5V Regulation from 4.8V

Go To Last Post
37 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,

I started a thread entitled "capacitor recommendation request" that I no longer think captures the essence of what I'm trying to find out... so I'm re-posting a simplified version of my question here. The question is....

If you would like to hold a steady, regulated 5V to power an AVR, sensor, and keep the AREF pin at a known, fixed value using a 4.8V battery (which ranges from about 6V fully charged down to 4.5V over time)... while simultaneously using that same battery to drive a couple servos which can draw a few amps of current in short bursts when making quick moves... what would the best approach be?

My current setup involves the 4.8V battery supply wired straight to the servos, and then also into a 100 uF cap which is the input to a 5v low dropout linear regulator, with another 100 uF output cap that drives the AVR, sensor, and AREF pin. The problem I've been seeing is that the momentary bursts of power required by the servos cause the voltage at the AREF pin to drop for an instant causing my ADC conversions to be inaccurate.... even though I have two caps (.1uF & 22pF) at the AREF pin. It was suggested to me (by Bob) that I use a Shottky diode to decouple the servos from the capacitor to the 5V regulator... which I tried, but did not have success with until increasing my battery supply voltage to a higher voltage. I noticed, however, that the problem is also fixed by just using the higher supply voltage alone... without the Shottky diode.

If I'd like this to work with a 4.8V battery, however, I'm assuming I would need to use one of those switched mode regulators that can output a higher voltage than its input. So what I'm wondering is whether I should get one that takes an input from some range such as 4V to 8V and outputs a regulated 5V... or would it be better to use two voltage regulators... a switched mode to bring the level up to about 6V and then a linear mode one to hold the 5V from the input 6V? I'm not familiar enough with how (or how well) the switched mode regulators work and am just wondering if by using only the one and expecting it to bring my 4.8V up to 5V will end up with the same problem I was previously seeing with my initial setup... i.e. voltage dropping any time the servos make a quick move.

Anyway, I hope this was clearly enough worded that it's understandable what I'm trying to do... and would greatly appreciate any feedback. Bob & Kartman, thanks for your replies to my previous post.

- James

Last Edited: Sat. Sep 1, 2007 - 09:46 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

James, since the voltage from the battery can vary from less than the required Vcc to more than the required Vcc, you need a smart switcher. I think it's called a buck-boost converter, and happen to know that Linear Technology has some very nice solutions to this problem. It was mentioned here recently .... but cannot remember the thread. LT's website will be the place to start your search.

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nard,

Thanks a lot for your reply... I've actually read several of those "buck boost" converter's datasheets which have just left me pulling my hair out in confusion. i.e. I'll find something that sounds just right, such as an input range of 4V-12V which outputs 5V... but then as I click on that & read the datasheet they'll show a schematic of how to set the regulator up and show it taking 7V-12V at the input and outputting 5V. I'll click back to make sure I got the right part that was supposed to have an input range from 4V and it's correct... the diagram just does not match what the thing is supposed to do. I cannot believe how confusing people always have to make those datasheets... They all have such a simple thing to explain but go out of their way to make the explanation as confusing as possible. I was kind of left wondering if those buck-boost regulators can only be configured as one way or the other? i.e. If you want to output a lower voltage than the input do you configure it one way... and if you want it to increase the voltage have to configure it a different way? Or can it with one configuration take a time varying voltage that starts a little higher & ends up a little lower and always output 5V? I'm kind of at a loss from what I've read...

James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I suggest you PM Jayjay1974 ... he has a much better memory than I do, and, what's even more important: he is quite familiar with LT-chips.

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There is a configuration called SEPIC that should do the trick

Link to a National Appnote:
http://www.national.com/an/AN/AN-1484.pdf

edit:added link

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Buck-boost is the way to go, though expensive. I also have had extensive experience with LT chips.

You might reconsider whether or not you REALLY need to hold the voltage a 5V. Most of the micros (and logic) will work down to 4V or so with little or no change in performance except that the output logic swings are also reduced.

