EEVblog #1293 - New NGP800 800W Power Supply

[Tek_TDS220]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

[nctnico]

Having mosfet, diodes transformers and cts is normal for every topology. For a dual switch flyback - where are the input flyback clamp diodes? Yes it can recirculate the energy, but you need the diodes! If you can find them it's a good argument.

If you look closely you can spot some devices which look like diodes near the MOSFETs. The flyback diodes don't need to be big since the primary current is relatively low (it also depends on how low the leakage inductance of the transformers is). I think there are also a lot of components on the solder side of the board. I have not spotted the gate drivers for example.

Also with an llc or flyback, how do you get down to zero volts output, with current control? Flyback maybe operate discontinuous, but what fun that would be from a control stand point.

I still can't shake the feeling of that extra big inductor doing more than a little filtering. And that big mosfet... Or is that just output enable?

I think the big MOSFET is used as an active load to get to 0V and to pull the output to a lower voltage (discharge the capacitors). The SOA graph looks good enough for that purpose. For an output enable it wouldn't need to be on a heatsink. Also switching the load using a MOSFET leads to a few problems with rush-in currents etc.

All in all it would be nice to have some high-res pictures of both sides of the board to be sure.

[rx8pilot]

This one single unit would replace all of my MASSIVE Agilent system power supplies as well as provide a vastly improved user interface.
I will seriously consider this unit in the next 6mos. for a specific project. The space savings, performance, and ease of use will likely make the price tag worth it.

[nctnico]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

Which supplies did you try? I can't imagine why any power supply would have problems delivering 1V at 10A if it is rated for that current.

[Phoenix]

If you look closely you can spot some devices which look like diodes near the MOSFETs. The flyback diodes don't need to be big since the primary current is relatively low (it also depends on how low the leakage inductance of the transformers is). I think there are also a lot of components on the solder side of the board. I have not spotted the gate drivers for example.

The gate drive transformers are right next to the CTs and it's easy enough to hide a couple of gate resistors somewhere. The US1M diodes might do the job and there could be stuff on the bottom as there isn't much decoupling I can see around the PIC32. But if it's a flyback why would you need 2 output diodes per channel - simply in parallel?

All in all it would be nice to have some high-res pictures of both sides of the board to be sure.

Yes please; it 'should' end the topology discussion.

[Phoenix]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

That's a pertient question for the topology discussion too. According to the datasheet this supply can produce/regulate rated current down to zero volts - what topology is capable of that besides a buck? Can a flyback or LLC readily do this (even with an active load)?

https://scdn.rohde-schwarz.com/ur/pws/dl_downloads/dl_common_library/dl_brochures_and_datasheets/pdf_1/NGP800_dat_en_3609-1927-32_v0100.pdf

[EEVblog]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

That's a pertient question for the topology discussion too. According to the datasheet this supply can produce/regulate rated current down to zero volts - what topology is capable of that besides a buck? Can a flyback or LLC readily do this (even with an active load)?
https://scdn.rohde-schwarz.com/ur/pws/dl_downloads/dl_common_library/dl_brochures_and_datasheets/pdf_1/NGP800_dat_en_3609-1927-32_v0100.pdf

Maybe that extra MOSFET is an LDO?

[Kleinstein]

A surprising point is that there is not much capacitance on the output side of the switcher.

The small toroids look like gate drive transformers and current sense transformers. With only 2 gate drive transformers it could be tricky with a full bridge (I don't see 2 well isolated secondary sections).  For me this more like points to a 2 phase configuration - with only one large inductor, likely 2 flybacks and the inductor as a common filter after the converter. The challenges for the switcher are more like reasonable fast reaction to load changes and the large voltage range.
The FETs may be optimized / good for resonant conversion, but this does not mean they have to be used in this topology only.

The extra MOSFET could be something like tap switching - but not sure about that.
I would expect the 3 rd heat sink near the output could  be a kind of linear LDO stage to get a really smooth and fast reacting output. The specs should tell if this is just a switcher of switcher + LDO.

[nctnico]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

That's a pertient question for the topology discussion too. According to the datasheet this supply can produce/regulate rated current down to zero volts - what topology is capable of that besides a buck? Can a flyback or LLC readily do this (even with an active load)?

Flyback = buck with a transformer in between. Maybe a better term would be isolated buck converter.

[Tek_TDS220]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

That's a pertient question for the topology discussion too. According to the datasheet this supply can produce/regulate rated current down to zero volts - what topology is capable of that besides a buck? Can a flyback or LLC readily do this (even with an active load)?

