Build your own charger

ElectricGod said:
Don't worry about all the specs for each part. All you need to do is identify what parts are the mosfets and ignore the others. If this is getting overwhelming for you...and I do understand....especially if electronics isn't your thing...post pictures of each part and I'll help you figure it out. I do this all the time.

amberwolf said:
Now the wire will be almost 3x thicker, even though it's in 3 "individual" wires, because they are all in parallel.

I already have 3 wires in parallel on toroid. Think i will leave it like this.

I have question about mosfets.
First "big one" is TO-247 - its rated to 125V 190A. Can i use some similar mosfet here? I cant find IRFP4110 here in Serbia.
I have IRFP4310Z (100V 120A).
Same for TO-220 - can i use lower voltage/amperage mosfets here? Not sure how this works. I have TO-220 IRFB4110 100V 180A so i guess its good enough because original chinese is rated to 80V 150A.

Can i use TO-220 instead of TO-247?
Will add active cooling of course.

Thanks!
 
depening on the cooling you can, but the 247 is a much bigger package to prevent heat spots and burning up the mosfet. if you are going to use a 220 package then make sure it has awesome RDSon otherwise it will burn up real fast.
 
flippy said:
depening on the cooling you can, but the 247 is a much bigger package to prevent heat spots and burning up the mosfet. if you are going to use a 220 package then make sure it has awesome RDSon otherwise it will burn up real fast.

Ok thanks. I found 247 but its not rated like original chinese. Its 100V 180A with lower rds. Guess will try with this one.
 
lower rdson is good.
 
flippy said:
lower rdson is good.

Can i use lower voltage/amperage mosfet here? Chinese is rated to 125V 190A. What factors are important here?
Why so big voltage/amperage? My charger wont use voltage higher than 40V and amperage higher than 20A.

Electronics is not my thing - i can fix something but i dont understand lot of things.

Thanks!
 
that is fine. i recon that chinese mosfet aint gonna do that voltage anyways.

those ratings is what they can do with "perfect" cooling keeping it at 25c or so. totally unrelalistic so when it gets hotter you need to downrate the mosfet. they will do a fraction of the rated specs with air cooling on a simple plate.
 
flippy said:
that is fine. i recon that chinese mosfet aint gonna do that voltage anyways.

those ratings is what they can do with "perfect" cooling keeping it at 25c or so. totally unrelalistic so when it gets hotter you need to downrate the mosfet. they will do a fraction of the rated specs with air cooling on a simple plate.

Ok thanks for your help!
 
krlenjuska said:
I have similar DC-DC converter and i have plan to make 24v 5-6A charger. Do you recommend mosfet replacing or i can leave default ones because my charger will use only 150-200W ?
You do not need any modification on those converters in this case. I personally would not go over 75% of rated specs for extended use but they worked fine at 100% while short test runs. If some component fails, then replace it.
 
Hi folks,
I've made several of these for myself or made them for others too. I've never bothered to document the entire process so that somebody else can make one. since I'm making myself an 82v charger, I thought it was time to document the entire process.

These are 20v PSU's so not ideal for anything other than 20v increments. In a later post in this thread, I'll show how to adjust the Lenovo PSU's so you can set the voltage to what you need. For now, 2 PSU's=40v, 3=60v, 4=80v and so on in 20v increments...ish. Typically these 170 watt Lenovo PSU's are not dead on 20v. They tend towards 20.5 -20.7v each.

Look on ebay or at the local computer recycle center and you will find tons of these 170 watt PSU's for $13-20 each. I paid $15.50 each for 6 of them on ebay.

Dell 170 watt PSU's will work too, but they typically cost more. There's a zillion of these older Lenovo PSU's "out there" so why pay extra? I've put 6 of these PSU's in series without issues. I bet you could do 10 or 15.

