My constant current load - 90W , 3A 30V

[Spikee]

I just made another video , the opamp problem is fixed . In some weird way mu arduino mega is getting reasonably hot. I might have to investigate that.



I just went around town to find a new lead acid or gel battery but i coulden't find one anywhere ... so i will have to wait until the weekend so i can take the one i have home to my dorms .

I also ordered a  0.1 ohm 100w resistor for low voltage , high current testing.

[Spikee]

It has been a while but i'm still buzy testing the current load.
All my gel / lead batteries are dead so i bought a second hand Delta Elektronika M5-10 (5V - 10A ) power supply.
In my youtube video i demonstrate the impact of a higher load resistance vs a lower one.
In particular i test 1 ohm versus 0.1 ohm 100W resistor. I test the load on 10A 5V and at the end you see a connector melting because of the current.

Youtube link:



If you like my videos just Comment , rate and subscribe .
Also a thumbs up really helps! 

[mazurov]

JDB: I made a circuit from your model and it works very well!

[mazurov]

Here's the circuit heating up old Pentium Pro heatsink. Holds 1.9A current within several milliamps from cold start at RT to ~60C.

[Jay_Diddy_B]

Hi,

Here are some pictures of the load that I made. As shown the load is good for about 10W. This can be increased to 30W with a bigger heatsink and a fan. For higher power I would suggest using the Master/Slave configuration.

I found a small load like this to be useful when testing multiple output power supplies.

Outside view:



Inside view:



Board top side:



Board solder side:




Board Design:




Regards

Jay_Diddy_B

[Spikee]

So i had some more time and i worked on the constant current load.

I have fitted 4 IRFP064N on a Hyper z600 CPU cooler and used one opamp for each mosfet. As a load resistor i use 0.1 Ohm 3W now but i will be switching them for 8 W equivalents. I will be testing the supply whit a 200W load tomorrow. But i already peak tested it at 63A , 10V = +-630W but some wires , shunt wattage etc was to low to run it for longer than a 5-10 seconds. This setup should be capable of 200W whitout getting hotter than 50-60 deg C. using one 120mm fan which has an resonable CFM. I can fit another fan to improve it ... but i have to make the pcb first. I need to get some connectors so i can hook up my 750W coolermaster pc power supply and see what the max of this cooling setup is. I think it is about 300-400W and then it will probably be around 60-80 degrees C.

Sorry for the bad photo's / video my phone was empty and it is really late now ... but i'm just soo exited 

Pic of setup:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032839.jpg

Breadboard:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032825.jpg

Feedback + gate wires:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032828.jpg

Close-up of resistors , mosfet construction:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_033010.jpg

Todo:

-Make pcb
-Bigger wiring on some places
-Add intelligence (STM32 + touch screen lcd)
-total current measurement (LT6100 whit 2m ohm shunt)
-Put in an enclosure
- ...

[blackdog]

Hi Spikee, :-)


How dit you connect the powerfets to the heatsink, glue?

Kind regarts,
Blackdog

[LEECH666]

I've talked to Spikee on Skype some time ago. The FETs are indeed glued onto the heatsink with some thermal adhesive.

[Spikee]

Hi Spikee, :-)


How dit you connect the powerfets to the heatsink, glue?

Kind regarts,
Blackdog

I used Arctic Silver thermal adhesive. That stuff is super strong , it's almost impossible to get it off without sanding the heatsink.
It conducts but that does not matter because all the drains are already connected whit wires. I'm making my schematic now in Altium and i will post it here. The fan puts alot of noise in the supply so extra schottky diode etc is advised. Without the schottky diode the output noise is around 20mV PP.

gr.spikee

63A peak current test:


A few moment before i had it at 50-60A for 20 seconds and the insulation of my wires was melting.

Schematic :
https://dl.dropbox.com/u/100819959/constant%20current%20load/Constant%20current%20EEVblog_Spikee.pdf

[rbola35618]

Nice video and projects Spikee  .

[grenert]

Jay_Diddy_B,
Did that case come with the heatsink?  Where did you find it?
Thanks!

[Spikee]

Nice video and projects Spikee  .

Thanks!

Jay_Diddy_B,
Did that case come with the heatsink?  Where did you find it?
Thanks!

He used an off the shelf case. Got a metal plate the same size as the normal end cap. Drilled some mounting holes in it and mounted the heatsink on the plate.

[Jay_Diddy_B]

Hi Grenert,

Spikee is correct. Here are the details:

The case was manufactured by Hammond part number 1598A.

The rear panel was cut from aluminum sheet 0.062, 1.5mm.

The Heatsink was cut from a length of aluminum extrusion. It is about 1 inch (25mm high)

I drilled and tapped the heatsink to attach it to the metal plate.

This design is good for about 10W dissipation.

Regards,

Jay_Diddy_B

[clcham]

So i had some more time and i worked on the constant current load.

I have fitted 4 IRFP064N on a Hyper z600 CPU cooler and used one opamp for each mosfet. As a load resistor i use 0.1 Ohm 3W now but i will be switching them for 8 W equivalents. I will be testing the supply whit a 200W load tomorrow. But i already peak tested it at 63A , 10V = +-630W but some wires , shunt wattage etc was to low to run it for longer than a 5-10 seconds. This setup should be capable of 200W whitout getting hotter than 50-60 deg C. using one 120mm fan which has an resonable CFM. I can fit another fan to improve it ... but i have to make the pcb first. I need to get some connectors so i can hook up my 750W coolermaster pc power supply and see what the max of this cooling setup is. I think it is about 300-400W and then it will probably be around 60-80 degrees C.

Sorry for the bad photo's / video my phone was empty and it is really late now ... but i'm just soo exited 

Pic of setup:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032839.jpg

Breadboard:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032825.jpg

Feedback + gate wires:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_032828.jpg

Close-up of resistors , mosfet construction:
https://dl.dropbox.com/u/100819959/constant%20current%20load/20130225_033010.jpg

Todo:

-Make pcb
-Bigger wiring on some places
-Add intelligence (STM32 + touch screen lcd)
-total current measurement (LT6100 whit 2m ohm shunt)
-Put in an enclosure
- ...

Hi Spikee, I've seen your schematic, I have a few questions:
1) why do you need a pull down resistor?
2) there is a 2.7k resistor in the feedback path, what is it for?
3) what is the maximum voltage capable of discharging with this design?

thanks

[Spikee]

1. So the mosfets turn off when you apply no current.
2. Probably some kind of filter (haven't looked at schematic , there is probably also a cap somewhere)
3. It is limited by the voltage of the mosfet. The IRFP064N has a VDSS = 55V . If you need more then you need to change the mosfet.

[ghent360]

Hi Spikee,

None of the links are working. Schematics gives 404- not found. Video does not play anymore. Would you kindly re-upload.