Flyback = buck with a transformer in between. Maybe a better term would be isolated buck converter.

Thanks for the responses.  I didn't phrase my question well.  I need to run a lot of current through a wire with a cold resistance of ~0.05 ohms.  I've tried a couple of powers supplies that should be able to do this, but they have protection circuitry that prevents it.  I would love to have the R&S supply, but I can't justify the cost. 

After thinking about it more, a beefy MOSFET with a big heat sink would solve my problem.  I'll try that.

[cgroen]

I don't want to disrupt this discussion on circuit topology, but I'm interested in the ability of this power supply to deliver at least 8 amps (or more?) at 1 volt.  Are there other power supplies (even a single channel device) that can do this?  I know there are supplies that deliver 10 amps and are capable of regulating at 1 volt, but they shut down when you ask for both at the same time.

I just tested my R&S HMP4040, that one does this without problems (I was using remote sense also)

[EEVblog]

Someone on the Youtube comments just mentioned this is likely a Phase-Shifted Full-Bridge, and I suspect they might be right.

[Phoenix]

Someone on the Youtube comments just mentioned this is likely a Phase-Shifted Full-Bridge, and I suspect they might be right.

It matches the components better if one of the transformers is the series inductor. But, if it is a full bridge primary why do they have 2 CTs, and 4 diode on the output? It would be a lot easier to have a single CT in the transformer path than say one in each phase leg and have to OR their outputs.

[nctnico]

Someone on the Youtube comments just mentioned this is likely a Phase-Shifted Full-Bridge, and I suspect they might be right.

I doubt it. It wouldn't explain the need for two transformers and why the switching MOSFETs are on a big heatsink. Unless they run them in parallel and leave one out in the 400W version but it would leave the question why there are two current transformers. However the first question to answer is whether the topology supports a wide output voltage range. So far most switching variable voltage output power supplies I have taken apart uses a flyback / buck topology. In a forward converter-ish design like the Phase-Shifted Full-Bridge the output inductor is the problem to make a wide range of output voltages possible. But it is not impossible. The very cheap PSUs use  a forward converter design. If it is a forward-converter design then I suspect they run two identical transformer / diode sections in parallel on the 800W version and one section in the 400W version. However, if it is a forward converter design I would expect an extra filter inductor because the large toroid would be the inductor used to charge/discharge the energy from the transformer. In a lab PSU I would expect some extra filtering (but this could impair transient behaviour).

It would be nice if you could take the power board out and take pictures of it. I think there are quite a few components on the solder side (including gate drivers).

[EEVblog]

It would be nice if you could take the power board out and take pictures of it. I think there are quite a few components on the solder side (including gate drivers).

Doesn't look that easy, but might give it a go some time, just not right now.

[Phoenix]

It would be nice if you could take the power board out and take pictures of it. I think there are quite a few components on the solder side (including gate drivers).

Yes please!

I'm not sure what you're missing in terms of gate drives, however. The gate drive transformers are next to the CTs. And you could hide a couple of gate resistors on the other side of the heatsink.

[ignilux]

Doesn't look that easy, but might give it a go some time, just not right now.

Wtf else do you have to do right now? This is your full-time job and that !@%$# virus has us all locked up inside with nowhere to go and nothing to do. I was so bored earlier that I actually cleaned my bench. Do you have any idea what it takes to get me to do that?!

[Messtechniker]

HMC8043

My most used power supply as well, love it.

Same here.

[nctnico]

It would be nice if you could take the power board out and take pictures of it. I think there are quite a few components on the solder side (including gate drivers).

Yes please!

I'm not sure what you're missing in terms of gate drives, however. The gate drive transformers are next to the CTs. And you could hide a couple of gate resistors on the other side of the heatsink.

Usually there is some additional circuitry together with gate drive transformers.

[EEVblog]

Doesn't look that easy, but might give it a go some time, just not right now.

Wtf else do you have to do right now?

Take care of the kids (they are now home from school full time)
Move into the new lab.
Business stuff.
Packing and shipping (my logistics person is in lockdown).
And, you know, videos of countless other stuff that is on the back burner, and ones that, you know, may interest me more than a THIRD video on this power supply.
In case you haven't looked at the view stats, that last video wasn't exactly popular, in fact my worst performing one in a long time.

[nctnico]

In case you haven't looked at the view stats, that last video wasn't exactly popular, in fact my worst performing one in a long time.

Karma in action for not taking it apart completely