Placing Laptop AC/DC Power Adapters in Series?

[LooseJunkHater]

Simple question; I'm thinking of placing 2x laptop power supplies in series to create ~38v. Will this be okay? Will it possibly pop the output FET's or capacitors? Is there a limit on how many power supplies can be placed into series?

[Andy Chee]

I’d first make sure the bricks are identical in model number. Not all 19V 5A bricks are made the same and any mismatch may have weird regulation feedback issues.

Secondly, the ultimate limit will be insulation voltage breakdown in the secondary winding, so don’t try and stack up dozens to get 1kV!

Thirdly you alluded to popping FETs. Hopefully a reasonable brick would have FETs of sufficient voltage/current rating to stack a handful of bricks. But remember, these things are made to a low price, so again don’t stack a whole bunch together.

Otherwise, I myself have done a similar thing, though with 12V bricks.

YMMV.

[CaptDon]

Make sure the earth ground is not carried through (if the power pack has a ground pin on the line side) to the negative power lead. Some power packs provide an earth ground to the laptop. I would also maybe give each power supply its own hefty filter capacitor such as 4700uf @ 16v or possibly 4700uf @ 25V for better stability and load sharing across the pair.

[ArdWar]

Most brick PSU works fine connected in series. I never found one with grounded output or otherwise non floating output, but it's always good idea to check first.

A (slightly) better approach is to add diodes parallel to the output. It protects the PSU from reverse voltage to their output. Especially during powering up/down or when one is powered while the other isn't.

https://www.cui.com/blog/power-supplies-in-series-or-parallel-for-increased-power#power-supplies-with-outputs-connected-in-series

[LooseJunkHater]

Hmm, I guess one last question; if for example the output cap of the AC/DC power adapter is 35v, shall I be concerned about that, or can that be ignored? Do I only really need to worry about the Vds (drain-source voltage) of the FET + adding diodes on each AC/DC power adapter's output?

[Konkedout]

BIG ISSUE: This can happen due to load capacitance during startup or due to an overcurrent shutdown of any one (or more) series connected units:

The unit(s) that are shut down (or not yet started up) can have reverse voltage applied across their output.  That might cause the power supply to go up in smoke.  No guarantees, but what would probably help is to connect a high current antiparallel schottky diode across the output of each one.  That means cathode to positive and anode to negative of each one.  That way, if the other series units try to reverse the output from the weakest or slowest one, at least the schottky diode should be able to handle most or all of the current.  It would CLAMP the output voltage from that weakest unit and limit any negative excursion.

[ejeffrey]

Hmm, I guess one last question; if for example the output cap of the AC/DC power adapter is 35v, shall I be concerned about that, or can that be ignored? Do I only really need to worry about the Vds (drain-source voltage) of the FET + adding diodes on each AC/DC power adapter's output?

During normal operation the output caps will each only see 20V or whatever.  The only part that should take a larger voltage is the isolation transformer and any EMI filter caps from ground to the output.  But all of those components need to handle 230 VAC.

However the reverse polarity in case of an output short circuit (including the load capacitance that acts like a short at power on) is a real problem, protection diodes are a good idea.

[Whales]

I thought all of the laptop bricks I had encountered were earth referenced (ie mains power earth is connected to output GND).  Maybe I've been mistaken all these years?

Definitely check first with a multimeter!

[ramussons]

Yes, this is very important. The diodes across the outputs.

[ramussons]

I thought all of the laptop bricks I had encountered were earth referenced (ie mains power earth is connected to output GND).  Maybe I've been mistaken all these years?

Definitely check first with a multimeter!

No, the outputs are floating. I have a couple of HP laptops that have only 2 pins for the AC input.

[CaptDon]

I have several power packs that have 3 pin line connectors on the inputs and EVERY one that has a ground connection on the input side carries it through to the laptop negative lead!! Naturally, none of the 2 pin line inputs can carry a ground through. I have a pair of Gateway ROG450 laptops and although the power packs are made by two different companies they both have grounded inputs that carry through. So do my older Panasonic / Lenovo units. My 2 pin line input packs radiate R.F.I. to a much larger degree than the packs with a ground connection. I agree about the diodes being needed across the outputs, sorry I didn't include that in my original post.