I would also reconsider whether you need to regulate the motor voltge. The torque will go down at low battery, but that should be OK in many apps. Except maybe for a battle-bot. It is likely that you are using PWM control on the motors anyway and that can be used to compensate for changes in battery voltage.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jim,

Well if I'm not going to try to hold exactly 5V, then I would at least like to try to better isolate my AVR & sensor side from the servo side of my circuit. I had hoped that by sending the battery power straight (unregulated) to the servos, but then going through a 5V linear regulator with input & output capacitors to drive the AVR, sensor, and AREF pin would somewhat isolate me from the servo noice... but it doesn't. I've placed two capacitors on my AREF pin, yet I still get really bad momentary voltage drops there when the servos make a fast move and my ADC results are corrupted... I originally thought maybe I just didn't have big enough capacitors at my 5V regulator, then tried using a Shottky diode between the battery source and input capacitor to the 5V reg at Bob's suggestion on my previous thread, but that did not seem to make a difference either... at least with my given setup. The only thing that helped a lot was when I used a larger battery supply, such as a 5-cell 6V battery instead of the 4-cell supply. With that, I got a pretty steady regulated 5V at the AREF pin no matter what the servos were doing.

So you could say that the problem has been fixed... but I'd really like for it to work with a 4-cell 4.8V battery supply instead... and that gives me really bad voltage drops on AREF when the servos twitch. So I had thought that maybe the answer was to use the buck-boost reg to get up to 5 or even 6V and then use the linear regulator after that to regulate back down to 5V to simulate the results I got with the 5-cell battery... but is there an easier way to accomplish a similar result? If I just stick with my 4.8V battery & linear 5V regulator, is there a way to better isolate the servos from the regulated AVR side of the circuit? Should I try using a lower voltage regulator, such as 4.5V or something? I know it sounds weird saying I'm driving a 5V reg with a 4.8V battery, but the voltage starts out at about 6V when that 4.8V battery is fully charged... I was reading about using an L-C network in series to help isolate one circuit side from another... and the AVR datasheet actually shows using an inductor as well as capacitor at the AREF pin, where I have only put a capacitor. Does the inductor make a big difference? Could that be a big part of my problem? I went to Frye's Electronics today to get a few to experiment with, but they do not have any... they've got a wall of capacitors... but no inductors! I guess I'll have to order one...

Thanks,
James

Last Edited: Sun. Sep 2, 2007 - 02:17 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Build or buy? There are certainly COTS DC-DC converters that put out 5V that will accept 4-12V. I think. I've looked at lots of power supply catalogs... seems like I remember some....

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Either... I'm just debating whether (if I get one of those) I should get one that outputs 5V, or something higher and then use a 2nd regulator to come down to 5V just because the only time I saw good performance was when driving my current circuit with a 6V battery? Or do you think I just didn't have the isolation setup right?

- James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You should certainly be able to isolate quirks of the circuit from quirks of the power supply. If it works on a lab supply, just get a ps with the current rating on the display and you're golden.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd consider running the micro at 3.3V to maintain enough voltage differential. Other option is a step-up switching regulator to 5.5 volts and a linear low drop postregulator to 5 volts. The output from the switcher would be higher than 5.5V with fresh batteries but that's not a problem in this case. The regulation loop must be fast enough to react to the input voltage drop caused by the servos.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
You might reconsider whether or not you REALLY need to hold the voltage a 5V. Most of the micros (and logic) will work down to 4V or so with little or no change in performance except that the output logic swings are also reduced.

Jim


I concur with Jim. What if you drop your ARef to 2.55V (internal reference) and measure with that ? You might have to scale down your input voltage with a simple voltage divider, but this is still much simpler and cheaper than a complex voltage regulation circuit.

Markus

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Okay, I think I just fixed the problem....

