Another dummy load ;-)

[madworm]

This is "YauDL" - Yet another useless Dummy Load.


YauDL flickr set

I needed one for testing DC switcher modules I acquired on bay. Building it was great fun. This is a very simple project, both in terms of making it and operation.

Latest Schematic + all design files for grabs. To get the schematic for THIS version, select the V0.1 tag in the repository.

Some of the specs:

• 0-20V input voltage
• 0-5A input current
• 20W continuous dissipation
• reverse polarity protection (using mosfets), overridable with a shunt wire
• 9V battery operation
• nice binding posts
• 10-turn current adjust potentiometer

Initially I used TO-220 fullpak mosfets, but they died pretty quickly due to overheating. Switching to standard TO-220 fixed that problem. Thermal compound also helps a bit. It works pretty well for DC, but it doesn't really like step changes of the DUT's input voltage.

As I've used Elecfreak's 10-circuit-board deal, I naturally have 9 boards left over. They are looking for a good home. If anybody wants a board to play with, send me a PM.

[Jad.z]

Nice bulid 

The binding posts are oozing quality.

[MikeK]

Very nice.  Could you please explain what U2A does?

[IanB]

Nice soldering. Very good looking joints.

[LEECH666]

Fellow German here. May I ask where you bought those binding posts?

Gruß,
Florian

[T4P]

Clip on heatsinks .... eh that's going to be your source of many heat problems to come
I would just directly screw on the mosfet

[madworm]

Fellow German here. May I ask where you bought those binding posts?

It took me quite a while to find these semi-insulated ones.

Die gibts bei voelkner.

The price is tolerable. But at least I only need an M4 hole to mount them.

[madworm]

Clip on heatsinks .... eh that's going to be your source of many heat problems to come
I would just directly screw on the mosfet

Hmmm. I will see. I've added some thermal goop to be sure it works. But if I start to drill holes, I also need to tap them to make it a proper thing. You can't just use a nut on the other sides of the heatsink. There's nothing but fins.

[madworm]

Very nice.  Could you please explain what U2A does?

Of course.

U2B is just the usual voltage follower. But as the current shunt resistors equate to 50mOhm, the voltage drop per 1A is only 50mV. U2A (non-inverting amplifier) amplifies that voltage by 20 to get 1V / 1A - which is conveniently read on the attached multimeter. As I didn't use precision resistors this 2nd opamp is also used to calibrate the device.

[MikeK]

U2B is just the usual voltage follower. But as the current shunt resistors equate to 50mOhm, the voltage drop per 1A is only 50mV. U2A (non-inverting amplifier) amplifies that voltage by 20 to get 1V / 1A - which is conveniently read on the attached multimeter. As I didn't use precision resistors this 2nd opamp is also used to calibrate the device.

Thanks.

If I were making a 1A (max) dummy load (which I am), would it make sense to alter the non-inverting amplifier gain so that I got 5V/1A reading?  I plan to read the voltage with a PIC and I figure I'd get better control and more precision if I scale it to use the full ADC range (0-5V).

[madworm]

Sounds very sensible to me.

[SeanB]

M4 tap is very cheap, just buy a tap designed to form threads in soft materials, as these do not cut but roll a thread, which is much stronger.

[Spikee]

Do you still have some pcb's ?

[madworm]

Yes I do.

The latest 'bugfix' for this version involves replacing all of the p-ch mosfets and accompanying zener diode with a single beefy schottky diode (like MBR1045, TO-220 + heatsink). It is a bit of a hack, but works quite well.


MBR1645 Schottky diode

The free-standing TO-220 should get some more support though. It's heatsink could be epoxied to the big heatsink, as they're the same potential anyway.

The voltage drop across the Schottky diode is only 0.35V at 1A and doesn't go above 0.5V for 5A, so the dummy load does work down to low input voltages.

If you can wait a bit, I've just ordered the next revision of this project and there will be a small number of DIY kits + a couple of bare pcbs as well. The next revision will have all TO-220s mounted to the big heatsink and will allow very basic remote control with 0-5V signals. I it works...