[u666sa]

create ~38v.

Why not pick proper resistor for internal voltage regular and get the desired 40 volts? 

[magic]

because it sucks

(and I just came up with another reason why it sucks and added it there)

[Zero999]

create ~38v.

Why not pick proper resistor for internal voltage regular and get the desired 40 volts? 

That could actually work, but it's not trivial to double the voltage. The output capacitor would need to be changed to a higher voltage rating and the series resistor a higher value. It's more feasible in 230V land, than 120V land.

[magic]

Very not trivial. I would worry more about other things than the output capacitor, and it actually is safer with 120V input (clarification: assuming a universal PSU). Unless there is APFC - then line voltage doesn't matter.

[dvas]

I took two 230W HP Power Adaptors that are coming with older docking stations. When connecting one to Bluetti AC200MAX for charging, AC200MAX recognized around 180-190W.
I took two bricks, connecting them in series, one was disconnected from ground (from line side) and connected to Bluetti. The charging was 573W?! Correct voltage, but 15A! Not clear where the extra power is coming from. No warming of any component.

[Zero999]

create ~38v.

Why not pick proper resistor for internal voltage regular and get the desired 40 volts? 

That could actually work, but it's not trivial to double the voltage. The output capacitor would need to be changed to a higher voltage rating and the series resistor a higher value. It's more feasible in 230V land, than 120V land.

Very not trivial. I would worry more about other things than the output capacitor, and it actually is safer with 120V input (clarification: assuming a universal PSU). Unless there is APFC - then line voltage doesn't matter.

It would be worse at 120V, since the duty cycle would have to be higher than it was initially designed for. Heck it might not even work. Doubling the output voltage is safer at 230V because the duty cycle would be the same, as the design output voltage, when the line voltage is 230V.

EDIT:
Quote added, to show I was responding to magic's response to one of my previous posts.

[Jwillis]

Putting two or more Switch Mode Power Supplies in series is fine.
The output capacitors will be in series which increase the voltage rating of the capacitors. Search "Electrolytic capacitors in series".
The output diodes will be put in series which will increase the voltage capability of the diode. Search "Diodes in series".
The secondary of the transformers in series will be no different from and other isolating transformers put in series.

If the one or the other power supply gets switch off or if they are switched on at different rates, this will not hurt to the output capacitors, Electrolytic capacitors can handle as much as 1.5V of reverse polarity. Page 9 https://www.cde.com/resources/technical-papers/AEappGuide.pdf  Even "junk" Made in China capacitors exhibit the same tolerances.
The potential difference between the positive side and negative side of any of the capacitors in series will not be more than the voltage drop of the diode. That won't exceed more than 300 to 700mV.
Putting extra "Safety" diodes between the power supplies will do nothing about the reverse polarity of the capacitors. The capacitor will still see the same potential difference because of the voltage drop of the internal diode in the power supply. All putting external diodes will do is drop the total series voltage. Nothing more.
There are no output Fets on those types of power supplies. Those are Diodes. Even if there were Fets they would be put in series with the other Fets. this would increase the voltage tolerance of the Fets.

The feed back of each power supply will not be affected because it will only see the voltage and current of each individual power supply and not the entire series.

The only time it becomes a problem putting power supplies in series is when  using different power supplies with different current output. The entire series cannot exceed the lowest rated current. For example if you had 3 12v supplies and 2 are rated for 10A and the third was rated for 5A, the output of the series would be 36V at 5A only.

[p.larner]

I dont know if it helps but i have used a lappy brick in series with a homebrew psu to charge a 48v ebike battery,it worked well for about 3 months before i got the proper charger,its ironic as when i got the propper charger,it overcharged and the battery went on fire,it was a chineese charger tho!.