120VDC in to appliances and devices?

[lufihengr]

Has anyone tested what happens if you put 120VDC in to a TV that has a switchmode PSU meant for 120-240VAC?

In theory it should work, no?

Asking because I would like to avoid using inverters, and modifying appliances like TV's with integrated power supplies is too time and cost intensive.

Thanks

[AVGresponding]

In theory yes, but bear in mind you're putting one half of the diode bridge on 100% duty cycle. Depending on the design specs, this might be problematical. In the unlikely event it's a half-wave rectified primary side, you'd have to get the polarity right.

[lufihengr]

Would it be better for the appliance to have 120VDC or 240VDC if the PSU is capable of taking 120-240VAC?

And would it technically be even better to have a 100% duty cycle polarity switching DC supply to it? Like have it switch polarity at 10, 50, 60Hz, essentially just a square wave going from +120VDC to -120VDC? Or is that worse than a pure sine wave? It would be easier to make than even a modified sine wave inverter.

[Gyro]

If it's a modern, reasonably large TV, you may need to consider how the PFC circuit would react to DC. I have a vague memory of this being discussed on the forum before, but not the outcome.

[Ian.M]

You are going to run into severe problems with switches and fuses, as they are extremely unlikely to be rated to break a >30V DC circuit.  Squarewave  AC isn't much better, as there is no 'dead' time round its zero crossings.  OTOH sinewave 120V AC has nearly a millisecond under 30V round its zero crossings, making low energy arcs much easier to quench.  Therefore, you'll probably need to replace fuses with DC rated ones and avoid powering with DC any devices with a mechanical or relay mains switch.

Then there's the issue of PFC circuits - some will be very unhappy with any input waveform that isn't somewhat sinusoidal.

On the subject of switching polarity, see https://www.quora.com/What-is-Quasi-Square-wave-or-Modified-Square-wave-Inverters
Even with dead time at zero crossings, and both peak and RMS voltages equal to sinewave mains, many appliances will run hotter or even fail if run from such a supply.

[Psi]

It's usually one of those, try it and see, things.

Sometimes it will work fine, sometimes it will not.
And occasionally you might burn something out, but it's rare.

[Berni]

In most cases switchmode supplies run perfectly fine from high voltage DC. The diodes won't be run that close to the rating to be a problem.

But yeah in some cases PFCs might get unhappy when they expect to see a sine wave, but a lot of them still work.

And as said above the is a small but non zero chance that a PFC might get confused enough to blow up. Also the safety protections like fuses are not designed to break DC, so it is a bit of a safety issue doing this for daily use.

[Circlotron]

On occasion I have used my Rigol DS2072 scope while sitting in the passenger seat of a moving car so I have had it running off an inverter that supplied a "modified sine wave", in other words +/- 340VDC at 50Hz, 2/3 on 1/3 off approx. The scope power supply has an X cap across the mains and it draws a bit of current during the fast transitions of the inverter power so I run it through an external bridge rectifier and feed that to the scope. Works great. Scope PSU does not have a PFC input.

[NiHaoMike]

I suppose one could have a MOSFET/IGBT driven by a 555 to make pulsed DC at 170V or so which will be fine for AC switches.

[pqass]

Has anyone tested what happens if you put 120VDC in to a TV that has a switchmode PSU meant for 120-240VAC?

Some PC power supplies rely on a two capacitor voltage doubler after the diode bridge. 
They won't work with 120VDC input.
But this may only apply to those supplies with a 120/240VAC selection switch. 
http://www.righto.com/2021/05/teardown-of-pc-power-supply.html

[tom66]

I've tried this to some extent with a few devices.

Rigol scope - works fine 50V upwards.  Draws 10W, so unlikely to wear the diodes even at the lowest input voltage.  Below 40V, once started, scope shuts down suddenly with no apparent side effects.  Current draw is minimal until startup works.

LCD TV - works ok up to 100V DC (had a PFC circuit) but below this power supply makes some very unhappy sounds and backlight/sound flutter badly.  Eventually LCD TCON lost its mind and started drawing all sorts of random patterns on the display in place of video, whilst sound continued OK.  Did not break anything.

various AC-DC PSUs - all seemed to work OK

Dell laptop PSU - did NOT work correctly, output pulsed on and off, laptop refused to charge.

[amyk]

120VAC is RMS so you should probably be aiming for ~170VDC instead (or 340 for 240VAC).

[Zero999]

120VAC is RMS so you should probably be aiming for ~170VDC instead (or 340 for 240VAC).

No, the switched mode power supply will be designed to work at a much lower steady DC volage, because the smoothing capacitor will discharge, during the zero crossings.

My Rigol DS1054Z oscilloscope works perfectly down to 36V. I even considered making a battery pack for it, but lost interest because I bought another oscilloscope with a rechargable battery.
https://www.eevblog.com/forum/testgear/battery-pack-for-rigol-ds1054z/msg717702/#msg717702

I'm not saying this will be the case for all switched mode power supplies, but I would expect it to work from a lower DC voltage, than the peak mains. As mentioned above, there are potential pitfalls, to running AC switched mode supplies off DC, but I've found most mains switched mode power supplies work perfectly well.

[lufihengr]

As a followup then, is there a point where it would just be easier/better to use DC power from batteries to spin a DC motor, that is connected to a generator? That would at least be a perfect sine wave

As you have probably gathered, the situation is a case where grid connection is not available, but the use of regular appliances is wanted.

