Post by thread:[EE]: Suggestions on how to generate 40V 1A DC fro

In my never ending quest to develop a robust and easy to build power supply for my mobile MP3 player, I've come across an interesting Lucent DC/DC converter from All electronics (http://www.allelectronics.com). The only problem is that the part requires 36-72V DC at about 1 amp for input. I've scoured the web for a useful boost setup, but every design I've seen that uses a regular transformer complains of very poor conversion efficiency. It's becoming clear from my readings that a flyback transformer setup is required to get efficient operation. However it's not clear how to obtain such a beast. The supply houses generally carry ordinary step down transformers. So any suggestions? The output doesn't require any rigid regulation, just the ability to provide 1A at about 40V or so. BAJ -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spam_OUTlistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

At 05:57 PM 3/26/01 +0000, Bob Blick wrote: >On Mon, 26 Mar 2001, Byron A Jeff wrote: > > > In my never ending quest to develop a robust and easy to build power supply > > for my mobile MP3 player... > >http://www.bobblick.com/bob/projects/yamm/ There is a variant of the boost that uses a tapped inductor to lower the inductor requirements. The only tradeoff is that the switch needs to be a bit higher voltage. (usually not a problem) -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

At 05:57 PM 3/26/01 +0000, Bob Blick wrote: >On Mon, 26 Mar 2001, Byron A Jeff wrote: > > > In my never ending quest to develop a robust and easy to build power supply > > for my mobile MP3 player... > >http://www.bobblick.com/bob/projects/yamm/ Nice design! :) Did you try multifilar winding to reduce the energy that ends up in the snubber? -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email .....listservKILLspam.....mitvma.mit.edu with SET PICList DIGEST in the body

On Mon, 26 Mar 2001, David VanHorn wrote: > >http://www.bobblick.com/bob/projects/yamm/ > > > Nice design! :) Thanks. > Did you try multifilar winding to reduce the energy that ends up in the > snubber? Yes, the input and +5 out are wound together. There's not much energy wasted in the snubber, if I had to guess it'd be under 1/4 watt judging by the resistor not showing any heat. It'll run without the snubber, but a little bit of zenering happens on the mosfet. The only drawback is that I used the LM3524 instead of something like the UC3843 that is meant to drive a mosfet, so there's a little extra waste with the pinch-off resistors. Of course the UC3843 has drawbacks too. BTW this mp3 player has been running more than 2 years under the seat of my car, from winter in eastern Oregon to summer in Phoenix. It powers the computer down to 8 volts input, so you can start the car while it's running. Cheers, Bob -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservspam_OUTTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

> >Yes, the input and +5 out are wound together. There's not much energy >wasted in the snubber, if I had to guess it'd be under 1/4 watt judging by >the resistor not showing any heat. It'll run without the snubber, but a >little bit of zenering happens on the mosfet. I'd be very tempted to try 20 and 40 turn windings, hexifilarly, with twisted conductors. >The only drawback is that I used the LM3524 instead of something like the >UC3843 that is meant to drive a mosfet, so there's a little extra waste >with the pinch-off resistors. Of course the UC3843 has drawbacks too. I like them. I've used the 384X several times. It's got a pretty beefy fet drive, so much so that I've had to slow down the gate charge to avoid transients on the source. >BTW this mp3 player has been running more than 2 years under the seat of >my car, from winter in eastern Oregon to summer in Phoenix. It powers the >computer down to 8 volts input, so you can start the car while it's >running. I may need to do something like this soon. I'm running out of what I can do with my libretto in the car.. I'm gonna end up with a LAN if i'm not careful! :) -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body

Byron A Jeff wrote: > > In my never ending quest to develop a robust and easy to build power supply > for my mobile MP3 player, I've come across an interesting Lucent DC/DC > converter from All electronics (http://www.allelectronics.com). The only problem > is that the part requires 36-72V DC at about 1 amp for input. I've scoured > the web for a useful boost setup, but every design I've seen that uses > a regular transformer complains of very poor conversion efficiency. > > It's becoming clear from my readings that a flyback transformer setup is > required to get efficient operation. However it's not clear how to obtain > such a beast. The supply houses generally carry ordinary step down > transformers. > > So any suggestions? The output doesn't require any rigid regulation, just > the ability to provide 1A at about 40V or so. Hi Byron, you want to generate 40v 1A from 12v auto power? This sounds typical of many car audio applications, I think some of the hobby electronics shops sell kits for similar PSUs that power their car amplifier kits. Try http://www.jaycar.net. :o) -Roman -- http://www.piclist.com hint: To leave the PICList @spam@piclist-unsubscribe-requestKILLspammitvma.mit.edu

