Adjustable transistor load

I am a fairly intermediate experienced technician, an apprenticeship in england. Could you talk me briefly how that would work? I have dealt with Bridge rectifiers before as tyhey are used in our power supply circuits and produces a continuous positive trace. I have used a LM317 voltage regulator to control the transistor. as I am dealing with high power transistors they are getting rather hot. I am using a large heatsink but now need to include some feedback to counteract the losses due to heat. Thank you for the alternative idea though its something to include in my project.

Dan

As always, depends on what you're trying to accomplish. What's your definition of load? Watts is the only unit of measure you mentioned. Do you want constant wattage load? Constant current load? Constant resistance load? Other kind of load?

"AC-DC Power supply"???? Do you want to load an AC source? Depending on what kind of load characteristics you want, this can become complex.

What about transient response. Do you change the load? How fast does it have to react? Does the source change? How fast does the load have to react? What level of accuracy/repeatability do you need? What do you want to happen when some limit parameter is exceeded? Like under 3V or over 60V? Fuse blows? Source voltage has AC component? Why do you want to minimize EMC noise? Minimize doesn't help much. If it matters, it has a number(s) and a procedure to measure it. What are the numbers?

FWIW, for DC power supply testing, I use a transient load that can slew almost an amp/nanosecond. Turns up all sorts of power supply layout and transient problems. For simple fixed current load, a Dynaload brand linear load works.

Your specs are challenging. 100W @ 3V is 33A. Whatever you do, you have to dissipate the heat somewhere. The 3V limit means that you can't do much with resistors unless you switch them which affects transient response. The 60V limit means that you have to be careful about secondary breakdown if you use bipolar transistors. Plan on using MANY more transistors than you'd think from the wattage requirement. Yes, I know that 100W @ 60V is not much current, but if you count on this, make sure you include electronic protection to make sure you don't have high current transients at high voltage that blast your transistors. It's probably cheaper and more reliable to use more transistors than to employ more sophisticated protection. And when you think about it, you're more likely to want more than 1.6A at 60V anyway. And put a circuit breaker to protect against catastrophic failure. Board layout or wiring paths matter! You have to make sure the current goes where you intend.

Once built a charge dump load for a wind generator. Used a PIC with a single 0-11Amp switching load and saturated FETs to switch in BIG resistors to change ranges 10-20A, 20-30A etc. Just measured the voltage and ramped the PWM up and down to keep the voltage below the target. Fixed voltage made it easier. Transient response was horrible, but ok for this application.

All depends on what you're trying to accomplish. 95% of the design is trivial. It's the last 5% that nobody told you about that is gonna bite you in the...anybody who tells you it's 100% trivial or that you only need an op-amp and a transistor has never built one. One size does not fit all unless you have unlimited funds. More complete specification can simplify your task...assuming you really don't need much. mike

Do you mean a supply with a dc output, and you want a controllable current sink? You can make a controllable current sink with an op-amp and a power transistor. Connect the opamp output into the base of the transistor (through an appropriate resistor), and connect the emitter of the transistor to ground through a current sensing resistor (say, 0.1 ohm). Then hook the junction of the emitter and the resistor to the inverting input of the opamp, and apply a voltage to the non-inverting input as your demand, or current-setting input. The opamp output will increase until the current sense resistor has the same voltage as the non-inverting input. So, 1 volt in for 10 amps out, 0.1 volt for 1 amp. It should also work with a power fet as the current sink, mutatis mutandis. You should take steps, like the opamp output resistor and shunt diodes, to protect the opamp from excessively high voltages.

-- john

How about a bridge rectifier with the transistor as a variable active load across the output?

Leon

Here are a few ideas for design from magazines that will get you started in a path that may get you what you want. You didn't say whether your load should be constant power, constant current, constant resistance, etc. Those types of loads require different techniques to get you there, so maybe if you could be more specific about how you want to test the power supplies, maybe someone can be more specific with a solution for you.

Additionally, ISTR an article in Nuts & Volts magazine that described an active power supply load.. I'll see if I can dig it up and scan it if you're interested. I think it was a constant resistance load.

Hi I work in a power supply design business. When we test for emc radiation we connect the units outputs to sets of resistive loads, to draw the correct amount of power from each of the outputs. I am trying to build a circuit that will achieve the same but can be used on a wider variety of outputs. once the circuit is connected it is left running. no transients are applied. I was hopin this would be a simple project but have become lost along the way. Thank you for your help Dan

schreef in bericht news: snipped-for-privacy@o13g2000cwo.googlegroups.com...

Dan,

Elector ever published an active controllable load. Bur it's some time ago and I don't have it at my hands at the moment.

petrus bitbyter