The death of transformers? AC Direct DC Enabler is a solid-state solution for AC-DC power supplies

(post is archived)

You are currently inside a comment thread.

Click here to see all the comments (18).

[–] • 1 pt

Output power – 0.2W to 15W

Dynamic input voltage – 50V to 480V

On-demand output voltage 3.3V, 5V, 12V, or anything in between

Highest power density at 5W per cubic inch

Short circuit, over-voltage, and thermal protection

UL safety certifications

So Amber can take an AC supply rated up to 480V as input power and gain a UL safety certification when there is no significant hi-side, low-side separation going on? Sure the proposed SOIC package has some missing pins which I assume is there for hi-side, low-side separation, but the pin pitch is small enough that I would not want to use that if the AC supply exceeds 100VAC. There needs to be better safety considerations for this to be any better than a simple capacitive dropper circuit. Linear power supplies may require bulky and expensive transformers, but they have the best hi-side, low-side isolation and built in current limiting should things go wrong with a stray finger/hand or other fault that sends HV AC to the output.

Also, 15W output power isn't much, but they don't say what the efficiency of the setup is. What is the TDP of this setup and later SOIC package? How much thermal consideration will consumer designs need when operating at max power output? Who cares if they shrink it to a 10x20mm package if you have to attach a 30W TDP heatsink to it to maintain operational specs. And what about the quality characteristics of the DC output? What kind of DC filtering will be needed? What is the ripple like? Are there drop out modes and does it provide any over-voltage/over-current protection? Reverse current? Thermal cutoff? Voltage stability over a range of current draws? Transient voltage/current spike isolation? Power factor? So many important technical questions need to be answered here before I would consider this as a replacement for a SMPS or linear PS. The IoT aspects of this only make it less desirable since we don't need anymore back doors into or consumer electronics than we already have.

>* Output power – 0.2W to 15W >* Dynamic input voltage – 50V to 480V >* On-demand output voltage 3.3V, 5V, 12V, or anything in between >* Highest power density at 5W per cubic inch >* Short circuit, over-voltage, and thermal protection >* UL safety certifications So Amber can take an AC supply rated up to 480V as input power and gain a UL safety certification when there is no significant hi-side, low-side separation going on? Sure the proposed SOIC package has some missing pins which I assume is there for hi-side, low-side separation, but the pin pitch is small enough that I would not want to use that if the AC supply exceeds 100VAC. There needs to be better safety considerations for this to be any better than a simple capacitive dropper circuit. Linear power supplies may require bulky and expensive transformers, but they have the best hi-side, low-side isolation and built in current limiting should things go wrong with a stray finger/hand or other fault that sends HV AC to the output. Also, 15W output power isn't much, but they don't say what the efficiency of the setup is. What is the TDP of this setup and later SOIC package? How much thermal consideration will consumer designs need when operating at max power output? Who cares if they shrink it to a 10x20mm package if you have to attach a 30W TDP heatsink to it to maintain operational specs. And what about the quality characteristics of the DC output? What kind of DC filtering will be needed? What is the ripple like? Are there drop out modes and does it provide any over-voltage/over-current protection? Reverse current? Thermal cutoff? Voltage stability over a range of current draws? Transient voltage/current spike isolation? Power factor? So many important technical questions need to be answered here before I would consider this as a replacement for a SMPS or linear PS. The IoT aspects of this only make it less desirable since we don't need anymore back doors into or consumer electronics than we already have.

link

[–] • 0 pt

They're targeting simple device chargers initially. Can't even quick charge a phone or tablet with 15W. I anticipate them being acquired by TI or another entity and having the patent applied to something better.

parent
link