DIY High Tech adjustable AC/DC dummy load

[robrenz]

I needed a dummy load that was capable of DC and AC so I could load check some of my bipolar supplies when they are sourcing AC. I also needed a load that would not introduce noise or ripple when loading supplies and measuring sub 100µV ripple and stability. So plain old power resistors sounded like an inexpensive solution.  I got (1) 2 Ohm 225W (3) 10 Ohm 50W  (3) 20 Ohm 50W (4) 50 Ohm 50W and (1) 250 Ohm 50W for $48.00.  I added stackable banana jacks 24 for $15.00 and about $6.00 worth of Pomona silicone test lead wire good for 20A. The leads are just long enough to reach a standard .75" jack spacing. The jacks allow putting these in series or parallel combinations to get more power or other resistances. If necessary these could be connected to the DUT with banana jumpers and the resistors put in a bucket of water to dissipate some more power.  A switch with banana connections could be used insert another resistance to get a step change in loading for transient response testing.

The whole set is $69.00 in materials


Parallel stacking.


Series connected on a ripple measurement.


Cutting the wires and glue lined heat shrink with the precision cutter.




Tinned and bent for insertion to the jacks.


Held in place with the wire grabbing hemostats to solder.



Soldered. The solder does not stick to the stainless steel hemostat. It takes a while for these to get hot enough to wet out on the nickel plating even with a JBC high thermal capacity tip.

[jaxbird]

Nice work, I like those resistors 

I usually do cheap 5-20W ceramics in series/parallel on a large fan to get the values/rating i need for specific testing

[free_electron]

How to make an ac current sink :
Take a bridge rectifier , put dc load at the rectified output.
Done. Ok , your minimum drop is 1.4 volt. But then dc loads typically dont work below 1 volt anyway.

Make bridge rectifier with shottkys

[jaxbird]

...
Done. Ok , your minimum drop is 1.4 volt. But then dc loads typically dont work below 1 volt anyway.
...

What do you mean DC loads don't work below 1 volt anyway?

Even with a 50 mOhm shunt they can sink up to 20A at 1V theoretically, no?

edit: bad grammar and spelling, sorry 

[jaxbird]

I'm a bit of a hack most of the time, but I have to say I enjoy your attention to detail, I assume this is required to be a good machinist

Cutting the wires and glue lined heat shrink with the precision cutter.

[Jay_Diddy_B]

Hi,
Nice work !! (as usual  )

For this kind of application I use the kind of rheostat that you might have used at school:



These have a slider so that can adjust the resistance.

I have several with maximum resistances between 20 Ohms and 400 Ohms. You can probably find then on eBay or at your local surplus outlets.



Regards,

Jay_Diddy_B

[krivx]

Hi,
Nice work !! (as usual  )

For this kind of application I use the kind of rheostat that you might have used at school:



These have a slider so that can adjust the resistance.

I have several with maximum resistances between 20 Ohms and 400 Ohms. You can probably find then on eBay or at your local surplus outlets.



Regards,

Jay_Diddy_B

Is there any way to know the power rating of these as you adjust them? Do you just stop when they get too hot?

[Jay_Diddy_B]

Is there any way to know the power rating of these as you adjust them? Do you just stop when they get too hot?

This is another one. The resistor is 12 inches (30cm) long and 2.5 inches (6.5cm) in diameter. It is actually hexagonal in this design.



The nameplate:



Indicates 33.7 Ohms and 5A

The maximum dissipation is 5 x 5 x 33.7 = 842 Watts. So this resistor will run hot. The maximum current is always 5A. The maximum dissipation depends on how much of the resistor you are using.

