Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:> Okay, thanks all--I think I've got it figured out. I'm already using > a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off > the free AC and then use regulated cap multipliers to make a nice > quiet +-40ish volts. > > Cartoon attached. > > Cheers > > Phil HobbsVery nice. You got rid of the horrible ringing at the drain on most of the other versions posted here. That is very nice. I was wondering why you want to dump 2.25 watts into R2, the 100 ohm going from -15V to ground. I couldn't see why it was needed, so I removed it. Circuit works fine. There may be some loading effect on either or both supplies that require it, but as currently shown, the reason is not apparent. Similarly, D6 which was in parallel with R2 doesn't seem to be needed either. Removed it also. Circuit works fine. I think the reason R2 is not needed is because whatever charge is introduced into C4 via D5 is removed on the next half cycle via D4 and dumped into R5. So the load presented by R5 is all that is needed to keep -15V on C4. Good guess? I really have to congratulate you. Yours is the best circuit I've seen so far, especially for eliminating noise from the ringing. You truly have a gift for design that most others lack. Thanks, JK
Flyback vs half-bridge
Started by ●July 16, 2013
Reply by ●July 18, 20132013-07-18
Reply by ●July 18, 20132013-07-18
John K <spam@me.not> wrote:> There may be some loading effect on either or both supplies that require > it, but as currently shown, the reason is not apparent.OK, I see it now. I let the analysis run further out, and with R2 removed the -15V tends to drift down towards -20V. So that part depends on the balance between the load on the -15V and -45V supplies. However, putting R2 back in the circuit and increasing it to 1k seems to solve the problem. An alternative might be to increase the load on the -45V supply, or simply keep the -15V heavily loaded. Altogether not bad at all. You have +/-15v and +/-45V supplies with a remarkably small bunch of parts. And no ringing! Thanks, JK
Reply by ●July 18, 20132013-07-18
On 07/17/2013 11:01 PM, John K wrote:> Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: > >> Okay, thanks all--I think I've got it figured out. I'm already using >> a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off >> the free AC and then use regulated cap multipliers to make a nice >> quiet +-40ish volts. >> >> Cartoon attached. >> >> Cheers >> >> Phil Hobbs > > Very nice. You got rid of the horrible ringing at the drain on most of > the other versions posted here. That is very nice. > > I was wondering why you want to dump 2.25 watts into R2, the 100 ohm > going from -15V to ground. >That's the load, i.e. it represents the circuit that needs the -15V.> I couldn't see why it was needed, so I removed it. Circuit works fine. > > There may be some loading effect on either or both supplies that require > it, but as currently shown, the reason is not apparent. > > Similarly, D6 which was in parallel with R2 doesn't seem to be needed > either. Removed it also. Circuit works fine.That was there to get rid of a weird polarity reversal at power-on. The power-on transient isn't representative of what the real circuit will do--the chip has built-in current limiting, whereas this version goes up to like 7A on power-up--but putting +3V on a -15V rail isn't usually considered too healthy for the rest of the circuitry.> > I think the reason R2 is not needed is because whatever charge is > introduced into C4 via D5 is removed on the next half cycle via D4 and > dumped into R5. So the load presented by R5 is all that is needed to > keep -15V on C4. Good guess? > > I really have to congratulate you. Yours is the best circuit I've seen > so far, especially for eliminating noise from the ringing. You truly > have a gift for design that most others lack.Thanks. Blind luck and bloody ignorance this time round though! Cheers Phil Hobbs>-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●July 18, 20132013-07-18
On Wednesday, July 17, 2013 7:56:00 PM UTC-4, Bill Sloman wrote:> On Thursday, 18 July 2013 01:32:53 UTC+10, George Herold wrote: >=20 > > On Wednesday, July 17, 2013 10:14:40 AM UTC-4, Joerg wrote:=20 >=20 > > > George Herold wrote: >=20 > > > > On Tuesday, July 16, 2013 6:12:17 PM UTC-4, Phil Hobbs wrote:=20 >=20 > > > >> On 7/16/2013 3:53 PM, Joerg wrote: >=20 > > > >>> Phil Hobbs wrote: >=20 > > > >=20 >=20 > > > > <snip original question><and more snipping>> > > One of the issues with flybacks and gapped cores is that the gap does=n't=20>=20 > > > know where it is supposed to end at its sides. It extends into the >=20 > > > surroundings and that can result in egg in the face at the EMC lab. >=20 >=20 >=20 > In RM and pot cores manufactured with a build-in air gap, only the centra=l face is relieved - the outside layer of ferrite is continuous, which does= help the shielding.