NOTE: PSU's with isolated outputs in series is just like batteries in series, except it's PSU outputs. Plus from PSU 1 connects to minus on PSU 2. Plus on PSU 2 connects to minus on PSU 3 and so on.

4 big things you are looking for:
1. The outputs are 100% isolated from the AC input. Without this, STOP NOW!
2. 170 watts is 8.5 amps at 20v. Expect to get more like 9-9.5 amps.
3. These specific PSU's are over load and over temperature protected. They won't die easily.
4. Buy legit Lenovo PSU's, not the Chinese knock-off PSU's.

So then. let's build a charger that will cost you virtually nothing, fanless and will deliver about 9 amps. Later on once it's built, if the voltage isn't quite right, that's alright, I'll take one of them apart and show you what to change to adjust the voltages. If you are impatient and want to do it now, here;s a good article on the process.

https://hackaday.io/project/3469-modifying-a-notebook-power-supply

Parts needed:
1. Legit Lenovo 170 watt PSU's complete with AC chords.
2. Solder, soldering iron, wire strippers, heat shrink, zip ties, some wire or chord
3. Cheap watt meter
4. Connector for your charging port.

Here's the PSU's:
***Notice the beveled top. Notice the single corner knocked off. Notice the Lenovo label on the bottom. Don't get fooled by some Chinese clone!
https://www.ebay.com/itm/Lenovo-170W-20V-8-5A-Laptop-Charger-AC-Adapter-Power-Supply-45N0113-45N0114/382856199240?epid=1719485080&hash=item592401fc48:g:dMcAAOSwFlNceUrS

This 20A watt meter will do the job for this charger.
https://www.ebay.com/itm/Digital-Multimeter-Charge-Discharge-Battery-Tester-DC-0-90V-0-20A-Volt-Amp-Meter/292938370933?hash=item44347ccf75:g:gLwAAOSwikBcUB-T

Let's get building!

1. First off, be sure everything is unplugged from AC power.

2. I'm doing an 82v charger so you'll see 4 PSU's and so on. It's the same process for 3 PSU's are 10 PSU's.

3. Right near the laptop connector is that large plastic bulge. There's a ferox core in there and the output chord is wrapped around it several times, to reduce noise. We don't care for a charger. Lay all your PSU's stacked together, stretch out the output chords and cut them all the same length right behind that ferox core. That will leave you with something like a 3" section with a connector on one end and the plastic bulge on the other. Keep, toss, whatever...you won't use it again for the charger. I pulled all of mine tight and then cut at the length closest to the bulges. As you can see they weren't all the same length.

PSU%20series%204.jpg


4. Pull off the velcro straps. You probably won't reuse but maybe one of them.

5. Strip back the outer insulation on the ends of all PSU's about 5/8". The loose strands are ground and the white wire is +20v.

PSU%20series%202.jpg


6. Strip back the insulation on the inner wire about 1/4" and twist together the loose strands on all exposed wires. Repeat for all PSU's.

PSU%20series%203.jpg


Note to people that want to take short cuts:
If you look at any of my threads, you will see that I pre-tin almost everything before soldering it to something else. This serves several purposes:

A. Stranded wire often times does NOT flow solder particularly well. A dip in some flux and then tinning the exposed wire end means good solder flow into all the strands.

B. Bare copper, steel and other metals that can be soldered to oxidize. Solder flux inside the solder is often times too weak to cut through the oxide layer to get down to bare metal. Prepairing bare metal in advance of soldering something to it makes the solder joint stronger and more reliable.

C. Below you will see where I solder multiple wires together. In one case 5 wires. Holding all those wires together, getting solder to flow into the many strands and NOT make a mess of things is not easy. You are likely to get poor solder flow, burnt fingers, a lumpy connection and melted insulation on the wires. Pretinning everything in advance means, touch the soldering iron to the joint, add a little solder and you are done.

D. Stranded wire is made of many thin strands. Tinning the end of the wire binds all those loose strands together.

E. Thoroughly cover bare wires and connections so that shorts can't happen. It's your build, why skimp on details that cost you reliability and safety?