It was manufactured by the 'British Electric Resistance Co.' BERCO. (Where else would a British guy buy his resistors?  )

Regards,

Jay_Diddy_B

[nctnico]

Thank God it's not the French (electric) resistance 

I'm working on a couple of 1.2kW dummy loads to discharge batteries. I'm using 16 Welwyn WDBR-2 resistors (http://www.welwyn-tt.com/pdf/datasheet/WDBR.pdf) on a big heatsink with fan. These resistors can dissipate over 700W each. I'm using MOSFETs to make combinations of resistances to cater for various battery voltages and discharge currents. On of the interesting aspects is that some kind of overvoltage protection is required. The self inductance of these resistors is enough to cause nasty high voltage spikes when switched off. I could have used semiconductors but those would have required a bigger heatsink. With 1.2kW the temperature rise in the heatsink is about 60 degrees which would mean semiconductors would run at a junction temperature close to 90 degrees or more. The resistors don't mind running at these high temperatures even if the load is on 24/7.

[Monkeh]

Thank God it's not the French (electric) resistance 

I was pissing by the door when I heard a shat.

I'll get my coat.

[saturation]

Beautiful work as always.  Note if these are wirewound resistors [ likely for high power] when used for AC you'll have an impedance load not just a resistive load.

As always, as they heat up their resistance and or impedance will change slightly compared to an e-load.

Wirewound resistors are least noisy but total noise rises proportionate to heat and resistance.

[mjkuwp]

robrenz,

thanks for taking the time to post this - I could have used it a few months back

I especially like your workmanship and also those stackable banana jacks you found.

fwiw, I used incandescent bulbs for  a similar purpose.



in the picture you see an LED bulb and CFL bulb but I used those to test my watt meter - not to act as dummy loads.  Old-fashioned light bulbs have very low inductance and surprisingly do not take a long time to stabilize.

The switch on my panel is used to bypass the DMM so that its sense resistor does not heat up while the rig stabilizes.

[robrenz]

Thanks all.

@ saturation,  These are wirewound and Yeah I forgot to mention measuring the total inductance of the combination with a LCR meter for adding in the AC impedance.  Thanks for the noise chart

@ mjkump,  nice job on that setup

@ Jay_Diddy_B,  I forgot about the adjustable resistors but doesn't the useable wattage goes down as you use less than full resistance on most of them?

@ jaxbird,  I think attention to detail probably helps in any vocation.

@ free_electron  I will have to try the full wave bridge out. I don't have an actual E-load I would be using a bipolar supplies as a load on a bipolar supply that is outputting a symmetrical sine wave. I wonder if the ripple left after rectification and the bipolar supply trying to force an constant current will get the two supplies oscillating?

[HKJ]

Nice idea, but have you looked at the working temperature for this type of resistors?

[Jay_Diddy_B]

Robrenz and the group,

Yes, the useable wattage does go down if you use part of the resistor. I forgot to mention that Ohmite manufactures a range of power rheostat (potentiometers) with maximum dissipations to 1kW. They can also be found at surplus stores.

This is the link to the Ohmite catalog: http://www.ohmite.com/cat/controls_rheostats.pdf

Here is a picture of one from my collection. It is useful for testing power supplies, it has a semi-log taper. This was achieved by making the Rheostat with three gauges of flat wire. The part is marked '4.5 Ohms Spec' It is about 4 inches (100mm) in diameter.



Regards,

Jay_Diddy_B

[robrenz]

Nice idea, but have you looked at the working temperature for this type of resistors?

The plan is to use multiples in series and/or parallel so that you are not running at the rated wattage. That is also why I mentioned using banana jumpers so you can drop the whole resistor group in a bucket of water to run at rated wattage but at a lower than normal operating temperature. If its a longer term test that would eventually boil the water in the bucket. I have a sink in my shop where I could feed the bucket from the bottom with a slow, continuous supply of tap water that overflows into the sink at a steady rate.

[TerraHertz]

Here's my DIY configurable passive dummy load.
The basic unit when I got it was a bunch of fans and big heatsinks, with 18 x 200R 250W resistors, wired up with a couple of relays to give about four different total load values via a single connector.

