Low-noise isolated DC/DC

On a sunny day (Sun, 9 Aug 2020 09:00:46 +0200) it happened Piotr Wyderski wrote in :

60W Edisonb lightbulb -> solar panel. No noise.

Was discussed here years ago, laser diode into fiber or something..

This is not that silly, TLP190 is actually good at providing quietly microamps of negative/floating bias currents.

Best regards, Piotr

This is a bit so and so. With a standard sun (1000 W/m**2) about 0.5 dm**2 of solar power is required to provide 0.5 W of electricity. A 7 cm x 7 cm panel or 80 mm diameter cell is required.

The question is, can a 60 W incandescent bulb provide the same illumination as the sun on that panel or cell area using concentrating mirrors and lenses ?

The question is further complicated by different spectral distribution of the sun, the solar panel and the incandescent lamp. The sun contains a lot of short wave radiation, while majority of incandescent lamp radiation is in the near IR. The panels quantum efficiency for IR is low.

It might be a good idea to use an array of blue LEDs (as used in "white" LEDs) and use a solar cell that can effectively use photons with higher energy (shorter wavelengths).

These were used e.g. for current transformer hanging on a 400 kV line. The power demand was much lower to send down the measurements.

On a sunny day (Sun, 09 Aug 2020 14:06:37 +0300) it happened snipped-for-privacy@downunder.com wrote in :

Yes, there is a lot online up to 500 W to power drones:

small to 250 mW:

small to 500 mW:

google results after 3 minutes search.

found from

The half-bridge operates at 500 kHz with a dead time of 200 ns. The voltage at its output is trapezoidal. The transistors are switched in ZVS mode with an additional 33-47 ?H i nductance and 1800 pF capacitance in parallel with the switches. Galvanic isolation is carried out by two transformers connected by a wire i n high voltage insulation. Information is exchanged through the IRDA interface.

Yes, it can.

No, it is fine for those IR wavelengths for which the photons have more energy than the bandgap of silicon.

Blue LEDs are not optimal. IR LEDs (~ 940nm or 880nm) are perfect. If the photons had any more energy than that, it would just be wasted in heating the solar cell, and creating higher-energy photons would require more voltage to drive the LEDs.

It may be desirable to run the solar cell at more than 1 sun intensity.

Something I have not tried, but which might be useful is to modulate the LED current with AC, and then connect the solar cell to the primary of a step-up transformer, so that only one solar cell is required rather than a series string of solar cells to generate the required output voltage. With a series string of solar cells, the current will be determined by the least illuminated cell, leading to a requirement for the solar cells to be uniformly illuminated, which might be inconvenient. Having only one solar cell removes this requirement, and may also result in a more compact physical arrangement. A variation would be to use two solar cells electrically connected in anti-parallel, connected to the transformer primary, and illuminate them alternately by using separate LEDs aimed at each one. This would provide push-pull or bipolar drive to the transformer.

I've been thinking about the same issue. I might need to design a very small, fast UV-photodiode-amp-ADC board that runs off +12. I'll need clean pos and neg analog supplies, and power for an ADC and a small FPGA, all in a few cubic inches. Any hard switching supply will spray spikes everywhere. I was doodling a sinewave-based converter or something. Or maybe use a switcher chip but add a passive lowpass filter before the transformer.

It might be easier to use a small commercial dc/dc brick and shield/filter it hard.

An older design just switches the power converters off some microseconds before the light pulse. I guess that's not a bad idea.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

PV isolators are great, for biasing up floating cascodes and such. Pity the power levels are so low.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

That's not right; the typical silicon solar cell peak efficiency is in near IR

and it's not 'quantum efficiency' that is the best metric here.

If you want an efficient system, you'd be looking at IR LEDs, or something other than mass-produced silicon solar cells.

The possibilities are many; one interesting approach was a ceramic rod, with an ultrasound excitation at one end, and pickup at the other. It amounts to two speakers, an exponential cone match/coupling, and a stick.

You can also spin the stick (motor-generator) and the belt-drive variant is good for megavolts of isolation in van de Graaff installations.

A less interesting, but viable, approach is to turn the voltage off and install fresh batteries...

These Silicon Labs isolated gate drivers are excellent parts, wicked fast and up to 5kV isolation. you could prolly drive a small forward converter directly from one of the 3 amp duals.

If one can live with the lower secondary supply rating by comparison they're as good or better than the equivalent from AD, and for about half as much $$$.

Luv these things:

If the load is constant an isolated Cuk converter could be an option.

The input current and output current are continuous in the ideal Cuk, so if the secondary load is constant the impedance looking into the primary is always well-defined so primary-side regulation should be easier.

Into the primary-side of the converter from the perspective of the primary-side supply, rather

Unless one actually needs just microamperes of current these schemes are an engineering rathole IMO. They never end up being cost-effective from a size, cost, or design-time point of view.

Sometimes it doesn't pay to be clever, just use a magnetic/transformer converter and get paid! Don't look for clever hills to die on cuz you're gonna die there :)

Er, use "investment-effective" as a term there, perhaps, to avoid being redundant

A dummy says "Hmm, magnetic converter seems too complex here, I'm going to try photovoltaic."

ERRRFFF

Unless microamps is actually all you need.

Sometimes, yes.

A PV makes the gate voltage for the upper cascode fet. Current required is zero.

We had that PV around, from making our own SSRs.

PVs sure are small and cheap and quiet.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

A gate bias generator that runs off the waste heat produced by its own power FETs would be amusing and aside from the startup problem doesn't violate any laws of thermodynamics AFAIK