OK...back to charger building...

7. Tin the ends of the wires. A container of flux is really useful so you get enough flux into the wire to get the solder to flow. All you need is the last 1/4" of each end tinned. Solder has flowed in between all the strands binding them together. Repeat for all the exposed wire ends.

PSU%20series%205.jpg


8. Pull out your multi-meter. Plug in each PSU one at a time to AC and check the output voltage across the 2 wires. I'm reading around 20.5v for all of them. 20.5 x 3 is 61.5v, 20.5 x 4 = 82v. Don't let those bare wire ends short together. There's 10 amps there!

Side line: I got ahead of myself and made a mistake which I'll present here. I soldered the outputs together in series and didn't think about properly insulating the bare ground wires. This is power outputs from 4 PSU's in series. The grounds from each output must not touch each other or else they create a short across at least 2 PSU's.

PSU%20series%206.jpg


See those exposed ground wires even AFTER heat shrinking? This is a short waiting to happen!

2019-08-05%2023_41_45-PSU%20series%207.JPG%20-%20Windows%20Photo%20Viewer.png


9. Cut 3/4" lengths of heat shrink that will slide over the outer insulation of each PSU wire. You will use them later. Slide them on for now.

NOTE: I use heat shrink. It is highly reliable, easy to use and cheap. This could be done with electrical tape to get the same results, but I'm not fond of tape for things like this and it gets gummy after a while. Yuck! Heat shrink is better in every way.

PSU%20series%208.jpg


10. Slide heat shrink over the bare ground wires as far down as you can get it. Leave a little less than 1/4" exposed end. As you can see there is still some exposed wire here that could short to the other exposed wires.

PSU%20series%209.jpg


11. Solder your wires together like this to make the series connections between PSUs. Now slide that heat shrink over the exposed ends and shrink it in place. This should have been 2 steps, but I forgot to take a picture before shrinking over the bases of the wires. AS you can see, those bits of exposed ground wires seen in step 10 are now covered and there is no chance of a short happening.

PSU%20series%2010.jpg


12. Put more lengths of heat shrink over the exposed solder connections.

PSU%20series%2011.jpg


13. Pull them all together and put a zip tie around everything so the connections can't move. This zip tie will get cut off later. It's here to hold everything together until later on so the heat shrink doesn't get stretched or solder connections stressed.

PSU%20series%2012.jpg


14. Your PSU's came with AC chords. However many PSU's you have in series, you need that many AC chords and that many AC outlets. I need 4 of each. We are about to test out our charger for the first time. Grab your multi-meter and a set of alligator clip test leads. Clamp them on the 2 wire ends and set your DMM to the correct setting. ***Make sure those bare ends do NOT touch.*** Plug in any 1 PSU and you should read 20v, plug in another and you'll read 40v, plug in a third and read 60v and so on. If you don't have alligator clip leads, just plug in all the PSU's and check the output voltage. You should see something like 20v x the number of PSU's. In my case, 82.3 volts...close enough to 82v to not bother with.

PSU%20series%2013.jpg


PSU%20series%2014.jpg


PSU%20series%2015.jpg


15. Unplug all AC chords from the PSU's and the AC outlets. Time to make a single AC chord.

16. It doesn't really matter which end of each chord you line up, just use the same end on all your chords. As you can see I have chords of various lengths.

AC%20chord%201.jpg


17. Line up all the connectors that plug into the PSU's. If yours are like mine, that's various lengths. You want to cut off all 4 wall plugs leaving a long length of chord with a PSU connector on each end. You also want to leave a fairly long pigtail so all 4 chords can be soldered to it. My longest chord was 8" longer than the shortest one. The 4 chords with a PSU connector on them are all the length of the shortest chord right at the wall plug...aka all the same length or about 18" long. That left me with a single 8" long AC plug "pigtail" and 3 shorter ones.