> > > > >Yes, the input and +5 out are wound together. There's not much energy > >wasted in the snubber, if I had to guess it'd be under 1/4 watt judging by > >the resistor not showing any heat. It'll run without the snubber, but a > >little bit of zenering happens on the mosfet. > > I'd be very tempted to try 20 and 40 turn windings, hexifilarly, with > twisted conductors. > > > >The only drawback is that I used the LM3524 instead of something like the > >UC3843 that is meant to drive a mosfet, so there's a little extra waste > >with the pinch-off resistors. Of course the UC3843 has drawbacks too. > > I like them. I've used the 384X several times. > It's got a pretty beefy fet drive, so much so that I've had to slow down > the gate charge to avoid transients on the source. > Ladies and Gentlemen, The above is what you get when two talented and competent engineers get into a heavy tech conversation. I do the same with my Linux group and we all do the same with PICs. But unfortunately it doesn't help very much. Way over my head. Bob. Your power supply doesn't meet the original specification. I know that you read the need to power a PC from car voltage. However I was really hoping the simplify the process of building the supply by using the self contained Lucent module and simply supply it the required 40V. Reason being is that because of the environment I need to build something that's as simple and rugged as possible. The more complexity, the more difficult to debug and repair. I understand the basics, but not the details. A few questions: - Why doesn't an ordinary transformer work efficiently in the opposite direction? I'm not clear why one cannot pump 12V @ 3A into a 24V 3A transformer and expect to get 60V out the other end. - Can such a device be built with a power coil? For example I can get my hands on a 10-50 uH coil with nearly 10 amps capacity. I used them in a National simple switcher project with good results. BTW the simple switcher doesn't seem to be an option because the final output current is 2.1A times the ratio of the input/output voltage. At 36-37 volts out from 12-14V input the max current is only about 700 mA. Maybe I can use a bigger external transistor to up the available current. - Where can I obtain cores and wires if I must roll my own transformer? What features are required in the core. From my RatShack.com catalog it seems the the permissivity of the core determines the inductance. How is that important in this application. Thanks for any help you can offer. BAJ -- http://www.piclist.com hint: To leave the PICList KILLspampiclist-unsubscribe-requestKILLspammitvma.mit.edu

{Quote hidden}>The above is what you get when two talented and competent engineers get into >a heavy tech conversation. I do the same with my Linux group and we all do >the same with PICs. >But unfortunately it doesn't help very much. Way over my head. >Bob. Your power supply doesn't meet the original specification. I know that >you read the need to power a PC from car voltage. However I was really hoping >the simplify the process of building the supply by using the self contained >Lucent module and simply supply it the required 40V. Reason being is that >because of the environment I need to build something that's as simple and >rugged as possible. The more complexity, the more difficult to debug and >repair. >I understand the basics, but not the details. A few questions: >- Why doesn't an ordinary transformer work efficiently in the opposite > direction? I'm not clear why one cannot pump 12V @ 3A into a 24V 3A > transformer and expect to get 60V out the other end. >- Can such a device be built with a power coil? For example I can get my hands > on a 10-50 uH coil with nearly 10 amps capacity. I used them in a National > simple switcher project with good results. BTW the simple switcher doesn't > seem to be an option because the final output current is 2.1A times the > ratio of the input/output voltage. At 36-37 volts out from 12-14V input the > max current is only about 700 mA. Maybe I can use a bigger external > transistor to up the available current. >- Where can I obtain cores and wires if I must roll my own transformer? What > features are required in the core. From my RatShack.com catalog it seems > the the permissivity of the core determines the inductance. How is that > important in this application. >Thanks for any help you can offer. First - An ordinary transformer will not work as you expect as it works on AC, not DC - but then I am assuming you do realise this, it just does not come across like that in what you wrote above. Second - Why build a converter to go from 12V to 40V to use a commercial converter to then go to whatever voltages you require? Why not just build the converter to go straight from the 12V to the voltages you require, it will be the same effort, and will have a cost saving. You may then be able to use the 10A inductor you mention above (without my knowing the end voltage you are looking for). -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

