Dummy load construction

[JAelec]

Hi!


I'm constructing a dummy load from 4x100W 4 Ohms effect resistors. I was thinking of using CPU thermal paste between the resistors and the heatsinks.

Is this the right way to do it?

The resistors can get very warm, a lot warmer than a CPU since they dissipate 100W. I don't want the paste to evaporate or become runny.

Any advice welcome!

[mozz]

Very thin coating is all that is needed. Best way is to lap both surfaces but probably overkill.

[x-pro]

Has the paste evaporated?
Read the process of producing thermal conductive paste.
It is produced at a temperature much higher than what you might experience even when radiators and resistors melt.

[Chuck H]

I used thermal paste on my similar build. I didn't use proper heat sinks since I didn't want to spend the money. I just used heavy gauge angle aluminum. I don't ask a lot from mine but I've done a couple of high wattage repairs using it. I think it's about fifteen years old and there's no drying, eveporation or running of the thermal paste

EDIT: Your heat sinks look great btw. More pro and higher functioning. I'm experiencing project envy.

[Mick Bailey]

Just an observation - the heat sinks will have limited dissipation from the finned side due to being bolted fin side down against the mounting plate. Ideally a fan is needed to remove the hot air. I have two switched 4" fans in my dummy load for high output amps or extended running.

[JAelec]

Thanks, I'll go ahead with the thermal paste then. Yes it would be better with a fan, and the heatsinks should not face down, admittedly I did it this way because it was the simplest design I could come up with.

I was thinking of mounting a faceplate on the front with the help of two angle brackets. Then on the faceplate; a BNC connector for the scope, a tele jack for the amp and maybe a switch for 4, 8 and 16 Ohms.

I am able to get these maximum wattage ratings for each Ohm setting:

16 Ohms @400W max (4+4+4+4=16)
Max current would be Sqrt(P/R)=Sqrt(400/16)=5 Amps

8 Ohms @200W max (4+4=8)
Max current would be Sqrt(P/R)=Sqrt(400/8)=7 Amps

4 Ohms @ 400W max (4+4 // 4+4 = 4)
Max current would be Sqrt(P/R)=Sqrt(400/4)=10 Amps


However I don't think I could get these settings to work with just one switch, probably two are needed. Also most switches are rated for 6 Amps @250V so that would not work with 400 Watts into 4 Ohm which would yield 10 amps of current.

How would you set this up for maximum wattage?

[trobbins]

Perhaps see if those heatsink/resistor modules can be turned upside down, and still use the heatsink bolts/holes to mount to that baseplate (although baseplate holes will be different) - the 'top' of each resistor then gets additional cooling from the baseplate, and the heatsink sections provide the best form of natural convection cooling (which may avoid the need for fan cooling).

Yeh, the 4 ohm and 16 ohm configurations provide max 400W capability, but 8 ohm can only provide 200W capability. I'd suggest wire them to a terminal block that then allows easy bolt/lug type reconfiguration - not as quick as a switch, but rationise that against why and when you really want to change impedance. Perhaps also allow the 8 ohm configuration to be reconfigured as 6 or 10 ohm using the spare two resistors, as a way of confirming that say 8 ohm provides max output from a certain amplifier (and 6 or 10 ohm provide somewhat less max output).

The thermal paste tries to fill the micro-voids between the psuedo-milled heatsink and resistor surfaces. Those surfaces aren't properly milled, so the paste is meant to exclude air pockets - but not act as a layer of 'brick mortar' - so needs to be as thin as practical. Too many people just lather on a layer of paste, thinking that is best.

You may also find that including a 10:1 or 100:1 resistor divider could be useful to then connect to a soundcard (instrument 1 Megohm input) and hence use a PC and software to 'view' the voltage waveform and frequency spectrum, in lieu of a more typical oscilloscope. The divider is used to reduce the max rated signal voltage (400W, 16 ohm, 80Vrms) to a level that is safe for a soundcard.

[Chuck H]

Or just use standoff tubes for mounting the heat sinks so the base plate acts more like a frame and the heat sink fins are up. That is actually how I did mine. The angle aluminum is fin up and mounted to a project box with the resistors mounted on the heat sinks and 'upside down' but inside the chassis.

These resistors are already aluminum housed for heat sinking purposes. The idea being that they should be chassis mounted (for a little more heat sink), but away from any heat sensitive components. And I've done this with them used as built in attenuators, cathode resistors and sag effect/rectifier tube effect resistors. But used as a sole dump for power any extra heat sinking can only help. So yeah, fins up if possible if only because hot air rises. A fan? Probably not needed for anything under 200W (just from my own experience with my unit). But short of a fan and since you have those bitchin' heat sinks I would use stand offs to flip everything and mount the fins up

[Mick Bailey]

I have 4x 4 ohm sections each terminated into 4mm terminals, the type that will also take a 4mm plug. Despite having been in full-time amp repair, it's not been inconvenient not having it switched. Also, it reduces the risk of inadvertent switching which could be a problem in some situations.

[Chuck H]

Also, it reduces the risk of inadvertent switching which could be a problem in some situations.

Agree. Because mine DOES have a switch and I HAVE had it set incorrectly a time or two. I remember thinking "Why does the waveform look odd and the power is low?" This while I was going in for tone tweaks on a unit so I knew the amp was performing otherwise already. I had the dummy load set for four ohms with an eight ohm amp plugged in.

[J M Fahey]

Just as an option, when testing very high powers, say bridged rack amps and such, I drop my load resistors inside a plastic water bucket.

Water does not damage it,just let it dry afterwards to avoid rusting and corrosion..
30 seconds st high power will definitely dry them.

Or do as Pros do, or Hams, and use an oil filled bucket.

Liquids dissipate WAY more than air, even with fans.

[tedmich]

NaCl solutions (or if using Cu electrodes, cupric sulphate is better) can be used to make big dummy loads.

Or the cheap water heater elements can be used with plain water (or antifreeze/ethylene glycol solutions)

a 2-4kW (4 or 8 Ohm) load is easily made that costs about $70, but its big and heavy (50lbs)


smaller pails get you faster heating of course!