The bare pcb of the current version is 5.50€ + shipping. In case you also need the heatsink (alutronic 'V PR127/94-M3') + clips (may be a bit hard to find), let me know and I'll add it to my next order (probably tomorrow). That would be an extra 2.77€ for that.

[Spikee]

I need the board + all the stuff on it , make me a price

[madworm]

Sent you a PM.

[IanB]

I can't help noticing how nice and shiny your solder joints look. Is that done with lead free, or with 60/40?

[SeanB]

Looks like lead free to me, the striations give it away.

[madworm]

I can't help noticing how nice and shiny your solder joints look. Is that done with lead free, or with 60/40?

I have a 250g spool of "the terrible stuff" somewhere. I used the stuff only ONCE. It cost me small fortune and I haven't used it for years. Maybe lead-free paste is easier to deal with, but the wire is just evil. It doesn't wet properly, it doesn't melt properly, when cranking up the temperature to make the flux happier it almost instantly oxidizes and burns. It is CRAP. It might work if I had protective gas flowing out of the iron's tip.

I exclusively use 60/40 0.5mm wire. And when the time has come to get started with hot-air + paste I will think twice which type of paste to get. I might try a tiny tiny sample of lead-free paste, but I won't put a lot of money into basically unusable stuff again.

[T4P]

I exclusively use 60/40 0.5mm wire. And when the time has come to get started with hot-air + paste I will think twice which type of paste to get. I might try a tiny tiny sample of lead-free paste, but I won't put a lot of money into basically unusable stuff again.

The paste ... you have to do the same thing again, cranking up the temperature and making sure you have enough power ( not some teeny weeny 300W 850b but more like 700-800W 858d )
for the airflow you have

i gave up on lead free paste, somebody lent it to me the next day i returned it to him ... it wasn't cheap he wasted his money

Anyway i can't help but notice the size of your 5W resistor ... i have 5W resistors and they aren't half as long ... Low temp change resistors?
Mine shifts 11C with 2W ( aka 5.5C/W ) so i assume yours is 2.75C/W?

[madworm]

5.5K/W? That is pretty low for such a small device. The heatsink of my dummy load has about 3K/W. The resistors I used have a thermal resistance of 50K/W.

[T4P]

5.5K/W? That is pretty low for such a small device. The heatsink of my dummy load has about 3K/W. The resistors I used have a thermal resistance of 50K/W.

I am not sure :\
But my ambient temperature that moment was about 35
But i am sure i was dissipating 2W as i was discharging a very healthy 18650 of exactly 3.7V at 6.8ohms
50C/W ... Wouldn't it be 250C at 5W 
But you better check your datasheet again, on another resistor i found on element14 claims 7W at 70C
http://sg.element14.com/multicomp/mchmv-m7-2r2-j/resistor-7w-5-2r2/dp/4455575
But i can also assume my temperature readings are out as i am using a shitty meter as a thermometer BUT i could still touch the resistors
it was a little warm
But the best test is always using a power supply, i shall report back at once ... Okay, i'll do it on tuesday

[madworm]

The most these resistors will see is 0.625W each (2.5A, 100mOhm), so they won't get that hot. Less thermally induced resistance change and no need to recalibrate all the time. The resistors will probably live 'forever', quite overspecced too ;-)

I use these: (KH-208-8)

http://www.vitrohm.com/download/series/18/

[T4P]

The most these resistors will see is 0.625W each (2.5A, 100mOhm), so they won't get that hot. Less thermally induced resistance change and no need to recalibrate all the time. The resistors will probably live 'forever', quite overspecced too ;-)

I use these: (KH-208-8)

http://www.vitrohm.com/download/series/18/

There you go! 10C/W
So it's for sure that crappy meter (Can't be my 61E ... that doesn't have temp) of mine is out by a huge margin
Anyway, i did buy a 2 port digital thermometer ... overkill for my purposes but beats having to get a fluke/agilent to get any accuracy on the temp readings
PLUS! It's probably more accurate then the temp range on the fluke ...
http://www.dealextreme.com/p/k-type-digital-industrial-thermometer-with-sensor-50-c-1300-c-7286

You know what i am going to do ... selling away the POS meter