AC%20chord%202.jpg


18. Strip back 1/2" of insulation from all the wires and twist their strands together.

AC%20chord%203.jpg


19. For ungrounded chords like these, it doesn't matter, but I'll do it anyway. On your 5 pieces of chord you will notice one conductor has a ribbed insulator and the other does not.

AC%20chord%204.jpg


AC%20chord%205.jpg


20. Take all 4 chords with PSU connectors on them, find the 4 wires with ribbed insulation. Twist them together and put a zip tie around them. Repeat for the 4 un-ribbed wires. Those zip ties will get cut off in a minute, but for now they are making your life easy by holding all 8 of those wires together in 2 sets.

AC%20chord%206.jpg


AC%20chord%207.jpg


21. Tin the 2 bundles of wires.

AC%20chord%208.jpg


22. This is the end of my 8" pigtail with an outlet plug on it. I've tinned the ends of those wires too and then cut them a bit shorter.

AC%20chord%209.jpg


23. Cut 3 sections of heat shrink. The 2 smaller ones just fit over the 4 wires from step 21. THey extend over both sides of each connection by about 1/2". The larger section looks like it will fit over top of all 8 wires plus added heat shrink. It extends beyond the inner heat shrink by at least 1/2" over each end.

AC%20chord%2010.jpg


24. This is the pigtail with the 3 sections of heat shrink on it BEFORE soldering the pigtail to the wire bundles. Slide the big piece on first over both wires and then the smaller sections over one wire each on the 8" pigtail.

AC%20chord%2011.jpg


25. Solder the AC outlet pigtail to the 2 bundles of 4 wires each. The pigtail wires are smooth and ribbed. I soldered the ribbed wire to the ribbed bundle and the smooth wire to the smooth bundle. There is no need to do more for this connection since the joints won't see vibration or much stress like an EV would see. A decent amount of solder bonding the 5 wires together is sufficient for many years of use. You can remove those 2 zip ties now. They were holding your wires together until this step was done. Trust me...soldering 5 wires together while holding all of them together is MUCH harder than wasting 2 zip ties! Pre-tinning everything before trying to solder them together makes these solder connections much easier too.

AC%20chord%2012.jpg


26. Slide the heat shrink over the solder connections and shrink it in place. As you can see, the solder connection has a good bit of heat shrink on either side of it. I wish this was 3X shrink so it closed up around the pigtail wires better.

AC%20chord%2013.jpg


27. Slide that larger section of heat shrink over all 8 wires and shrink it. You want this piece to completely cover the underlying heat shrink on both ends.

AC%20chord%2014.jpg


28. Zip tie all 8 wires together to keep them from stretching the heat shrink later. This joint will not fail easily and getting shocked here by 110v AC is not likely. The 4 chords going to the PSU's can't pull apart easily thanks to the zip tie. This is pretty secure!

AC%20chord%2016.jpg


29. Congrats, you have just made a 4 way AC chord. I've made several of these for the laptop PSU chargers I've done. This chord will work for a very long time and not overheat. At most it sees is maybe 10 amps.

AC%20chord%2015.jpg


30. I doubt it happened, but just for kicks. lets test that 4 way chord. Without it plugged in, put your multimeter on ohms and the test leads across the wall plug prongs. If you measure anything other than infinite resistance, you did something wrong.

31. Now plug in your 4 way chord into all your PSU's and then into the wall. You should measure 20v x the PSU number.


More to come later! My watt meter is on order and won't get here for a few weeks. I'll finish up this charger when that happens.
 
Hi folks, the watt meter arrived a couple of days ago so my 82v charger is done now. The next post about adjusting the voltage is still coming.

These PSU's do get warm so letting them dump heat is important. The best way to put them together is side by side so the heat sinks under the flat plastic sides can dump heat the easiest.