>=20<snipping stuff>> > Say are there any good books about transformers/ inductors/ magnetic=20 >=20 > > materials? >=20 > >=20 >=20 > > The subject seems to go from the trivial freshman physics transformer. =To=20>=20 > > the "full hair ball" real world situation, with nothing in between. >=20 >=20 >=20 > There are books, and E.C. Snelling of Mullard, which got taken over by Ph=ilips, wrote several of them. They were horrible - Snelling produced the mo= st confusing and misleading technical expositions that I ever came across.>=20 >=20 >=20 > When I ran into the Siemens Magnetic Materials data book in the late 1970='s, it was a revelation - the application notes were clear, concise, well o= rganised and informative. The late great Tony Williams shared my enthusiasm= and was known to forward several mega-bytes of .pdf files to interested pa= rties.>=20 >=20 >=20 > When Siemens became EPCOS, and the magnetic business was eventually sold =off to TDK the situation got a bit confused, and I've lost touch with the r= elevant web-sites.>=20 >=20 >=20 > If Winfield Hill ever does get around to releasing the third edition of "=The Art of Electronics" it may include a decent section on magnetics - I ce= rtainly put in a personal plea for one a few years ago.>=20 >=20 >=20 > At the time I laid a fair bit of emphasis on the transformer equation >=20 >=20 >=20 > V1 =3D L1. dI1/dt + M. dI2/dt >=20 >=20 >=20 > V2 =3D M. dI1/dt + L2. dI2/dt >=20 >=20 >=20 > where M =3D k.(L1.L2)^0.5=20 >=20 >=20 >=20 > and k is around 0.99 for a well-coupled transformer. L1 and L2 are the in=ductances of the coupled windings and M is the mutual inductance of the two= windings (in Henries).>=20 >=20 >=20 > Snelling never seems to have used it, and it came as a revelation to me w=hen I found it in the Siemens application notes.>=20 >=20 >=20 > For advanced stuff, look for the John Chan model of hysteritic inductors =in the LTSpice help files where they do include the reference "John Chan et= la. in IEEE Transactions On Computer-Aided Design, Vol. 10. No. 4, April 1= 991".>=20OK thanks Bill, So I should go looking for the Siemens book. I'll see wha= t the web has to offer. =20 George H.>=20 > --=20 >=20 > Bill Sloman, Sydney
Reply by ●July 18, 20132013-07-18
George Herold wrote:> On Wednesday, July 17, 2013 7:56:00 PM UTC-4, Bill Sloman wrote:[...]>> >> When Siemens became EPCOS, and the magnetic business was eventually >> sold off to TDK the situation got a bit confused, and I've lost >> touch with the relevant web-sites. >> >> >> >> If Winfield Hill ever does get around to releasing the third >> edition of "The Art of Electronics" it may include a decent section >> on magnetics - I certainly put in a personal plea for one a few >> years ago. >> >> >> >> At the time I laid a fair bit of emphasis on the transformer >> equation >> >> >> >> V1 = L1. dI1/dt + M. dI2/dt >> >> >> >> V2 = M. dI1/dt + L2. dI2/dt >> >> >> >> where M = k.(L1.L2)^0.5 >> >> >> >> and k is around 0.99 for a well-coupled transformer. L1 and L2 are >> the inductances of the coupled windings and M is the mutual >> inductance of the two windings (in Henries). >> >> >> >> Snelling never seems to have used it, and it came as a revelation >> to me when I found it in the Siemens application notes. >> >> >> >> For advanced stuff, look for the John Chan model of hysteritic >> inductors in the LTSpice help files where they do include the >> reference "John Chan et la. in IEEE Transactions On Computer-Aided >> Design, Vol. 10. No. 4, April 1991". >> > > > OK thanks Bill, So I should go looking for the Siemens book. I'll > see what the web has to offer. >It's probably long gone. If you want to also learn about switcher design you might want to take a look at a book by Pressman. I don't have i but heard it being lauded by lots of engineers. Maybe someone can shed light as to the exact title or ISBN. On some online stores you can take a preview before buying. -- Regards, Joerg http://www.analogconsultants.com/
Reply by ●July 18, 20132013-07-18
