Compact Stereo Dummy Load for Amplifier Testing

In a past post, I’ve attached a picture of the load I was using for speaker amplifier testing. I have a box full of these 50W wire-wound resistors and a heatsink (HS) I’ve tapped to be able to attach these resistors easily. I was simply connecting as needed for the specific case. In practice, I rarely change the default 8×2-ohm resistors which are split into 2 loads of 8-ohm each. When I needed to dissipate significant power I would normally point a fan at that HS and be done with it. However, this wasn’t very convenient, and I wanted something more “user friendly” to replace it, this is what will be described in this post.

The main goal was to build a dummy load that could handle similar amount of a power in a smaller and more “desktop friendly” package (which means forced cooling). While at it I wanted it to have standard banana jacks, two independent load channels, and a switch to select between 4-ohm and 8-ohm modes.

The secondary goal was to reuse what I could from parts I already had. The case was something I’ve had from a past project that ended up going into a larger case. The HS was taken out of an old 5-CH receiver that I took apart some years ago. I had to cut the HS down in length to fit into the case, but other than that it was a great fit, especially in height.

I’ve had to order a few extra parts, such as 60mm fans, a switch that can handle high current with two poles at each switch, and some peripherals. I chose to leave the previous load unaltered, which meant I couldn’t reuse the resistors from that. Instead I’ve decided to try and order 100W resistors from AliExpress to check them out and see if they are good enough. As typical with such projects, one part is typically holding down progress. In this case it was the thermostat (explained later) which took a few months to get here after the first order was lost in transit, while everything else was completed months ago.

The structure I chose to use for the switchable load is as depicted in Figure 1. If you crunch the numbers, assuming 100W max power handling for each resistor, you will see it can handle >250W at both 4-ohm (272W) and 8-ohm (292W) modes. This is more than I will need, but since the Hafler I recently tested can drive 250W per channel this seemed like a reasonable power handling capacity, at least for short term until thermal limit is reached.

Fig. 1. Single Channel Load Schematic

The nominal resistance (assuming ideal switches and wires with zero resistance) will be 7.9-ohm for 8-ohm mode, and 3.97-ohm for 4-ohm mode. However, with the non-idealities of the switches and wires this will be slightly higher and closer to target. The spec on the wire-wound resistors is typically +/-5%, but as with other parts, they are typically closer to target than these limit values. In my case, I was happy to see that after assembly, the resistance was <1% out from the 4-ohm/8-ohm targets for both channels.

Fig. 2. Resistors Mounted to HS

The HS with resistors fit snuggly into the case, which also needed some work in the form of air intake holes at the front of the bottom plate. The resistors aren’t touching the bottom metal plate of the case so the case won’t get too hot to the touch. There are some spacers there that limit this distance.

Fig. 3. HS Mounted to Case

The rear panel holds the 2 fans (5V 60mm each) and a DC jack for 5V input. As always, I use a USB power brick to power as many of my DIY projects as I can. Its simple, easy, cheap, and always available.

Fig. 4. Rear Panel

The front panel received similar treatment, with the help of a p-touch printer to make it nicer and easier to use.

Fig. 5. Front Panel

The case holds a few extra components as means of safety. The first is an 80C normally-open (NO) thermostat in series with a red LED mounted to the front panel, the second is a 90C NO thermostat in series with a buzzer. These are there as backup in case I dissipate too much power over the load. The thermostats were mounted at the center of the HS using JB-weld to hold them in place and conduct heat from the HS to the thermostats. Figure 6 shows everything in place before closing the lid.

Fig. 6. Everything in Place

I’ve measured a few parameters of the load to know its capabilities. The resistors were an unknown to me, so I’ve measured the temperature coefficient, and it was ~80ppm/C which is what I except for plain wire-wound resistors of this value. The inductance was measured at 10KHz and 100Khz and showed similar numbers. The values were 1.3uH at 4-ohm setting, and 2.5uH at 8-ohm. There were small variations (~3%) between the two channels. These numbers of inductance are low enough for my audio needs, and can easily cover the band of frequencies I will be testing at. In comparison, my previous load made up of 4 series Dale RH-50 2ohm resistors (8-ohm/200W overall) measures at 1.9uH. Therefore, considering the higher power rating of the new load, the 30% increase in inductance is completely reasonable.

Finally I wanted to check the power handling capacity of the unit for continuous operation without tripping either of the thermostats. This means I can allow ~55C temperature increase for the HS, assuming a 25C ambient temperature. I’ve connected the Horizon DTPS 6012 to supply the power for the test, and also monitored the HS temperature as well as the temperature of the air exhaust at the back of the load box. The thermocouple mounted to the HS will obviously not monitor the actual HS temperature since its also exposed to the air around it and therefore only gives some rough estimation of the temperature. However, by monitoring these readings I could see if a steady state temperature was already achieved or if things are still heating up.

Fig. 7. Testing Under Load

The power supply was set to 61.2V at the load terminals, the load was connected in series for a total resistance of 16ohm. The power dissipated at the load was 234W. At these conditions I could see the temperature readouts of the meter settle in ~10 minutes, with only small changes following this point (which is why I will assume this is the steady state in my calculations below) . Despite this, after 20 minutes, the “Temp Warning” LED turned on. I was glad to see that the case was only warm to the touch at this point, which means there is no safety risk with handling the unit while its active.

Since ambient during the test was ~21C, assuming the thermostat is indeed fairly accurate at 80C (the external thermocouple read 70.5C at this point, close enough consider its only partially coupled), we can calculate the thermal resistance. 234W were dissipated over the load during the test, and assuming this is close enough to steady state, we get 0.252C/W thermal resistance from the HS to the air given the air flow of the fans. Therefore, for the 55C temp rise I’m using as a rule of thumb (as stated earlier), it will be able to handle ~220W continuously, we can round that down to 200W, which I’m more than happy with for such a small enclosure. I’ve repeated the test at 200W, and the “Temp Warning” LED didn’t turn on even after significantly longer periods of time.

This case is small enough to fit onto my bench so that its always accessible, much more “user friendly” than my previous load which I was constantly hiding in the parts bin.

Leave a Reply

Your email address will not be published. Required fields are marked *