,,,,,,when i get close to the voltage feedback starting to bias the error amp the thing gets unstable and oscillates. soooooooo. i need to make the voltage feedback circuit now.
i have looked at browns book and some other posts on the diy. i am seeing that everyone is using tl431's and optos.
my control is alraedy isolated form the input. WHAT IS THE BEST way to proceed without a opto, or is that the best???????
Jimbo,
Since you are operating off the AC Line, galvanic isolation is a MUST. You could either use a complex transformer sensing scheme, or simplify things using the venerable TL431-opto combo. Since the overwhelming majority of research is done using the TL431, I would recommend going this route.
Good to hear the MJEs arrived OK. Also, got the insulators and showed them to our FD station guru. He likes 'em so we're gonna use 'em
Steve
NOW............. the biggy ,,,,,,,,,,,,,,when i get close to the voltage feedback starting to bias the error amp the thing gets unstable and oscillates. soooooooo. i need to make the voltage feedback circuit now.
i have looked at browns book and some other posts on the diy. i am seeing that everyone is using tl431's and optos.
my control is alraedy isolated form the input. WHAT IS THE BEST way to proceed without a opto, or is that the best???????
Jimbo,
Since you are operating off the AC Line, galvanic isolation is a MUST. You could either use a complex transformer sensing scheme, or simplify things using the venerable TL431-opto combo. Since the overwhelming majority of research is done using the TL431, I would recommend going this route.
Good to hear the MJEs arrived OK. Also, got the insulators and showed them to our FD station guru. He likes 'em so we're gonna use 'em
Steve
voltage feed back loop
well the feedback loop doesnt work!!!!!!
tried to follow app.B in browns book, could not do it too many assumptions.
tried to cross values and variables as well as functions and could not in different parts of the book.
Ffp
G2
A2
Gxo
different and incorrect formulas for the same name.
I am lost.
So after this and several other errors in the book I complained to Amazon, Newnes and, EDN all to no avail............oh well.
Don't get me wrong, I think there is a lot of good information, but to a neophyte in solid state smps too much stuff leads me astray.
so i will try to get the other two books somewhere that I can affford.
So if any one has one that I can buy borrow, I would appreciate it.
Having said that.
i wanted to do a two pole two zero voltage feedback. Suggestions.
OR AM I ALL WET????????
well the feedback loop doesnt work!!!!!!
tried to follow app.B in browns book, could not do it too many assumptions.
tried to cross values and variables as well as functions and could not in different parts of the book.
Ffp
G2
A2
Gxo
different and incorrect formulas for the same name.
I am lost.
So after this and several other errors in the book I complained to Amazon, Newnes and, EDN all to no avail............oh well.
Don't get me wrong, I think there is a lot of good information, but to a neophyte in solid state smps too much stuff leads me astray.
so i will try to get the other two books somewhere that I can affford.
So if any one has one that I can buy borrow, I would appreciate it.
Having said that.
i wanted to do a two pole two zero voltage feedback. Suggestions.
OR AM I ALL WET????????
- Status
- Not open for further replies.