On Thu, 18 Jul 2013 08:32:13 -0400, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 07/17/2013 11:01 PM, John K wrote: >> Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: >> >>> Okay, thanks all--I think I've got it figured out. I'm already using >>> a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off >>> the free AC and then use regulated cap multipliers to make a nice >>> quiet +-40ish volts. >>> >>> Cartoon attached. >>> >>> Cheers >>> >>> Phil Hobbs >> >> Very nice. You got rid of the horrible ringing at the drain on most of >> the other versions posted here. That is very nice. >> >> I was wondering why you want to dump 2.25 watts into R2, the 100 ohm >> going from -15V to ground. >> > >That's the load, i.e. it represents the circuit that needs the -15V. > >> I couldn't see why it was needed, so I removed it. Circuit works fine. >> >> There may be some loading effect on either or both supplies that require >> it, but as currently shown, the reason is not apparent. >> >> Similarly, D6 which was in parallel with R2 doesn't seem to be needed >> either. Removed it also. Circuit works fine. > >That was there to get rid of a weird polarity reversal at power-on. The >power-on transient isn't representative of what the real circuit will >do--the chip has built-in current limiting, whereas this version goes up >to like 7A on power-up--but putting +3V on a -15V rail isn't usually >considered too healthy for the rest of the circuitry.Running it on LTspice: The current thru M1 goes up to 12A, with a half sinusoidal envelope, base width of nearly a millisecond. What is the actual schematic/chip that "has built-in current limiting"? [snip] ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by ●July 18, 20132013-07-18
On 07/18/2013 10:06 AM, Joerg wrote:> George Herold wrote: >> On Wednesday, July 17, 2013 7:56:00 PM UTC-4, Bill Sloman wrote: > > [...] > >>> >>> When Siemens became EPCOS, and the magnetic business was eventually >>> sold off to TDK the situation got a bit confused, and I've lost >>> touch with the relevant web-sites. >>> >>> >>> >>> If Winfield Hill ever does get around to releasing the third >>> edition of "The Art of Electronics" it may include a decent section >>> on magnetics - I certainly put in a personal plea for one a few >>> years ago. >>> >>> >>> >>> At the time I laid a fair bit of emphasis on the transformer >>> equation >>> >>> >>> >>> V1 = L1. dI1/dt + M. dI2/dt >>> >>> >>> >>> V2 = M. dI1/dt + L2. dI2/dt >>> >>> >>> >>> where M = k.(L1.L2)^0.5 >>> >>> >>> >>> and k is around 0.99 for a well-coupled transformer. L1 and L2 are >>> the inductances of the coupled windings and M is the mutual >>> inductance of the two windings (in Henries). >>> >>> >>> >>> Snelling never seems to have used it, and it came as a revelation >>> to me when I found it in the Siemens application notes. >>> >>> >>> >>> For advanced stuff, look for the John Chan model of hysteritic >>> inductors in the LTSpice help files where they do include the >>> reference "John Chan et la. in IEEE Transactions On Computer-Aided >>> Design, Vol. 10. No. 4, April 1991". >>> >> >> >> OK thanks Bill, So I should go looking for the Siemens book. I'll >> see what the web has to offer. >> > > It's probably long gone. If you want to also learn about switcher design > you might want to take a look at a book by Pressman. I don't have i but > heard it being lauded by lots of engineers. Maybe someone can shed light > as to the exact title or ISBN. On some online stores you can take a > preview before buying. >Pressman, Billings, & Morey, "Switching Power Supply Design" 3rd Ed, McGraw Hill. I bought a copy the other day but haven't spent much time with it yet. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●July 18, 20132013-07-18
On 07/18/2013 10:32 AM, Jim Thompson wrote:> On Thu, 18 Jul 2013 08:32:13 -0400, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > >> On 07/17/2013 11:01 PM, John K wrote: >>> Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: >>> >>>> Okay, thanks all--I think I've got it figured out. I'm already using >>>> a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off >>>> the free AC and then use regulated cap multipliers to make a nice >>>> quiet +-40ish volts. >>>> >>>> Cartoon attached. >>>> >>>> Cheers >>>> >>>> Phil Hobbs >>> >>> Very nice. You got rid of the horrible ringing at the drain on most of >>> the other versions posted here. That is very nice. >>> >>> I was wondering why you want to dump 2.25 watts into R2, the 100 ohm >>> going from -15V to ground. >>> >> >> That's the load, i.e. it represents the circuit that needs the -15V. >> >>> I couldn't see why it was needed, so I removed it. Circuit works fine. >>> >>> There may be some loading effect on either or both supplies that require >>> it, but as currently shown, the reason is not apparent. >>> >>> Similarly, D6 which was in parallel with R2 doesn't seem to be needed >>> either. Removed it also. Circuit works fine. >> >> That was there to get rid of a weird polarity reversal at power-on. The >> power-on transient isn't representative of what the real circuit will >> do--the chip has built-in current limiting, whereas this version goes up >> to like 7A on power-up--but putting +3V on a -15V rail isn't usually >> considered too healthy for the rest of the circuitry. > > Running it on LTspice: The current thru M1 goes up to 12A, with a half > sinusoidal envelope, base width of nearly a millisecond. What is the > actual schematic/chip that "has built-in current limiting"? > > [snip] > > ...Jim Thompson >I'll probably hang it on the output of an LM2594. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●July 18, 20132013-07-18