> >The only drawback is that I used the LM3524 instead of something like the >UC3843 that is meant to drive a mosfet, so there's a little extra waste >with the pinch-off resistors. Of course the UC3843 has drawbacks too. Did you ever wonder why we slam those gates to V++ and ground, when we really only need to get them above and below threshold a bit? :) -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com hint: To leave the PICList spamBeGonepiclist-unsubscribe-requestspamBeGonemitvma.mit.edu

> >Bob. Your power supply doesn't meet the original specification. I know that >you read the need to power a PC from car voltage. However I was really hoping >the simplify the process of building the supply by using the self contained >Lucent module and simply supply it the required 40V. Reason being is that >because of the environment I need to build something that's as simple and >rugged as possible. The more complexity, the more difficult to debug and >repair. All you'd need to do, is replace his output windings with a single winding. Looks like about 2 turns per volt, so 80 turns, and then set up the feedback resistors to keep it in regulation at 40V output instead of 5V. >I understand the basics, but not the details. A few questions: > >- Why doesn't an ordinary transformer work efficiently in the opposite > direction? I'm not clear why one cannot pump 12V @ 3A into a 24V 3A > transformer and expect to get 60V out the other end. You can do this. It's much less efficient, and noisy because of the lower chopping frequency. Look up a "Royer" converter. You'd use a 110-24VCT (or 48V, I forget) and two NPN transistors. >- Can such a device be built with a power coil? For example I can get my hands > on a 10-50 uH coil with nearly 10 amps capacity. This looks ok, for the inductor of a boost converter. Single transistor, schottky diode, output cap, and controller chip (could be a 555 of you're phobic of switcher chips) You have to switch fast enough that the current through the inductor in the "ON" state dosent' get past the 10A limit. 50-100kHz is a good range to work in. >I used them in a National > simple switcher project with good results. BTW the simple switcher doesn't > seem to be an option because the final output current is 2.1A times the > ratio of the input/output voltage. At 36-37 volts out from 12-14V input the > max current is only about 700 mA. Maybe I can use a bigger external > transistor to up the available current. The simple switchers are designed to be an "all in one" Look at the 384X series, or even a 555. You'll need a moderate FET to do this. >- Where can I obtain cores and wires if I must roll my own transformer? What > features are required in the core. From my RatShack.com catalog it seems > the the permissivity of the core determines the inductance. How is that > important in this application. Here's the biggest argument for running a boost converter. Cores can be hard to find, especially if you want specs with them. Simple inductors are easier. -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com hint: To leave the PICList TakeThisOuTpiclist-unsubscribe-requestEraseMEspam_OUTmitvma.mit.edu

> In my never ending quest to develop a robust and easy to build power supply > for my mobile MP3 player, I've come across an interesting Lucent DC/DC > converter from All electronics (http://www.allelectronics.com). The only problem > is that the part requires 36-72V DC at about 1 amp for input. A standard telco "48V" DC-DC converter... ooooh... Sudden realization - if they ever actually convert car electrical systems to 42V, all those telco devices should be usable off-the-shelf (not that they're very affordable that way, but...) I've wondered for "sensitive" car electronics whether it's a good idea to do this sort of "double conversion" for added isolation. Ie to run a laptop, use a 120VAC inverter with the standard laptop power supply, rather than looking for a device to produce/isolate the laptop voltages directly from 12V. Does this add any real protection, or not? (the standard laptop supply probably isn't set up to filter out the sort of transients that might occur in a car, and the typical inverter isn't necessarilly set up as a filter, either, so I'm worried that any safety would be "accidental" vs a well-designed car supply...) BillW -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

