Adjustable CC/CV PSU using laptop AC adapter?

[nbritton]

I have a 230W +19.5VDC laptop AC adapter that I'd like to adapt into a triple output variable constant current / constant voltage power supply. I'm sure something like this has already been done, any pointers where I can find more information about this? How would you accomplish this without using grossly inefficient linear regulators?

[ez24]

I have a 230W +19.5VDC laptop AC adapter that I'd like to adapt into a triple output variable constant current / constant voltage power supply. I'm sure something like this has already been done, any pointers where I can find more information about this? How would you accomplish this without using grossly inefficient linear regulators?

230W +19.5VDC laptop AC adapter  ==>  what brand ?

I'm sure something like this has already been done  ==> good luck

[Kleinstein]

If you don't want a pure linear regulator, it will be a switched mode regulator. Except for the input voltage there will be no big difference to starting from rectified mains voltage, except that one channel might use a step down without isolation.

[Fungus]

I have a 230W +19.5VDC laptop AC adapter that I'd like to adapt into a triple output variable constant current / constant voltage power supply. I'm sure something like this has already been done, any pointers where I can find more information about this? How would you accomplish this without using grossly inefficient linear regulators?

Browse eBay, there's dozens of things like this: http://www.ebay.com/itm/351276332291

[rjeberhardt]

I've done exactly that using one of these: http://www.ebay.co.uk/itm/5A-0-8V-29V-Constant-Current-Voltage-LED-Driver-Step-Down-Module-Charger-Module-/131584272548?hash=item1ea308a8a4 and an old Dell 20 V 3.5 A laptop power supply.  It works O.K. but when tested at 1.5 A load it gave 50 mV triangular wave ripple at about 300 kHz with 100 mV switching spikes with a fundamental frequency of about 10 MHz.  So, if you want to use it to power anything sensitive you need to be careful with the filtering.

Russell.

[ez24]

an old Dell 20 V 3.5 A laptop power supply. 
Russell.

Did you have to do anything with the center pin to get it to work?  (Dell PS use 3 wires)

[ez24]

There is no 3 output buck off the shelf item.

Here is a +- 12v buck  (rare item), using this and one of the ones you have picked, would give you a triple output

http://www.ebay.com/itm/251352773668?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

There are a lot of dual bucks that put out 12v and 5v

[Fungus]

There is no 3 output buck off the shelf item.

No, but you have have 3 single output bucks.

[rjeberhardt]

Did you have to do anything with the center pin to get it to work?  (Dell PS use 3 wires)

No.  The one I had just had a two wire output.

Russell.

[jay]

I've done exactly that using one of these: http://www.ebay.co.uk/itm/5A-0-8V-29V-Constant-Current-Voltage-LED-Driver-Step-Down-Module-Charger-Module-/131584272548?hash=item1ea308a8a4 and an old Dell 20 V 3.5 A laptop power supply.  It works O.K. but when tested at 1.5 A load it gave 50 mV triangular wave ripple at about 300 kHz with 100 mV switching spikes with a fundamental frequency of about 10 MHz.  So, if you want to use it to power anything sensitive you need to be careful with the filtering.

So did the ripple come from the module you got from that Ebay seller or from the Dell laptop power brick?

I've used Lenovo & IBM laptop PSUs for many projects including a bench power supply. The ripple is surprisingly low even at high currents (IIRC, I tested up to 4A). There was a bit of voltage drop at high currents, but that probably comes from the long cable and replacing the cable with something thicker would probably fix that. There might be a clear quality difference between brands.. At least I had several old power bricks from other brands and many of them didn't work at all or had serious issues with regulation or ripple.. Well, many were 10 years old so the output caps had died. Good to check them carefully before using.

[Fungus]

I've used Lenovo & IBM laptop PSUs for many projects including a bench power supply. The ripple is surprisingly low even at high currents (IIRC, I tested up to 4A). There was a bit of voltage drop at high currents, but that probably comes from the long cable and replacing the cable with something thicker would probably fix that.

I always thought the little center pin was some sort of voltage sense feedback to adjust for that. I might be wrong though, I'm just guessing.

[ajb]

I've used Lenovo & IBM laptop PSUs for many projects including a bench power supply. The ripple is surprisingly low even at high currents (IIRC, I tested up to 4A). There was a bit of voltage drop at high currents, but that probably comes from the long cable and replacing the cable with something thicker would probably fix that.

I always thought the little center pin was some sort of voltage sense feedback to adjust for that. I might be wrong though, I'm just guessing.

Dell at least uses the center pin to read an identifier from a little one-wire device in the adapter.  That way the laptop can identify whether a 65W or 90W adapter is connected and nag at you if you connect the wrong one.

[rjeberhardt]

So did the ripple come from the module you got from that Ebay seller or from the Dell laptop power brick?

The Dell PSU gives very little ripple on no load but when loaded the ripple shoots up to about 100 mV PP.  However it is switching at 200 kHz.  The ripple I'm getting on the Ebay module output is 300 kHz so is coming from that module.

Thinking about it, a switching regulator feeds a pulse waveform into an LC tank circuit which converts the pulse waveform into a triangular waveform (the ripple).  The feedback is taken from the capacitor.  There must always be that triangular ripple on the output unless further filtering is added.  The Ebay module has the output connected directly to the tank circuit so external filtering is needed if low ripple is required.

Russell.

[ez24]

The Dell PSU gives very little ripple on no load but when loaded the ripple shoots up to about 100 mV PP.  However it is switching at 200 kHz.  The ripple I'm getting on the Ebay module output is 300 kHz so is coming from that module.

Just curious - on the output of the buck is there any 200 KHz from the Dell?  Also what is the PP of the ripple from the buck, you say it is 100 mV PP on the input to the buck?

[rjeberhardt]

Just curious - on the output of the buck is there any 200 KHz from the Dell?  Also what is the PP of the ripple from the buck, you say it is 100 mV PP on the input to the buck?

O.K. here is the output of the buck converter on 10 V output with a 10 ohm load and with no additional filtering:

The triangular wave is 40 mV pp at 277 kHz.  (Ignore the frequency counter it is triggering on the switching spikes).  There is no sign of the ripple from the Dell PSU getting through.   The ripple changes very little with the load.

Russell

[ez24]

Just curious - on the output of the buck is there any 200 KHz from the Dell?  Also what is the PP of the ripple from the buck, you say it is 100 mV PP on the input to the buck?

O.K. here is the output of the buck converter on 10 V output with a 10 ohm load and with no additional filtering:

The triangular wave is 40 mV pp at 277 kHz.  (Ignore the frequency counter it is triggering on the switching spikes).  There is no sign of the ripple from the Dell PSU getting through.   The ripple changes very little with the load.

Russell

thanks

[Siwastaja]

Do the three outputs need to be isolated from each other, or can they share common ground?