50V 100A 500W Load Tester

THE_RB

Joined Feb 11, 2008
5,438
Are you expecting the heatsink temp to get to 80'C? That's hot! But even then (because of the fan forced airflow) the rest of the cabinet interior and parts should be at MUCH less than 80'C and not that much hotter than room temp.

It's always a good idea to do testing on high power heatsinks, that was the way I was taught to do it in industry. To just connect a pot etc to drive all your FETs and the PSU and turn it up to full power and measure how hot it gets for the given power dissipation with your chosen fan, then finalise the physical design based on that information. I would aim for max temp of the heatsink itself of about 60'C.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
At present, I'm not sure if I'm going to fully enclose the unit, and if I do, I'll probably build it out of cardboard or scrap metal. The current case doesn't fit with the MOSFETs on it. The advantage to enclosing it is guiding the air (better cooling), and less risk of getting electrocuted by the potentially 50V rails.
 
Last edited:

Thread Starter

tom66

Joined May 9, 2009
2,595
Got the fans today and some heatsinks for the regulators on the main board. I will be doing some airflow analysis possibly with talcum powder or smoke. I need to see where the air is going across the heatsink.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Here's the progress of the main board layout.

This is a bit cluttered, but it needs to fit within 10cm x 10cm to cost less at Seeed Studio. It's a 2 layer board.

On board so far: 5V buck converter (3A max output), -12V/-8V voltage inverter & regulator, +8V regulator, current generator.

I'd like to add a display for voltage and current/watts/sim.resistance - preferably two groups of 4 x 7 segment displays. I've not got many pins free on my micro (a PIC18F4550) so I've decided to go for some kind of 7 segment display driver IC. Any suggestions? It needs to control eight seven segment displays and work with SPI or I2C.

The board will also have USB as the microcontroller supports it. Not entirely sure how I'll use it, but I'll put a port there for it. I might use it for computer control, or for data-logging to a memory stick...
 

Attachments

Last edited:

thatoneguy

Joined Feb 19, 2009
6,359
You can always move more air over it, or if it will be used for short loads, put a gallon of water on each side, with one side of the water in contact with the heat sink, and the container on the other 4 sides to keep the water in.

Water is Awesome at soaking up heat, and if the load isn't 100% continual, such as in a computer, or car, you don't need to pump it through a radiator. Just having it in contact will soak up most all of the heat, though in-package temps might get high, it should overall stay cool for a load test. Then your cooling fans can cool down the sink and water to room temperature again between uses (or dump out the water between uses if not frequently used).

For Display: Maxim 8 Digit Starburst display driver with built in fonts MAX6954

The user can configure the MAX6954 with digit-by-digit or even segment-by-segment blink control. The driver can use an internal RC oscillator to set multiplex and blink frequencies, or an external clock to enable easy synchronization of multiple drivers.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
You can always move more air over it, or if it will be used for short loads, put a gallon of water on each side, with one side of the water in contact with the heat sink, and the container on the other 4 sides to keep the water in.

Water is Awesome at soaking up heat, and if the load isn't 100% continual, such as in a computer, or car, you don't need to pump it through a radiator. Just having it in contact will soak up most all of the heat, though in-package temps might get high, it should overall stay cool for a load test. Then your cooling fans can cool down the sink and water to room temperature again between uses (or dump out the water between uses if not frequently used).

For Display: Maxim 8 Digit Starburst display driver with built in fonts MAX6954
Interesting idea with water, but I fear it will cause problems for carrying the unit; I'd need some sealed containers full of it and then it would be very heavy. I would like to be able to dissipate at least 500W continuously and ideally 900W. 1,800W would be a peak dissipation for 10-20 seconds (dependent on heatsink temperature not exceeding 125°C), such a dissipation could be used to test power supply over-current/over-power features.

The display chip is interesting especially being in DIP but I could solder a SSOP. I think I'll put the display stuff on a separate board (possibly with the 7 seg displays on it.)
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Progress of main board so far:

Current sense & voltage sense all finished.