I need to use AVCC as my AREF because my sensor's output ranges from 0.25 to 4.75V... but it looks like the problem was just fixed by using the Schottky diode that Bob kept suggesting... I had tried it before but didn't get good results for some reason... maybe it wasn't hooked up right? But this time, I added a slightly larger electrolytic capacitor at the input of the linear 5V regulator (300uF), and put the diode between the 4.8V battery and the input capacitor. I thought I had it there before... but anyway... the results look really good now. If I probe the voltage before going across the diode (i.e. what the servos see) there are spikes from 5.25V down to about 4V whenevere the servos move... When I probe on the other side of the Shottky diode the voltage is almost flat... and then on the output of the 5V regulator is even more solid while I move the servos rapidly back & forth. The voltage is a little low... it's only at about 4.8V on the AVR side, but that's okay! I'm just excited that it's no longer jittering with every move of the servos!!!! I guess the diode is dropping my battery voltage from about 5.25V down to 4.8V (I couldn't find a 0.2V diode, mine is a 0.45V), and the voltage regulator must just be letting it pass thru unaltered? But I guess this way it'll be in there to regulate in case the battery is charged up at 6V, or a bigger battery is used... I think the results are really pretty good now... Should I also look into adding an inductor in series with the diode, or just leave it alone? Also, the datasheet shows putting an inductor at the AVcc pin along with the capacitor... I found an article on the web (forgot the link) where a guy gave this long dissertation on the need for including inductors along with every one of your 0.1uF capacitors on the board. He was saying that everyone who thinks it is fine to just put 0.1uF capacitors on all of their ICs just because that's how it's been done for the last 40 years do not understand that they should also be placing inductors along with them... but then he went on to say that you need to be careful, however, because the wrong value inductor could cause oscillations... but then did not really recommend a good value... The datasheet shows using a 10uH inductor at the AVcc pin, which I guess does not require much current, but to add one at the voltage regulator would need a lot more current and I was noticing that they get pretty huge if you want them to conduct 400-500mA's or so of current....

So now I'm just contemplating whether or not I should still try out the switching regulator in place of the linear one & whether to try adding inductors at the input & AREF pin... or just leave well enough alone? Any thoughts on what, if any, improvements those changes might bring?

Thanks,
James

Last Edited: Sun. Sep 2, 2007 - 07:04 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The 0.25W resistor-like 10 uH's I have take 360 mA. But there are also those small I-cores that take even more.

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

how are you going to deal with varying vcc (== varying aref)?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was just going to sample the built-in 1.1V ref in my control loop, low pass filter the result, and use that as my conversion factor from digital counts to volts...

Hey, let me ask you this, though... if I add an inductor in series with the diode to help isolate the servos from the AVR side, should it be put before or after the diode... or does it not make a difference? Also, when I purchase an inductor (I was going to get a 330 uH) do I need to get one with a max current spec to support the servo usage (4-5 amps) even though I'm only wanting to run the AVR & a couple sensors off of the diode and inductor output... which will only be more like 28 mA total?

Thanks,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

When the two are in series then it does not matter which of the two is first or second. It makes absolutely no difference.

The inductor needs the current capacity for the current flowing through it. If this is only the AVR supply current a couple of mA are sufficient. How much current the servos need is completely irrelevant, as long as their current does not traverses the inductor.

Markus

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks Markus.... I knew that, in general, the order of two things in series made no difference... I just wondered if because the diode acts as a one-way valve it might matter whether you were on the input side or output side of it. Thanks for your answers... they are very much appreciated!

James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Werent there some USB powerchips designed to give 5v , for inputs in that range ??

Afaik the usb specs also allowed for less than 5v.

/Bingo

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So you dont have the low dropout 5v reg anymore? Try it again with the reg? One more silly suggestion that I havent heard for 20 or 30 years... a Germanium diode only has .2V drop... wonder if they will handle the 100ms or whatever it takes to run the avr?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
So you dont have the low dropout 5v reg anymore? Try it again with the reg?

No, no, no... I am still using the low dropout 5V linear regulator... The whole thing is working fine now, it seems, with the addition of that Schottky diode you suggested. When I tried originally, I must've had something setup wrong... cause now it is working great! It's amazing to see how much of a difference that diode makes... I setup an experiment to see the difference on my oscilloscope... and without the diode, as I'd apply a little pressure to the servos you'd see these sharp dips across the whole screen dropping down to about 4V momentarily... With the diode in place & doing the same test the voltage would remain completely flat as I'd press on the servos... The only negative was that the output voltage is only like 4.6 to 4.8 V by then... but the Atmel chip seems to be happy with that... and as long as the voltage at the AREF pin stays stable... I'm fine with that. I ordered one of those 5V switching control regulators to play with also in a DIP package... so I'll see how that does by comparison... but I'm happy with what the 5V linear reg is giving me right now with the Schottky diode in place. Thanks a lot for the suggestion.

Let me ask one more question regarding the use of an inductor to help isolate my battery power from the AVR. Would it be more effective to place one before or after my voltage regulator, or would that still make no difference... as with before or after the diode?

Thanks,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The preference seems to have moved away from buck/boost (SEPIC) regulators now, but for the record, here's one I did for a piece of battery-powered gear. It delivers 5V at 300mA from 3-15V input. It's a little difficult to prototype since the regulator is an SOT23-5 package, but possible with care. All the caps are ceramic, the inductors are Coilcraft 2012 size and the whole circuit is incredibly compact. If you substitute an LT1615 chip it will work down to 1V (single AA cell).

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hey, that looks great!!! Is the output pretty stable from one of those as compared with a linear regulator? I just added a Schottky diode to the input of my linear regulator (between battery power & voltage reg input) which helped out enormously with servos trynig to pull down the voltage every now & then... would I still need that if using this SEPIC design since it looks like it already has a diode at its output? And same question for the inductor... it looks like your design has a couple inductors already... so would an additional one from the battery input to the regulator no longer be necessary? Why are people moving away from these lately? With a regulator being in a package that tiny (the SOT-23) do you have to worry about giving it a heat synch of some sort?

Thanks,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have used the LT1615 and it is a great little chip. Has a wide input and output range. There is a current limited version the LT1615-1 which limits the output current to 100mA.

Linear Tech has a demo board that uses the LT1613 which you can replace w/the LT1615.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The output is rock steady. Switching frequency is 1.2MHz so it only needs tiny caps. It does need two coils for the SEPIC action. These can either be separate, as my design, or coupled on the same core - it doesn't seem to make much difference, except space. Many coil makers have come out with a range of dual-winding cores, aka coupled inductors, since many small battery devices these days have a SEPIC. This circuit doesn't get hot at all and needs no heat sink. However, it will not survive a reversed input, so an input Schottky diode (before the caps) would be fairly important and would help hold the output up when the input voltage dips. A small inductor on the input would block high frequency spikes from the servos, if that's a problem, but anything over a few microhenries would interfere with the switching action and regulation.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
A small inductor on the input would block high frequency spikes from the servos, if that's a problem, but anything over a few microhenries would interfere with the switching action and regulation.
Would it be safer to just use one extra (larger) capacitor at the input instead, like just a 100 uF for instance, rather than the inductor? Or would the larger capacitor possibly interfere with the switching action too?

I'm trying to incorporate your schematic into my design right now.... Thank you very much for sharing that!

Sincerely,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can calculate the capacitance you need from the formula C(uF) = Amp.microsec per volt. So if the average input current is 0.1A and you keep my values of 4.4uF, and you figure it will go out of regulation when the input drops by 2V, then you get (4.4 * 2)/0.1 microseconds (88us) of grace on a brownout. 100uF would give you about 2ms. But because the current is taken in short, sharp pulses, an aluminum alone won't work here - you also need the ceramics. Check the regulator data sheet for recommended layout, too.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks a lot for your help!

James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Okay, so I had ordered this
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=576-1514-5-ND
before getting your link to what looks like a better one thinking since it came as an 8-pin DIP I could experiment with it on a breadboard... and am now noticing that it only regulates at 5V if the input is greater than 5V even though the description says that it will output 5V if from an input from 4-40V. I just followed the diagram of Figure 1 in its datasheet... am I missing something?

Thanks,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Here are three common types of switching regulator - Buck, Boost and SEPIC. In each case, the switch represents the transistor in the IC. With a Buck converter, the switch starts current flowing in the coil and when it opens, the diode maintains the current path to the output. This kind always steps down - the maximum output is when the switch is permanently on, and it can never be higher than the input. With a boost converter, the switch starts current flowing in the coil and when it opens, the current tries to continue flowing and raises the voltage at the switch end until it reaches the output capacitor voltage. This kind generally steps up, and can go to quite high voltages. The SEPIC is a combination - a boost converter first, followed by a buck coil-diode combination. When the switch opens, the first coil dumps its energy through the capacitor into the second coil and the output, reaching whatever voltage is necessary; the second coil keeps it going after the initial spike.

Most regulator ICs are either buck or boost; few can do both because of the way the transistor switch is arranged. The LM2574 is a buck converter only, I'm afraid. You might look at the MC34166, but it uses a lot of external parts and I don't know if it will work at the voltages you need.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hey, thanks a lot for that lesson on regulator types......