On Tue, 27 Mar 2001, Byron A Jeff wrote: > Bob. Your power supply doesn't meet the original specification. I know that > you read the need to power a PC from car voltage. However I was really hoping > the simplify the process of building the supply by using the self contained > Lucent module and simply supply it the required 40V. Reason being is that > because of the environment I need to build something that's as simple and > rugged as possible. The more complexity, the more difficult to debug and > repair. Hi Byron, I don't really see how it's simpler. If you build a 12-to-40 converter you need almost as many parts as building a triple output supply. The Lucent converter is not 100% efficient, so you have losses in two places. On a side note, 40 volts at 1 amp is only 40 watts - not enough to power a motherboard and 3.5 inch hard disk at startup. You can build a 12-to-40 volt converter with either a transformer or an inductor. The latter is called a boost regulator. The duty cycle is relative to the boost, and 50% is ideal, so a 12-to-24 is the ideal. 12-to-40 is less than ideal, so your caps will need to be bigger and better, and the switching transistor will need to be bigger, and the efficiency will be slightly less. Other than that it saves you a small bit of winding over a transformer. I was just trying to help by pointing you to my project - hey, it works and is documented and powers a PC quite well in a car environment. No math required. If you can't find the Arnold core I can point you to one at Digi-Key that works. Cheers, Bob -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestEraseME.....mitvma.mit.edu

> >ooooh... Sudden realization - if they ever actually convert car electrical >systems to 42V, all those telco devices should be usable off-the-shelf (not >that they're very affordable that way, but...) It's more fun than that! Lots of high power ham gear gets really easy with 30V+ supply. > Does this add any real protection, or not? (the standard laptop supply >probably isn't set up to filter out the sort of transients that might occur >in a car, and the typical inverter isn't necessarilly set up as a filter, >either, so I'm worried that any safety would be "accidental" vs a >well-designed car supply...) A boost reg will protect you from reverse voltages, but not overvoltage. It's series inductor and output cap will tend to soften any hard spikes, but there's a time limit, and it's not large. A flyback converter, (or forward, etc) will protect you absolutely, since no energy can get through the core except that the converter chip permits it. -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestmitvma.mit.edu

> Did you ever wonder why we slam those gates to V++ and ground, when we > really only need to get them above and below threshold a bit? :) 1 - Because FETs don't suddenly switch at a magic threshold. True, most of the action of over a small region, but you want to make sure the FET is fully on and fully off with some margin to spare under all conditions and device variations. 2 - In a power switching application, you want to minimize time spent in the transition region. Slamming over a larger range maximizes the slew rate in the vulnerable region, thereby minimizing device dissapation. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, RemoveMEolinEraseMEEraseMEembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestspam_OUTKILLspammitvma.mit.edu

At 01:08 PM 3/27/01 -0500, Olin Lathrop wrote: > > Did you ever wonder why we slam those gates to V++ and ground, when we > > really only need to get them above and below threshold a bit? :) > >1 - Because FETs don't suddenly switch at a magic threshold. True, most of >the action of over a small region, but you want to make sure the FET is >fully on and fully off with some margin to spare under all conditions and >device variations. > >2 - In a power switching application, you want to minimize time spent in the >transition region. Slamming over a larger range maximizes the slew rate in >the vulnerable region, thereby minimizing device dissapation. I know that.. I was thinking maybe it would be interesting to have two settable voltages, and have the SMPS chip drivers slam the gate between those voltages. Maybe not. -- Dave's Engineering Page: http://www.dvanhorn.org Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu

> >I understand the basics, but not the details. A few questions: > > >- Why doesn't an ordinary transformer work efficiently in the opposite > > direction? I'm not clear why one cannot pump 12V @ 3A into a 24V 3A > > transformer and expect to get 60V out the other end. > >- Can such a device be built with a power coil? For example I can get my hands > > >Thanks for any help you can offer. > > First - An ordinary transformer will not work as you expect as it works on AC, > not DC - > but then I am assuming you do realise this, it just does not come across > like > that in what you wrote above. Yes. I realize that the DC would need to be chopped. > Second - Why build a converter to go from 12V to 40V to use a commercial > converter to then go to whatever voltages you require? Why not just build the > converter to go straight from the 12V to the voltages you require, it > will be the > same effort, and will have a cost saving. You may then be able to use the 10A > inductor > you mention above (without my knowing the end voltage you are looking > for). Not the same effort, and the cost savings is debatable. Last year I rigged up a board using National simple switchers that ended up being 4 separate circuits. It worked, but it was noisy and not very reliable. With the new proposal I only have to build one switcher and use the $11 module to reliably generate the required voltages. BAJ -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestspamspamBeGonemitvma.mit.edu