ad: M2Ant-1

Experiments with dummy loads

Discussion in 'Homebrew and Kit Projects' started by K4VBB, Aug 4, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-Geochron
ad: Left-2
ad: HRDLLC-2
ad: L-MFJ
ad: MessiPaoloni-1
ad: abrind-2
ad: Left-3
  1. W2WDX

    W2WDX Premium Subscriber QRZ Page

    I am sorry if I sounded flip Marcus, that was not my intention I assure you. Let me explain the methodology.

    While I understand the idea here was to build a load with these devices with things you had on hand; however, the devices you have chosen also come with very specific cooling methodology. You are using them in a way they are not designed for. It's an application error, not a bad concept. Using what's at hand is an amateur radio tradition after all. However, you are approaching this like you have a large hollow Carborundum resistor or a carbon-on-film resistor designed for immersion in a dielectric coolant. With those the resistive component is in direct contact with the coolant and are designed or chosen with sufficient surface area for that specific application. The devices you have are neither, and are certainly not designed for immersion. Their overall surface area for one is much too small to be effective using that methodology.

    High temperature devices designed with heat transfer in mind use interfaces that relate to two factors. Those two factors are conductivity and specific heat. At the devices you need a material with high heat conductivity and low specific heat. However, due to the fact that this material cannot store heat well, you need a means to quickly remove that heat. In oil cooled dummy loads the oil (which has high conductivity and low specific heat) is contained in large high surface area radiators to remove that heat quickly from the oil. In these loads the oil draws the heat directly from the source and the heat is rapidly transferred out via the radiators either by convection or in some cases a circulating pump. The load devices in those systems are in direct contact with the oil and themselves have large surface areas since they are specifically designed or chosen to transfer heat into a fluid.

    The devices you are using are not designed for that application, and require a different methodology in order to remove heat at an adequate rate. They are designed to transfer heat directly via an interface with a metal surface of sufficient mass and high heat conductivity with low specific heat. Such as a large mass copper plate, which also has high conductivity and low specific heat. That heat then needs to be transferred from the copper plate to some sort of radiator to release that heat from the copper plate. Two methods that can be employed to radiate that heat being either water cooling or forced air cooling via a large surface area radiating device.

    Reducing the number of heat transfer interfaces is tantamount to a system being efficient in transferring heat away from a heat source. Discounting the radiator to air interface, in these two methods above there are two internal interfaces; the source to the oil and the oil to the radiators; or the source to a plate and the plate to the radiators. In your case you have three. You have a plate, then radiators (both low in mass and surface area and of lower heat conductivity than copper) and then you have oil without its own radiating surface other than the smooth surface of the can. The efficiency of being able to transfer heat away from the devices is much lower in your case, since you are using devices employing a method in which they were not designed to operate.

    I alluded to you about the issue of derating the devices, which in the this case is 100% rated power at 100C and it quickly drops off to 0% at 150C. Not a large margin. So removing heat quickly is very important or else these devices will fail in short order due to rapid rise in temperature based on their derate figure. Feeling the temperature of the oil or can is not a good indicator of the actual temperature of the devices (the crucial figure for predicting failure modes), especially if heat transfer is diminished. You should be using a thermocouple applied directly to the devices when testing. It also has nothing to do with power, it has to do with cooling rate and temperature at whatever power level. With a less than proper cooling solution, the power capacity of these devices drops off quickly due to higher temperatures presented with the derating at a specific temperature. In theory, the devices can't even handle 1W if their temperature is above 150C. Without knowing the specific temperature rise rate of the devices, you cannot make a prediction of how long you can use them at a given power level before a failure mode occurs. And a failure mode in a dummy load at these power levels can be very ... well ... sparky! :p

    Again, my writing skills are sort of dry and sometimes it makes my posts sound a bit condescending. :oops: However, I assure you that was not my intent. It's always great when people get creative and try stuff, and write about it. I was just trying to help you avoid frustration and potential harm. Those devices are expensive, as is the amp.

    Have fun!

    John, W2WDX
     
    Last edited: Sep 1, 2021
  2. W0RIO

    W0RIO Ham Member QRZ Page

    Be careful about sealing too well, your heat sink could blow up and release a lot of hot oil.

    I bought a used "cantenna" that was modified with an oil reservoir container attached to the side via a plastic tube,
    if the oil heats up and expands, it has a safe way to exit the can.
     
    KI4ZUQ likes this.
  3. KF5FEI

    KF5FEI Ham Member QRZ Page

    Yeah, the old Cantennas had a spring-loaded vent, IIRC. Still remember going to the dairy supply house to get a gallon of mineral oil.
     
  4. KI4ZUQ

    KI4ZUQ Ham Member QRZ Page

    That means I should put a vent in the one I made!
     

Share This Page