PSU%20charger%201.jpg


We'll see how well this holds or doesn't. Hot glue gets soft as it warms up. The PSU's will get warm and then my hot glue joints may not hold up. For now however, they are working great. JBweld is what I've used on previous chargers like this and that works really well. This hot glue is an experiment...fingers crossed.

PSU%20charger%202.jpg


These are 14 awg...a bit overkill, but it's fine. The black wire is about 5" long and the red 12" long. The black wire goes to the watt meter, while the red wire goes direct to the charge connector. AS you can see, it's all covered in heat shrink so shorts don't happen.

PSU%20charger%203.jpg


In the above image, you can see I had a zip tie on the wires from the PSU's. I cut that off and then slid larger heat shrink over everything, shrunk it down and secured the ends with zip ties. This keeps movement and stress on connections down to a minimum. This ought to last many years. Electrical tape wont do as well.

2019-08-23%2015_07_44-PSU%20charger%2013.JPG%20-%20Windows%20Photo%20Viewer.png


The 5" black wire got a screw lug soldered to it and then screwed down at "input". This specific watt meter reads current flow in both directions so there's not really an input or output side persay other than that you get positive current flow from input to output and negative flow output to input. The second black wire is the same length as the red wire so their ends can connect to the connector. Notice the bare section on the red wire. This lets me solder a small wire here for connecting to VCC+ on the meter. You don't need Vext (4th screw lug) at all.

PSU%20charger%204.jpg


Protect those connections! Positive power to the meter is ready to go.

PSU%20charger%205.jpg


Once it was all connected, securing the wires together with zip ties is a good idea. The screw terminals are pretty resilient and work well for securing the wires so they can't move and break.

PSU%20charger%206.jpg


Ends of my charger wires are tinned and the same length.

PSU%20charger%207.jpg


Like everywhere else, heat shrink is your friend. These 2 pieces will insulate the solder joints to the XT90 connector.

PSU%20charger%208.jpg


PSU%20charger%209.jpg


A few pics of the watt meter. I've compared the current reading with another panel meter and with my $200 clamp on meter and they are all within 1/10 amp of each other...good enough. That list of stuff on the right side of the LCD are options you can set so the meter works within a specific voltage range or Ah range.

PSU%20charger%2010.jpg


PSU%20charger%2012.jpg


My completed charger.

PSU%20charger%2013.jpg
 
riverie said:
This is very informative and helpful EG. Keep 'em coming...you know I need them :D

More coming...waiting on the watt meter and time to do the PSU modding...
 
Started a charge on a 20S pack tonight. The pack is 32Ah so at 75 volts it will absolutely suck down every dribble of current this PSU charger can deliver. If I don't limit my 30 amp charger, this pack will overload that charger so it shuts down to protect itself without current limiting.

The PSU's don't want to exceed 9 amps so they throttle back to 9ish amps, but peak at 10 amps.

PSU%20charger%20amperage%201.jpg


PSU%20charger%20amperage%202.jpg


PSU%20charger%20amperage%203.jpg


This is an acceptable margin of error: .07 amps. $23 watt meter vs $140 clamp on meter...not bad!

P1040349.jpg
 
I checked the PSU's with my laser temp meter. Its now 30 minutes into the charge. Three are at 122F and the 4th at 124F. I might need to add a current limiter so they stay a bit cooler. That's just a section of nichrome wire to add .2 or .3 ohms resistance. On my last PSU charger build the shunt in the watt meter was the exact resistance needed to keep the PSUs from getting too warm. They are rated for 8.5 amps and I've obviously exceeded that a good bit...proof of the quality of the PSU's! They have thermal protection so they will shut down after they reach some temperature threshold.
 
ElectricGod said:
"4 big things you are looking for:
1. The outputs are 100% isolated from the AC input. Without this, STOP NOW!
2. 170 watts is 8.5 amps at 20v. Expect to get more like 9-9.5 amps.
3. These specific PSU's are over load and over temperature protected. They won't die easily.
4. Buy legit Lenovo PSU's, not the Chinese knock-off PSU's."