Let me ask you something else... In experimenting with the "buck" regulator that I got, I noticed that it's output is a rock solid 5 volts when the input is something higher, like 6 volts. However, when I try integrating it into my design with two servos feeding off of the same battery source... even using a Shottky diode between the battery pack and input capacitor to the regulator, the regulator's output looks horrible when the servos are activated. i.e. The scope trace immediately turns from a rock solid 5 volts to more like a saw-tooth wave oscillating about 5V at maybe +/- 0.1 or 0.2 volts... The only way to diminish this seems to be adding a really large input capacitor... I tried 470 uF and it clears it up... 100 uF was borderline... So my question is... if I try implementing your SEPIC design, if I replace your capacitors C4 & C5 (2.2 uF each) with one large 470 uF capacitor instead... would that cause any problems with the SEPIC action or anything as far as you know? I'm going to try having a board made and am trying to figure out which capacitors to use... and it seems like, in light of what I saw with the buck regulator, that going with as large an input cap as possible might be the safest way to go... as long as it doesn't interfere with the regulator at all.

Thanks again,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
if I replace your capacitors C4 & C5 (2.2 uF each) with one large 470 uF capacitor instead... would that cause any problems with the SEPIC action

Keep the low-ESR ceramics in parallel with the big cap. Actually a couple of 0.1uF should be enough, close to the regulator. The ground between the input cap, output cap, regulator and second coil needs to be short and thick as it will be pulsing up to an amp at 1.4MHz.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

This has an example layout for the LT1613 SEPIC configuration.
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1003,C1042,C1035,P1740,D8197

I have also pulled off the LT1613 and stuck in an LT1615 with success.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks a lot for your advice....

James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Peret,

On your schematic, where you use the 10k and 30k resistors to build an output of 5V per the datasheet formula... if I wanted the output to be something larger, such as 6V or 7V by changing those resistor values appropriately, would everything else still work as anticipated? I would assume so, except that I'm slightly confused as to what factors go into limiting the final output current or determining the input voltage range, etc... I just wanted to make sure there's nothing else I'm overlooking if I try changing those resistor values to get a slightly higher voltage output? The reason I'm contemplating this... is that I'm thinking about making a dual regulator system where I use the SEPIC to produce a voltage of about 7V and then send that into a 5V linear regulator. My thinking is (correct me if I'm wrong) that I may get a better result that way than with just the "diode and capacitors" scheme to isolate my servos from the rest of the circuit elements... especially the 5V AREF ADC reference. Although the diode and capacitor scheme significantly help with isolating the servos drawing current/voltage from the same battery source, I'm still getting slight dips in voltage at the AREF pin whenever the servos twitch... the bigger and bigger I make the capacitors the smoother it seems to get, but I noticed that the AVR's 1.25V reference stays absolutely rock solid... which made me think maybe if I had a higher voltage (such as 7V) being produced by the SEPIC, and trying to regulate that down to 5V, that perhaps it would stay rock solid as well... even if the 7V was dropping to 6 or 5.5 every now and then? Any thoughts on that? And the main thing I was wanting to know is just whether modifying the SEPIC regulator's resistor values to produce a higher voltage will change anything else that I'm unaware of... such as input voltage range, output current, etc...

Thanks,
James

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
if I wanted the output to be something larger, such as 6V or 7V by changing those resistor values appropriately, would everything else still work as anticipated?

I believe that would be alright. You might check the data sheet to see what is the maximum output voltage, but I seem to recall (without looking it up) that it's a good deal higher than 7V. Just pick the nearest preferred resistor values that give you about the right reference point. The engineers at National, Linear Tech and so forth seem to use values like 142.4k, 403.6k, or even weirder values; maybe they have a bench-top laser trimmer and always cut to size. I generally just pick the nearest preferred values that give me the output. Sometimes I pick off the 1% range, like 30.1k or 57.6k, but in truth, the 5% range 30k and 56k would probably be good enough because the internal reference on the regulator has a few percent tolerance.

The best way to avoid losing regulation is to isolate the regulator from the servo, like with a diode and a huge capacitor - or a small battery - but I guess post regulating would do, for a one off anyway.

As an aside, I've seen servo systems where the linear servo output stage is bracketed top and bottom by switchers, which hold the voltage to the output stage a couple of volts off the setpoint. The switchers themselves are too slow to do the actual servo job so a linear output stage is needed, but it's always operating with just a few volts across the output transistors and the dissipation is kept very low. Probably not applicable to your application but it comes in handy when the load is a big DC motor drawing 10A at 100V.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Okay, thank you... I've designed a PCB with a few different regulation options built into it that I can connect/disconnect with jumpers... so I can experiment with a few different options. I'll let you know how it turns out... Thanks a lot for your help.

Sincerely,
James