Graham Gunderson's Energy conference presentation Most impressive and mysterious

[Farmhand]

I think he is measuring "input power" as input to the transformer, that is, between the H-bridge output and the input to the transformer.
There has been some discussion on this matter. It is my opinion that the input power should be measured as input to the power supply to the H-bridge and should also include the inputs to the other necessary power supplies, but obviously not everyone agrees with me.

I do and I think it is important as well to define the actual input (start point).

If we were to take even a device that is usually powered by grid connection to a place with no grid power we would quickly realise that to run the grid powered device we would need to generate the AC input ourselves which incurs losses.

In my opinion any device should have the input considered to be from a source of stored energy such as a battery. Doing that assures us that we can get the same result from a battery bank charged by solar or wind or whatever.

However if we consider the grid a source of stored energy then the input would be from the wall socket. Much easier to determine input at 60 Hz.

Using grid power is even cheating because it costs money, the input AC is bought with money. The grid is part of the system.

Using a H bridge then the power consumed by the H bridge itself must be considered as the H bridge is part of the system. Measuring the input to the H bridge should be easier than measuring the output from it shouldn't it ?

A transformer and load itself is not a system, it requires switching or an AC input.

..

[tinman]

I do and I think it is important as well to define the actual input (start point).

If we were to take even a device that is usually powered by grid connection to a place with no grid power we would quickly realise that to run the grid powered device we would need to generate the AC input ourselves which incurs losses.

In my opinion any device should have the input considered to be from a source of stored energy such as a battery. Doing that assures us that we can get the same result from a battery bank charged by solar or wind or whatever.

However if we consider the grid a source of stored energy then the input would be from the wall socket. Much easier to determine input at 60 Hz.

Using grid power is even cheating because it costs money, the input AC is bought with money. The grid is part of the system.

Using a H bridge then the power consumed by the H bridge itself must be considered as the H bridge is part of the system. Measuring the input to the H bridge should be easier than measuring the output from it shouldn't it ?

A transformer and load itself is not a system, it requires switching or an AC input.

..

So how far back do you go?
When we are doing power measurements on the bench,do we go all the way back to the source,and take into account heat losses at the power generation plant?.

When doing power measurements where we are using a DC input from a bench top power supply,we measure from the bench top power supplies DC output-not from the AC input to the power supply. So in other words,we circumvent all the dissipated power within that power supply,and do not take that into account.

If the power delivered from the H bridge go's only into the transformer,then we would only measure the P/in after the H bridge,as this is our bench top power supply.
If !!if!! there is power being delivered to the drive mechanism of the synchronous rectifier that bi passes the transformer,where it may add to the P/out,then this power has to be taken into account as part of the input power.

Only once a full and correct schematic is presented,will you know what has to be measured where.


Brad

[TinselKoala]

What "bench top power supply" delivers a power signal to its load like this, at 50-75 kHz?

[TinselKoala]

The "H-bridge" is pretty complicated. I can't find any good photos of the mosfets in the H-bridge but this is, I think, part of the driver circuitry. There is far more going on here than just switching 4 mosfets. Why not just use an appropriate high-power, high frequency H-bridge driver chip like HIP4080 if all you are doing is coordinating the switching of 4 mosfets?

[Farmhand]

So how far back do you go?
When we are doing power measurements on the bench,do we go all the way back to the source,and take into account heat losses at the power generation plant?.

When doing power measurements where we are using a DC input from a bench top power supply,we measure from the bench top power supplies DC output-not from the AC input to the power supply. So in other words,we circumvent all the dissipated power within that power supply,and do not take that into account.

If the power delivered from the H bridge go's only into the transformer,then we would only measure the P/in after the H bridge,as this is our bench top power supply.
If !!if!! there is power being delivered to the drive mechanism of the synchronous rectifier that bi passes the transformer,where it may add to the P/out,then this power has to be taken into account as part of the input power.

Only once a full and correct schematic is presented,will you know what has to be measured where.


Brad

Good question, The way I see it with a setup that plugs into the grid we can do a couple of things, we can consider the grid as a source of stored energy which it kinda is.

1, We can use the grid and measure the device input at the wall socket or where the non standard equipment begins such as one socket powers the function generator/whatever changes the input frequency or wave shape and the actual device input. In other words whatever is needed to run the device from the wall socket.

2, Or we can use a DC supply and measure the device input right at the DC supply as with the grid.

In other words everything that is required to run the device from a standard AC wall outlet or a standard DC supply outlet.

When I look at the efficiency of stuff I think of it that way. I don't consider the losses in a DC supply as long as I can recreate it with batteries. eg. if it's a HV DC supply then I consider the losses in the DC supply.

eg. I don't consider the losses in the solar or wind system only the output of the batteries, however if a setup is 50 Hz for
here"Aus" I don't think we need to consider the losses in the DC to AC inverter either as we would most likely have one anyway
even if living off grid. But if the device is using a 60 000 Hz input to the transformer then whatever changes the 50 Hz AC to
60 000 Hz is part of the device as it isn't for anything else and it is necessary.

We have to start somewhere.

Still it is fine and dandy to measure the efficiency of the transformer itself. But what good is that if we cannot just plug it in.

As TK  said. Where do you buy a supply with an output like that. One that can power a decent load as well.