I modified it in 2010 by ripping out all the original wiring and relays, adding some more heatsinks and resistors from the scrapbox, a bunch of terminal blocks, and wiring all individual resistors to the terminal blocks. Except for paralleling up the 200R ones in pairs, for 9 x 100R 500W.
Then made up a bunch of jumper leads with crimped ring and spade lugs. Lastly there's an insulating block with 8 isolated M6 bolts, to use as combining and final connection points.

The resistors available are 12 x 1R 100W, 4 x 9R 200W, 9 x 100R 500W.
Its first configuration was 0.8R, 2KW (for testing a HP 6269B 40V 50A supply), but it's been reconfigured a few times since then.

Currently it's in the machine shop, hence the mess in the pics.

[robrenz]

Ok, I concede, that is more high tech than mine 

[rob77]

Here's my DIY configurable passive dummy load.
The basic unit when I got it was a bunch of fans and big heatsinks, with 18 x 200R 250W resistors, wired up with a couple of relays to give about four different total load values via a single connector.

I modified it in 2010 by ripping out all the original wiring and relays, adding some more heatsinks and resistors from the scrapbox, a bunch of terminal blocks, and wiring all individual resistors to the terminal blocks. Except for paralleling up the 200R ones in pairs, for 9 x 100R 500W.
Then made up a bunch of jumper leads with crimped ring and spade lugs. Lastly there's an insulating block with 8 isolated M6 bolts, to use as combining and final connection points.

The resistors available are 12 x 1R 100W, 4 x 9R 200W, 9 x 100R 500W.
Its first configuration was 0.8R, 2KW (for testing a HP 6269B 40V 50A supply), but it's been reconfigured a few times since then.

Currently it's in the machine shop, hence the mess in the pics.

beast and the beauty in one piece of gear ! very nice job 
and as a side effect you can crank up the supply and heat up your lab during cold nights

[lowimpedance]

Here's my DIY configurable passive dummy load.
The basic unit when I got it was a bunch of fans and big heatsinks, with 18 x 200R 250W resistors, wired up with a couple of relays to give about four different total load values via a single connector.

I modified it in 2010 by ripping out all the original wiring and relays, adding some more heatsinks and resistors from the scrapbox, a bunch of terminal blocks, and wiring all individual resistors to the terminal blocks. Except for paralleling up the 200R ones in pairs, for 9 x 100R 500W.
Then made up a bunch of jumper leads with crimped ring and spade lugs. Lastly there's an insulating block with 8 isolated M6 bolts, to use as combining and final connection points.

The resistors available are 12 x 1R 100W, 4 x 9R 200W, 9 x 100R 500W.
Its first configuration was 0.8R, 2KW (for testing a HP 6269B 40V 50A supply), but it's been reconfigured a few times since then.

Currently it's in the machine shop, hence the mess in the pics.

Now that's one serious 'dummy' there, I assume that's teflon wiring used. What power supplies do you have then??.

@robrenz , nice well thought out and simple implementation, (when some one shows you how   ).

Now for when you don't need to dunk it in water you now have to build a nice bench cradle, and show pics of course 

[TerraHertz]

Now that's one serious 'dummy' there, I assume that's teflon wiring used.

I wish it was just teflon insulated wire. Unfortunately it's teflon wire salvaged from a military flight system, and it is teflon with molybdenum additive, that makes the wiring extra slippery for abrasion resistance. Unfortunately it also makes it poisonous to handle. I found this out to my misfortune while building that dummy load, and I'm the other dummy.

What power supplies do you have then??.

A few. 
https://www.eevblog.com/forum/chat/how-many-psu's-do-you-have/msg393388/#msg393388

[Richard Head]

This is what you guys need.
Pulling 600A at 54VDC.
Each unit can do 500A each!
Very nice but pricey at about $2500 each.

Dick

[saturation]

An Italian product I see, Marel XP500C load bank. 

http://www.marelweb.it/it/prodotti/2-xp-500c

[suryaputhra]

I have been looking info on rheostat does the current is based on the resistance we selected or it is always rated current of the rheostat.

[coppercone2]