..................
Two questions . First, when you are charging with an LED CV/CC supply , if you set the voltage to 58.8 and whatever current , 6 amps , and you didn't turn it off , would it continue to push 6 amps until it went up in flames? Second question , When you say " The outputs are 100% isolated from the AC input , what exactly does that mean ?
 
Hot glue on the PSU charger...
I did say it was an experiment.
Just don't, the glue warms up, gets soft and no longer holds.
Use JB weld, it will hold and stay put even when on a hot engine block!
 
Dak77 said:
if you set the voltage to 58.8 and whatever current , 6 amps , and you didn't turn it off , would it continue to push 6 amps until it went up in flames?
A power source only makes X amps available.

It is the load / battery that demand / pulls the current level it wants.

Overload protection, or even better adjustable current limiting is what protects the source device, and also the battery from too high a C-rate.

_____
> Second question , When you say " The outputs are 100% isolated from the AC input , what exactly does that mean ?

----
It means there is no connection electrically between the primary and secondary circuits. Separate common/return (ground or floating).

Not sure if this is required for parallel stacking of outputs, but definitely is when charger / supplies are connected in series.
 
john61ct said:
Dak77 said:
if you set the voltage to 58.8 and whatever current , 6 amps , and you didn't turn it off , would it continue to push 6 amps until it went up in flames?
A power source only makes X amps available.

It is the load / battery that demand / pulls the current level it wants.

Overload protection, or even better adjustable current limiting is what protects the source device, and also the battery from too high a C-rate.

_____
> Second question , When you say " The outputs are 100% isolated from the AC input , what exactly does that mean ?

----
It means there is no connection electrically between the primary and secondary circuits. Separate common/return (ground or floating).

Not sure if this is required for parallel stacking of outputs, but definitely is when charger / supplies are connected in series.

Ohhh ok. So no grounding the input to the same case that the output is grounded to . Thanks for clearing that up for me. So if I had 2 meanwell led supplies to connect in series, does that mean I would have to find outlets that aren't on the same breaker?
 
That would be to work around available current limitations.

Would not help, since the whole building supply likely shares a common (actual Earth) ground.

Floating return / common on DC output is one option.

Only one of the PSUs is chassis (case common) "grounded", the rest are isolated in that sense.

Lots of HowTo threads here and lots more on RCgroups as well.
 
Dak77 said:
Two questions . First, when you are charging with an LED CV/CC supply , if you set the voltage to 58.8 and whatever current , 6 amps , and you didn't turn it off , would it continue to push 6 amps until it went up in flames? Second question , When you say " The outputs are 100% isolated from the AC input , what exactly does that mean ?

1. No, by definition...CC means it delivers a set amount of current. If the PSU is capable of 6 amps as it upper limit and you set it to deliver 6 amps, it will deliver 6 amps for a very long time. Try setting that CC PSU to 7 amps and it might burn up.


2. isolated Disconnected from AC power, set your DMM on ohms. You have 2 output wires or screws and 2 or 3 input wires or screws. On ohms, put your probe leads on either output and the other lead on any input wire or screw. Try all possible combinations of input to output options. Do you measure a resistance anywhere less than a few megohms? If you do, it's not isolated.
 
Dak77 said:
Ohhh ok. So no grounding the input to the same case that the output is grounded to . Thanks for clearing that up for me. So if I had 2 meanwell led supplies to connect in series, does that mean I would have to find outlets that aren't on the same breaker?

By definition, ground literally means the dirt under your feet with a long copper bar driven 6 or 8 feet into it...AKA the term "earth ground". Every electrical panel in a modern house has a long copper stake somewhere nearby with a big copper wire going to the breaker panel for earth ground. This gives you a specific voltage potential relative the planet. The ground pin on your AC chord connects to earth ground. More than likely if your PSU has an AC ground connection, then the shell of your PSU is also connected to earth ground.

