Is this the best way to make a resistive load bank for SMPS testing?

[Flyback]

Hello,
We need a 2kW load bank so that we can do full-load-to-no-load transient testing on our 2kW SMPS. Also, We also need this load bank to load the SMPS with to see if the SMPS is stable. As you know, Electronic loads cannot be used for this purpose, because they have an effective capacitance, and that effects the SMPS's feedback loop parametrics, so obviously electronic loads cannot be used in this cause.
So anyway…
Do you agree that the attached 2kW resistive load bank is the best way to actually make such a load bank?
The Twenty-four TO247 power resistors are LTO series

LTO series TO247 resistors:
https://www.farnell.com/datasheets/1817173.pdf

The reason the attached load bank is the best way to do a resistive load bank is because…

1..The TO247 resistors are simply & quickly removed and replaced by simply unscrewing and desoldering the legs off the surface mount pads. –Therefore, you can quickly and easily change it to get almost exactly what resistance value you want.

2…The fan means that the load bank occupies a small volume.

3…The PCB on which the resistor legs are soldered on to provides the connections between the resistors, so you don’t need to fiddle with multiple wires and connections. (this ‘connection’ PCB is screwed to the heatsink)

4…There are solder blob pads across each resistor, so any resistor can be quickly shorted out by just “blobbing” solder across these pads.

 

[throbscottle]

I'm very far from being an expert, but it looks as if it needs to be bigger. I like the idea of being able to adjust the resistance so easily though.

 

[ronv]

Maybe water cooled?

 

[bountyhunter]

Hello,
We need a 2kW load bank so that we can do full-load-to-no-load transient testing on our 2kW SMPS. Also, We also need this load bank to load the SMPS with to see if the SMPS is stable. As you know, Electronic loads cannot be used for this purpose, because they have an effective capacitance, and that effects the SMPS's feedback loop parametrics, so obviously electronic loads cannot be used in this cause.
So anyway…
Do you agree that the attached 2kW resistive load bank is the best way to actually make such a load bank?
The Twenty-four TO247 power resistors are LTO series

LTO series TO247 resistors:
https://www.farnell.com/datasheets/1817173.pdf

The reason the attached load bank is the best way to do a resistive load bank is because…

1..The TO247 resistors are simply & quickly removed and replaced by simply unscrewing and desoldering the legs off the surface mount pads. –Therefore, you can quickly and easily change it to get almost exactly what resistance value you want.

2…The fan means that the load bank occupies a small volume.

3…The PCB on which the resistor legs are soldered on to provides the connections between the resistors, so you don’t need to fiddle with multiple wires and connections. (this ‘connection’ PCB is screwed to the heatsink)

4…There are solder blob pads across each resistor, so any resistor can be quickly shorted out by just “blobbing” solder across these pads.

We had tons of them at Power Ten. We built them out of large power resistors in banks and put fans on the ends to blow through as required. We used similar to these:

**broken link removed**

The check loop stability you do load step testing and look at the output voltage response.

 

[Nigel Goodwin]

An easy, and cheap, way to make high power loads is to use electric heater elements.

 

[throbscottle]

That was my first thought, but then I thought as it gets hot it's resistance will increase, like a light bulb.

 

[Flyback]

I think the top post is the best way to get a very variable but accurate resistance load of 2kw...but its not that great.....it has to be somewhat disassembled to change the resistors. Ive seen loads of them in compnaies where ive worked, but all are usually just loads of cylinder power resistors with shedloads of series paralell wiring.....changing the resistance value is a nitemare. At least the top post saves that, but its not too great.

 

[Nigel Goodwin]

That was my first thought, but then I thought as it gets hot it's resistance will increase, like a light bulb.

Nowhere near as much as a bulb though - it's a nice simple and cheap solution.

 

[Flyback]

thanks, ive used cheap heater fans (£7 each).....its good, but they are about 30 ohms resistance....so you need a lot in parallel to get the 1.152 ohms we need for our 48v, 2kw load.

