OPEN SOURCE: The "Heavy Load" - REALLY high current (10 Amp) test load

[rhd]

Maybe this is too simplistic a solution, but why not simply attach the meter connections right to the pad of the resistors? Ie, zero trace length. It doesn't matter if the rest of the traces drop voltage, only if they drop voltage between the resistor and the point of the meter's measurement.

 

[djQUAN]

Yes that's how it's usually done. But only with one resistor. Once you parallel multiple resistors, then that's when it gets complicated.

Here's a nice read for current sense resistor PCB layout

http://www.analog.com/library/analogdialogue/archives/46-06/shunt_resistors.html

 

[IWIRE]

Better jumper ideas anyone?

What about little black slide switches ? Not sure if they come small enough at that current level :thinking: does seem it would be more convenient than jumpers. Just have to remember to open one switch when closing another.

 

[rhd]

What about little black slide switches ? Not sure if they come small enough at that current level :thinking: does seem it would be more convenient than jumpers. Just have to remember to open one switch when closing another.

Doubtful that they could handle the current.

I'm actually not sure whether these jumpers are a problem now that I think of it. The jumper polls are actually reasonably thick, and the jumpers themselves usually have a piece of metal inside with a fair amount of both thickness and width to it. I would be surprised if the jumpers were the weak link.

 

[IWIRE]

From an engineering standpoint they are probably a weak link. From a real world using it standpoint, they seem fine. I haven't actually checked the jumpers with my IR thermometer but I will.
I haven't mounted the load on its permanent home yet, I'm waiting on a fan and a piece of black acrylic
I can do a voltage drop check across the jumper under load also. Seems to work fine though.
I have several different test loads and this one is the most universal for doing everything.
Great job. Thanks again.

 

[BShanahan14rulz]

I demand knife switch for jumper, and 4 AWG wire with banana plugs for ammeter connection. DO IT NAO!

Nah, I kid. Cool project, though. I like the thought of using a piece of open source equipment that has been "group-thunk" rather than something made by one person.

 

[IWIRE]

I like the thought of using a piece of open source equipment that has been "group-thunk" rather than something made by one person.

I`m not sure about that. The first one rhd came up with is pretty nice. I'm betting rhd can make this version even better all by himself. I'm sure I'm just an unneeded distraction.

 

[rhd]

Hey all.

I haven't had time to get to this yet, so I thought I'd upload the board / schematic files in case anyone else is itching to move this forward.

http://www.filedropper.com/hctl08 (schematic)
http://www.filedropper.com/hctl08_1 (board)

EDIT: NVM - I see that I already did this in an earlier post!

 

[djQUAN]

Here's one I worked with this afternoon.

I worked with RHD's initial schematic and worked from there.

I still used the pin headers for jumpers as I have not found a better substitute. Another way could be to use a 2x2 pin header and use two jumpers in parallel. The external connections are through commonly found 508mil pitch terminal blocks.

Instead of two rows of diodes, I decided to use just one so that we could use only one large heatsink for the entire test load. I also used a ceramic heatsink footprint but the pin pitch of TO-220 resistors can fit in the PCB with the same placement line as the diodes.

Then you could cut off part of the PCB or leave it as is and mount the diodes underneath to make a low profile or infinite size heatsinked test load.

I can email the eagle files to anyone interested for further refinement as this was just a quick work. The voltage drop chart hasn't been added yet.

 

[crazyspaz]

Alrighty everyone. This was quite a while ago, but I am FINALLY ready to get some of these made. With help from DjQuan, I got the input and test pads changed a bit to fit these parst: screw terminal lever terminal. I will probably use screw for the input, and then get a few lever terminals and see if they fit DMM probes. If they don't I will use just header pins and use alligator clip probes.

I am using SeeedStudio like RHD did, simply because they are quite cheap. I am thinking of going with white since its a bit cheaper and I think it would the black/white contrast would look really nice. Anyone else interested in these? I would be happy to get rid of most of them, might ask a few $ to cover shipping and part of the PCB cost. We could also order 10x or 5x (depending on interest which determines how many boards I get) sets of parts to drive the cost down a bit on them.

part list:
-terminal 1: lever(optional)
-terminal 2:screw(optional)
-diode (x7): FFPF10UA60ST
-resistor (x1): .1 Ohm. As RHD stated in post #4:

Even though the board itself suggests using a 1 ohm resistor, I would strongly suggest moving to a 0.1 ohm resistor instead. It doesn't really make the conversion hard (just move a decimal place in your head), and it saves you from dropping 4V over the resistor when you're testing 4A, etc. That's a lot of voltage drop.

-headers (x9): 3M9447-ND
jumper: this or that. Doesn't matter which, but you need 2 of them. These seem a bit ridiculous, because its either 30cents from Digikey, or get them for pennies a piece, but wait on the slow boat/rickshaw from china. Personally, I would rather pay the 30cents and get it here in less than a month.
Here is a shared digikey cart with everything in it (not all is necessary):http://www.digikey.com/short/7hq59c

so what I need to know from people:
1) number of people interested in how many boards
2) is the PCB color OK (white vs. RHD's black)
3) do people want me to order all the parts so we can get a discount on them


edit: just to clarify, this is the board from post #4 of this thread, not DjQUAN's board.

 

[rhd]

I would strongly suggest making those traces larger.

 

[rhd]

sorry, which traces? I didn't change the size of any traces, just the contact point pitch- its the file you put up for download.
edit: wait, are you referring to DjQ's post from a few months ago?

Oh, sorry - thought you were making djQUAN's

 

[djQUAN]

Yeah he was going to make your version. My version was just to test the arrangement of the parts so it was not final.

 

[crazyspaz]

Giving this a bit of a bump now that everything is sorted...Anyone interested?

 

[rhd]

Giving this a bit of a bump now that everything is sorted...Anyone interested?

Looks good!

I'm guessing you're not using a 1ohm resistor, but a 0.1 ohm resistor?

In that case, the note at the top should probably be revised to indicate the correct voltage drop.

 

[djQUAN]

I just noticed. I forgot to add a restrict in one of the points to connect the voltmeter to the resistor. Without it, the thin trace that connects directly to the resistor pin is bypassed by the copper pour. I'll send a revised one asap

edit: Crazyspaz, revised file available in the link I sent you before.