On Thu, 18 Jul 2013 10:39:22 -0400, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 07/18/2013 10:32 AM, Jim Thompson wrote: >> On Thu, 18 Jul 2013 08:32:13 -0400, Phil Hobbs >> <pcdhSpamMeSenseless@electrooptical.net> wrote: >> >>> On 07/17/2013 11:01 PM, John K wrote: >>>> Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: >>>> >>>>> Okay, thanks all--I think I've got it figured out. I'm already using >>>>> a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off >>>>> the free AC and then use regulated cap multipliers to make a nice >>>>> quiet +-40ish volts. >>>>> >>>>> Cartoon attached. >>>>> >>>>> Cheers >>>>> >>>>> Phil Hobbs >>>> >>>> Very nice. You got rid of the horrible ringing at the drain on most of >>>> the other versions posted here. That is very nice. >>>> >>>> I was wondering why you want to dump 2.25 watts into R2, the 100 ohm >>>> going from -15V to ground. >>>> >>> >>> That's the load, i.e. it represents the circuit that needs the -15V. >>> >>>> I couldn't see why it was needed, so I removed it. Circuit works fine. >>>> >>>> There may be some loading effect on either or both supplies that require >>>> it, but as currently shown, the reason is not apparent. >>>> >>>> Similarly, D6 which was in parallel with R2 doesn't seem to be needed >>>> either. Removed it also. Circuit works fine. >>> >>> That was there to get rid of a weird polarity reversal at power-on. The >>> power-on transient isn't representative of what the real circuit will >>> do--the chip has built-in current limiting, whereas this version goes up >>> to like 7A on power-up--but putting +3V on a -15V rail isn't usually >>> considered too healthy for the rest of the circuitry. >> >> Running it on LTspice: The current thru M1 goes up to 12A, with a half >> sinusoidal envelope, base width of nearly a millisecond. What is the >> actual schematic/chip that "has built-in current limiting"? >> >> [snip] >> >> ...Jim Thompson >> > >I'll probably hang it on the output of an LM2594. > > >Cheers > >Phil Hobbs0.5A Limit, or is there a way to set it lower? ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by ●July 18, 20132013-07-18
On 07/18/2013 10:49 AM, Jim Thompson wrote:> On Thu, 18 Jul 2013 10:39:22 -0400, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > >> On 07/18/2013 10:32 AM, Jim Thompson wrote: >>> On Thu, 18 Jul 2013 08:32:13 -0400, Phil Hobbs >>> <pcdhSpamMeSenseless@electrooptical.net> wrote: >>> >>>> On 07/17/2013 11:01 PM, John K wrote: >>>>> Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: >>>>> >>>>>> Okay, thanks all--I think I've got it figured out. I'm already using >>>>>> a Simple Switcher buck to make -15 from +16-20, and I'll piggyback off >>>>>> the free AC and then use regulated cap multipliers to make a nice >>>>>> quiet +-40ish volts. >>>>>> >>>>>> Cartoon attached. >>>>>> >>>>>> Cheers >>>>>> >>>>>> Phil Hobbs >>>>> >>>>> Very nice. You got rid of the horrible ringing at the drain on most of >>>>> the other versions posted here. That is very nice. >>>>> >>>>> I was wondering why you want to dump 2.25 watts into R2, the 100 ohm >>>>> going from -15V to ground. >>>>> >>>> >>>> That's the load, i.e. it represents the circuit that needs the -15V. >>>> >>>>> I couldn't see why it was needed, so I removed it. Circuit works fine. >>>>> >>>>> There may be some loading effect on either or both supplies that require >>>>> it, but as currently shown, the reason is not apparent. >>>>> >>>>> Similarly, D6 which was in parallel with R2 doesn't seem to be needed >>>>> either. Removed it also. Circuit works fine. >>>> >>>> That was there to get rid of a weird polarity reversal at power-on. The >>>> power-on transient isn't representative of what the real circuit will >>>> do--the chip has built-in current limiting, whereas this version goes up >>>> to like 7A on power-up--but putting +3V on a -15V rail isn't usually >>>> considered too healthy for the rest of the circuitry. >>> >>> Running it on LTspice: The current thru M1 goes up to 12A, with a half >>> sinusoidal envelope, base width of nearly a millisecond. What is the >>> actual schematic/chip that "has built-in current limiting"? >>> >>> [snip] >>> >>> ...Jim Thompson >>> >> >> I'll probably hang it on the output of an LM2594. >> >> >> Cheers >> >> Phil Hobbs > > 0.5A Limit, or is there a way to set it lower? > > ...Jim Thompson >No, there isn't, but half an amp isn't awful. The current limit on that chip is a bit squishy, as well--the datasheet limits are pretty wide, but it's guaranteed to be less than 1.25A iirc. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