The entire breaker panel and all circuits inside and all AC wiring in the entire building had better ALL have the exact same ground plane back at the panel that connects to earth ground. If this is not the case, you have what is called a floating ground. For a lot of things you don't care. A single lamp is not electrically connected to a lamp in another room on another circuit except back at the breaker panel.

I worked as a sysadmin for 28 years. There was a customer that I wired for ethernet around 2001. They had some amateur wire the building for AC and he screwed it up quite a lot. They had never had a situation where multiple offices in the building could be electrically connected together. I kept blowing network cards, ethernet switches and mother boards in PC's. For a good while I had no idea why until I got the bright idea to connect a wire to a ground pin in one outlet and then take the other end of that wire to the adjacent office. I then measured 68v AC from one ground prong to the other between 2 adjacent offices. No wonder I was blowing up computer parts! There was no reliable earth ground! I found this situation in several places around the office.

So then on the output side of the PSU...assuming it is fully isolated means that negative connection is NOT ground. It's just the negative pole of the power supply output. If it was actually ground, then you could NOT put multiple PSU's in series. Isolated PSU's means a negative and positive output pole that are NOT in any way electrically connected to the input source. Without this you can't connect those poles in series with PSU poles on another PSU.

If you look at my Lenovoe PSU charger build all 4 AC inputs are 100% in parallel and plug into a single outlet. AC inputs are in parallel. Isolated DC outputs are in series.
 
john61ct said:
Only one of the PSUs is chassis (case common) "grounded", the rest are isolated in that sense.

The shell or case of all the PSU's can be at earth ground potential and NOT effect the isolated output of each PSU. If you look back in this thread a little, you'll see that I have 8 meanwell PSU's in physical metal shell to metal shell contact. The shells are all at ground potential. The maximum output for all 8 PSU's is 164v.
 
I wish I had thought to check under the rubber feet first. I assumed the shell was glued together and I needed to break the glue at the shell halves to get inside. I pried at the shell in several places and pretty well ruined the plastic shell when I noticed that the shell moved everywhere except at the corners This indicated to me that there were hidden fasteners only there. Some nice person decided to go for serviceability instead of cost and used screws hidden under the rubber feet to hold it all together. All you need is a security torx bit. The shell fell into two halves easily with the 4 screws removed.

Lenovo%20PSU%20teardown%201.jpg


Lenovo%20PSU%20teardown%202.jpg


First look under the plastic covers. These are thin aluminum sheet, They will transfer and radiate heat fairly well.

Lenovo%20PSU%20teardown%203.jpg


The two posts in this aluminum cover will later on need to be unsoldered from the board and it takes a good bit of heat to the solder to flow.

Lenovo%20PSU%20teardown%204.jpg


There's catches on either side hold the top and bottom aluminum sheets together. Pop them loose and the top sheet comes right off.

Lenovo%20PSU%20teardown%205.jpg


Lenovo%20PSU%20teardown%206.jpg


Under the top sheet is another sheet of aluminum with 2 welds in it.

Lenovo%20PSU%20teardown%207.jpg


The outer bottom sheet is stuck down by 2 strips of double sided tape and 2 solder posts. A little prying under the sheet and the glue gave way pretty easily. Getting the solder to melt on the posts took a good bit of heat.

Lenovo%20PSU%20teardown%208.jpg


This plastic shield was under the bottom aluminum sheet. It took a bit more effort to get it loose as it was held down by many spots of thermal glue. A plastic or wood tongue depressor would get it loose and not damage the electronics under it.

Lenovo%20PSU%20teardown%209.jpg


Lenovo%20PSU%20teardown%2011.jpg


Finally we get to see the bottom of the PSU board. I seriously doubt using a dremel to cut a hole in the shell to get access to the board is the right option. Taking it all apart is the only way to go. There's lots of aluminum layers here and you will only get metal bits on the components under them.