Microcontroller is next. That includes DAC, EEPROM, voltage reference and LCD connector, among others. The 7-seg display controller will be a separate board.

There are two quad op-amp packages and five relays on the board.

4 x M3 holes around the board will be used to mount it in the case.
 

Attachments

Thread Starter

tom66

Joined May 9, 2009
2,595
I finished the design of the main board.

I ran out of space on the top with through hole parts -- had to resort to using SMD in a few places, but nothing other than 0805 and SO8. There is also one jumper wire.
 

Attachments

Thread Starter

tom66

Joined May 9, 2009
2,595
Okay, I'm going to put an order in for the main PCB soon. Still got to finish the main chassis and think of a casing arrangement.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Soldered the board up - what can I say, soldering a 1 layer through hole PCB is very nice and easy. But, I honestly have to say, I prefer SMT. Seriously; it's far easier to do, and can be done with a reflow system involving a hot plate or toaster oven. I might make a future fan controller in 0805/SO8 SMT, but for now I'm sticking with this.

I used a low ESR high quality 105°C cap for decoupling the 12V rail to the fan, mainly because that's under the most ripple current. The other parts use generic caps, only 85°C, so I will have to see how well it stacks up when it gets hot.
 

Attachments

Thread Starter

tom66

Joined May 9, 2009
2,595
Okay, so I built up the fan controller. I had to wire in a separate 5V supply as that's what the TC642 wants. It does spin the fan, but even with very hot water does not go very fast. In testing, I eliminated the controller; the PWM increases with temperature but this does not increase fan speed. I also eliminated the fan, as it spins fast when powered directly from 12V. This leaves either the base resistor (470 ohm) or the NPN itself (2N2222) as being faulty or incorrect.

I think I may need to drop the resistor to around 100 ohms, but I fear the controller IC won't like sourcing 50mA. Maybe I'll switch to a signal FET.
 
Last edited:

Thread Starter

tom66

Joined May 9, 2009
2,595
So, I replaced the NPN with a high power MOSFET (overkill, but voltage controlled) and tested the fan controller, and it's kinda disappointing, to be honest. It does vary the PWM width with temperature, but the fault modes (the main reason I chose it) don't do auto-restart attempts and don't throttle the fan to high power in an attempt to get it to run.

I think I'll probably just use software PWM or pulse skipping on an internal 50 Hz loop, measuring the temperature of the heatsink through an I2C sensor. That will free up some pins.

Main board is finished, will be ordering PCBs for it soon.
 

mcgyvr

Joined Oct 15, 2009
5,394
Not sure if you are interested.. but google "acme ps2L manual" it gives complete schematic/parts list,etc.. for a 60V 1000W max solid state load. I have 25 of them at work and they are great. I use them as a 50V 15A load all the time.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Not sure if you are interested.. but google "acme ps2L manual" it gives complete schematic/parts list,etc.. for a 60V 1000W max solid state load. I have 25 of them at work and they are great. I use them as a 50V 15A load all the time.
Thanks for that - could be useful.
 

Phil916

Joined Feb 3, 2010
2
Not sure if you are interested.. but google "acme ps2L manual" it gives complete schematic/parts list,etc.. for a 60V 1000W max solid state load. I have 25 of them at work and they are great. I use them as a 50V 15A load all the time.

Hello,

did someone succed finding this acme ps2L manual nowdays ?

still no result on my side :(
 

bolm

Joined Feb 3, 2014
4
I built one here (called Big Dummy);



The two heatsinks came from a junkyard piece of equpment and were already populated with the big transistors, already wired as darlingtons. That was common in high power linear power supplies, 1960's 70's 80's etc.

It uses a pot to set the current. The biasing is set up at the moment to work with about 5v-48v and 0A to 30A.
Hello THE_RB. How about schematic of Big dummy load. I'm interesting it. Thank a lot
 
Last edited:
Top