[BeBuLamar]

I work on a machine which has a PC used as the operator interface. They power the standard power supply with 220VDC and it works. I really don't know why they do it that way as they could feed the 220VAC directly to the power supply but they use a bridge rectifer and a large capacitor to create the 220VDC (actually it's more than 220VDC) to run the computer.

[AVGresponding]

As a followup then, is there a point where it would just be easier/better to use DC power from batteries to spin a DC motor, that is connected to a generator? That would at least be a perfect sine wave

As you have probably gathered, the situation is a case where grid connection is not available, but the use of regular appliances is wanted.

Depending on funds, a better solution might well be solar PV/a wind turbine and a decent inverter, or indeed a generator.

[Zero999]

As a followup then, is there a point where it would just be easier/better to use DC power from batteries to spin a DC motor, that is connected to a generator? That would at least be a perfect sine wave

As you have probably gathered, the situation is a case where grid connection is not available, but the use of regular appliances is wanted.

That's known as a motor-generator set and is the old fashioned way of doing it. Quite often the primary and secondary windings were on one armature, in which case it was known as a rotary converter. Imaging a DC motor, with three poles on the armature, each with a secondary winding, connected to a slip ring. The DC going to the commutator will be converted to three phase AC.

The disadvantages of motor generator sets are: low efficiency, more noise and higher maintenance, as bearings and brushes need to be replaced. The only advantages are they're more electrically robust and can withstand ionising radiation better than electronics.

Nowadays static, semiconductor inverters are used. A boost converter, takes the battery voltage up to the peak voltage of the AC output and an H-bridge converts it to AC. Cheap inverters produce a modified sine wave, whilst more expensive ones use PWM and a filter to give a pure sine wave. The only downsides are they're more vulnerable to over-voltage and ionising radiation, compared to motor generator sets.

If all you want to do is convert one DC voltage to another, use a DC to DC converter.

[Siwastaja]

I have ran random switch-mode stuff out of DC and usually it just works. BUT:

The diodes won't be run that close to the rating to be a problem.

That's quite an assumption. Cheap consumer electronics is often dimensioned close to destruction. Diodes specifically are annoying to design in because they dissipate surprisingly lot of power for quite small packages. Suddenly doubling (+100%, not something like +10%) the average power dissipation in a component is something I would not classify as "definitely no problem". I would keep an eye on it, or just bypass the bridge.

[Zero999]

I think assuming the diodes are over specified is sensible with small power supplies, given 400V parts normally have current ratings of at least 1A.

Does it really increase the power dissipation in the diodes that much?

Don't forget that although the diodes only run at 50% duty when on AC, the peak currents are much higher, in a non power factor corrected design, whilst with DC it's more constant. I suspect the it's not that bigger difference.

[langwadt]

I have ran random switch-mode stuff out of DC and usually it just works. BUT:

The diodes won't be run that close to the rating to be a problem.

That's quite an assumption. Cheap consumer electronics is often dimensioned close to destruction. Diodes specifically are annoying to design in because they dissipate surprisingly lot of power for quite small packages. Suddenly doubling (+100%, not something like +10%) the average power dissipation in a component is something I would not classify as "definitely no problem". I would keep an eye on it, or just bypass the bridge.

but with ac the diodes are carry a higher peak current at the peaks of the period, with DC the diodes carry current all the time but it must be lower 

[Zero999]

I have ran random switch-mode stuff out of DC and usually it just works. BUT:

The diodes won't be run that close to the rating to be a problem.

That's quite an assumption. Cheap consumer electronics is often dimensioned close to destruction. Diodes specifically are annoying to design in because they dissipate surprisingly lot of power for quite small packages. Suddenly doubling (+100%, not something like +10%) the average power dissipation in a component is something I would not classify as "definitely no problem". I would keep an eye on it, or just bypass the bridge.

but with ac the diodes are carry a higher peak current at the peaks of the period, with DC the diodes carry current all the time but it must be lower

I just said that.

Don't forget though that with DC, only two of the diodes in the bridge are conducting continuously, whist with AC, all four are used, with each pair conducting alternately.

I suspect the losses are over all lower with DC, probably not quite half though. If it's a bridge rectifier module, rather than discrete diodes, it's even better because the connections to the unused diodes will still provide a bit of heat sinking.

[amyk]

I have ran random switch-mode stuff out of DC and usually it just works. BUT:

The diodes won't be run that close to the rating to be a problem.

That's quite an assumption. Cheap consumer electronics is often dimensioned close to destruction. Diodes specifically are annoying to design in because they dissipate surprisingly lot of power for quite small packages. Suddenly doubling (+100%, not something like +10%) the average power dissipation in a component is something I would not classify as "definitely no problem". I would keep an eye on it, or just bypass the bridge.

I've seen this in cheaper PC PSUs that are based on a voltage doubler, where in 120V mode two of the diodes in the bridge rectifier are not used, so to save the miniscule amount of extra cost of two higher-rated diodes, they use two bigger and two smaller diodes instead of a single bridge rectifier.

[Berni]

The diodes are on the AC side where the continuous current is pretty low due to the high voltage. These diodes also have to survive the inrush current of the capacitors filling up if potentially plugged in at the peak of the sine wave. So they have to be a reasonably robust power diode rather than some tiny fragile signal diode.

Fair point on the capacitor doubler input stages. I live in the 230V land so i don't come across those (apart from perhaps the 110/220 switchable PSUs). Those would definitely not work with HV DC on the input. But id think those are pretty rare these days since most modern switching PSUs seam to be universal input 90 to 250V and those won't have one.