Lenovo%20PSU%20teardown%2010.jpg


Using a plastic stick, I cleaned off all the thermal glue blobs and carefully got it out from around the small parts. An old tooth brush is a great tool for scrubbing off the remaining bits and crumbles. You see two thermal pads here. The bigger one is under the main isolation transformer. The small one is under a large inductor in the high voltage DC section. Upper edge, center of the board is an SMD capacitor. To the left of it is the temp sensor.

Lenovo%20PSU%20teardown%2012.jpg


This is the "brains" of the PSU a DAP019AT. Once I look up the data sheet for this part, I'll know where the adjustment circuit is specifically. It's right here around this IC. I can probably figure it out without reading about the IC. It's more than likely, just 2 resistors creating a voltage divider.

Lenovo%20PSU%20teardown%2013.jpg


Lenovo%20PSU%20teardown%2024.jpg


I don't know what this IC does yet. Maybe it's some kind of mosfet driver or op-amp comparitor? It's labeled DA505. I bet it's a driver that monitors the current overload circuit.

Lenovo%20PSU%20teardown%2014.jpg


Notice next to the IC, the small plastic part around a single solder connection. No idea what that's protecting or why. This solder connection is holding a heat sink to the board and nothing more.

Lenovo%20PSU%20teardown%2015.jpg


Back to the top of the PSU. Those two welds broke loose fairly easily with a little prying at them. The aluminum is quite thin so getting it flat again was easy and the welds were quite weak too.

NOTE: Access to the top of the PSU is NOT needed at all. Since I'm in here taking things apart, I kept going. Access to the bottom of the board is all you need to adjust the output voltage.

Lenovo%20PSU%20teardown%2016.jpg


The heat sink it was welded to.

Lenovo%20PSU%20teardown%2017.jpg


That heat sink is stuck down by it's entire underside to a large electrolytic cap, several small ceramic caps and 3 inductors. All those components are also glued together with thermal glue. It won't come loose easily and I don't really need to get under it anyway so I'll leave it alone. At this point looking deeper into the PSU is just curiosity.

Lenovo%20PSU%20teardown%2018.jpg


I knew this was in here somewhere. It's an opto-isolater and the ONLY connection from the high voltage DC side to the low voltage DC side. AND it's electrically isolated. All signals happen via light transfer. No opto...NOT fully isolated. Of course this is just one element of output isolation.

Lenovo%20PSU%20teardown%2019.jpg


Close-up of the output power switching section.

Lenovo%20PSU%20teardown%2020.jpg


High frequency rectifier diodes off the output of the isolation transformer to make the AC back into DC again. Then it is filtered through 3 electrolytic caps before it exits to your laptop.

Lenovo%20PSU%20teardown%2021.jpg


This section is for current sensing and overload limiting. Notice the R003 or 3 mOhm shunt. There's a 6 pin IC labeled DAS2 to the left of the shunt. More than likely it is an op-amp comparitor that monitors the voltage across the shunt and compares it with a voltage reference. The mosfets that drive the isolation transformer are then enabled or disabled based on the state of current draw at the shunt. It might happen some what different that this, but that current or over load circuit disables something to protect the PSU.

Lenovo%20PSU%20teardown%2023.jpg


This is the AC input isolation section...which is pretty nice. You have a double choke for supressing high frequency noise on the AC line. Then a couple of MOV's and a ceramic cap for more filtering. Finally after the full wave bridge rectifier you have another inductor for choking any remaining noise. Behind the aluminum is a large electrolytic for final ripple leveling. At this point you have around 150v DC inside the PSU. The next step is the switching circuit that takes that 150v DC and converts it back into AC at something like 10khz before getting stepped down by the isolation transformer.

Lenovo%20PSU%20teardown%2022.jpg
 
Back
Top