 

[picbits]

I was involved in a project designing an active 58Kw load bank - many finned power resistors of varying values switched in and out controlled by low RDSon Mosfets and air cooling (although we have access to chilled water cooling) with a set of balanced Mosfets in their linear region for trimming the current. The majority of the power was driven into the resistive elements while the last couple of hundred watts was absorbed by the Mosfets.

Never got past the pricing up stage after initial simulations as we have a bit of kit which can already source or sink a couple of hundred kW but an interesting exercise.

 

[simonbramble]

I once saw a teachest full of Arcol Gold resistors (50W each) with a huge fan at the bottom to cool them. It worked pretty well. They had lugs on each end of each row to enable them to series/paralleled.

Either your way or the Arcol way are both as good as each other. I see your sketch as forced cooling which is a must and like the previous post said, water may also be needed.

Sounds like a good intern project.. Yep _ I agree with you. this looks like a good way to do it

Simon

 

[picbits]

The resistors were were looking at had a rating of around 1kW peak each and 500W continuous - they were ceramic with metal fins to increase their surface area.

A few of these switched in or out plus a variable current sink (with massive heatsink) using Mosfets for finer control should work nicely.

 

[bountyhunter]

Nowhere near as much as a bulb though - it's a nice simple and cheap solution.

I used to use some spare halogen driving lights as a 60W/12V load (each). They don't get hot like resistors do.

 

[picbits]

All my 60 watt bulbs get almost as hot as my 60 watt resistors

 

[spec]

At work we used giant wire would variable resistors for dummy loads. They looked like they were made in the 1930s. If I remember correctly a 2KW version would be about 3 feet long and 9 inches in diameter. They had a large cast metal wheel at the end coupled to a threaded bar that ran the full length of the wire wound resistor and moved the wiper when you turned the wheel. The heat was conducted via the core and also by convection. They had some abuse, even had parts of the wire glowing, but they were dead reliable and never failed. I think similar rheostats were used in plating shops and cinemas.

Just speculating, but for your application I would investigate a bank of rotary rheostats in a case with the control knobs in a row so that the load could be continuously varied from min to max.

spec

**broken link removed**

https://www.alibaba.com/showroom/power-resistor-2kw.html

https://www.alibaba.com/product-det...238031703.html?spm=a2700.7724857.29.72.nsv7fX

https://sikes.en.alibaba.com/produc...le_high_power_ceramic_wirewound_resistor.html

I think Micheal Faraday used this rheostat:

 

[schmitt trigger]

[/CENTER]

Any idea of the type of material used for the resistor base?

It must be insulating and able to withstand high temperature, so I would think that in Faraday's day it would have to be ceramic.
However, the mounting holes appear oxidized, and in some areas the coating has been worn off......

 

[spec]

Any idea of the type of material used for the resistor base?

It must be insulating and able to withstand high temperature, so I would think that in Faraday's day it would have to be ceramic.
However, the mounting holes appear oxidized, and in some areas the coating has been worn off......

No, I don't know. It was just a pic off the net. As you say perhaps a ceramic of some sort.

On second inspection it looks like granite or even high temperature glass, like Pyrex.

spec

 

[Reloadron]

Years ago I got involved with an online computer parts review site and lent some help with testing the SMPS ATX Form Factor power supplies. Actually built a little demo PSU test box and then a large 1 KW unit. I used Ohmite resistors which can be seen in the links. I used the resistors I had at the time in the prototype and eventually went with MOSFETS to switch the loads in and out. Not much left of the original but here are a few remaining web pages fromwhen we worked on the thing.

https://www.bearblain.com/psu_test_2.htm
https://www.bearblain.com/psu_test.htm

Pots were used in the final versions to trim the power. The resistors need to be high power rated or they get hot and they drift. When it was done it worked well and we automated it and interfaced to a PC.

Consider Eric's mention of using a heating element.

Ron