money maker circuit, just design it (and find right buyer)

rs. The concept of going from 50Hz transformer to smpsu is an obvious one, and is normally avoided on grounds of cost. This is how it goes with most i deas, even if novel they normally have no financial value.

It isn't with microwaves.

NT

er

rers. The concept of going from 50Hz transformer to smpsu is an obvious one , and is normally avoided on grounds of cost. This is how it goes with most ideas, even if novel they normally have no financial value.

Meeting efficiency mandates mainly. There's not much difference in transport costs in the case of microwave ove ns.

NT

Switcher microwaves have been around for a while. Copper is just too expensive nowadays. And switchers allow better power control.

Has anyone replaced the magnetron with semiconductors? I've recantly read conjectures that it might be done with GaN in the future.

Oh, very carefully calculated. Logistics is a big deal for retailers and *everything* gets specified down to the exact dimensions and markings on each outside master carton. The number that can fit into a TEU (39m^3) is what is critical, not the weight. Surface shipment costs of this sort of thing are dominated by the physical volume of the cartons. You can't make a microwave oven of given internal capacity much smaller by fancy power supply technology- it's inherently an empty metal double-walled box with a magnetron and connecting waveguide off to the side.

A small microwave oven - you might be able to fit 500 in a TEU so shipping costs from the PRD in China to Europe or the US midwest will be maybe a buck or two each, and dock-to-dock time only a few weeks.

Probably better control over the power (conventional ovens are controlled strictly by on/off time and the power varies with line voltage etc.), and the lighter weight might appeal to some buyers. Downside is that they'll almost surely not last as long. We still have a chrome "Radarange" from ca. 1970 with red LED display and all. Works fine (not as even as modern units with a carousel- it depends on a microwave stirrer to spread the waves).

--sp

It's certainly an obvious thing to do. Someday surely.

I would suspect it's also still hard to approach the ~75% peak efficiency of a magnetron, cost aside. The guys pushing space-based solar generating stations have some optimistic estimates of what can be achieved with particular topologies- I looked into it a few years ago.

I remember seeing the actual volume cost of a consumer microwave oven magnetron some years ago (memory fades so I can't reliably quote dollars but it's not an especially expensive component compared to microwave semiconductors - maybe $5-6 or so - which includes the heat sink and output plumbing with mounting lugs and so on).

I guess if you could convince the GaN transistors to operate from mains voltage directly there might be some cost savings, and the size of the product could *definitely* be reduced.

Perhaps the minivan glove-box microwave for the soccer mom on-the-run will be the killer app.

--sp

Copper is not that expensive. It is currently around $2 a pound while the high water mark in 2011 was about $4.50. This is about as low as it has been for the last 10 years. Transformers are still very common.

My Philips microwave is from the eighties, and indeed I would not expect a SMPS based microwave to last that long. However, that would be another reason for manufactures to use SMPS. Electrolytics in an SMPS provide the timed failure they like to see.

I'm dismantled at least one MOT that had windings of aluminium wire. Disappointing, because it was wire that I was looking for.

Is there any sign of power GaN costs coming down explosively? Do they even serve any massive quantity markets? Cell phones yes, but only a little bit at a time (

take an existinf FPS like half-life, or cube2 and modify it.

mainly you need new artwork and new scenarios.

Well there goes the MOT! Now, what can the hackers do with a 1000 watt, high voltage switcher? Mikek