load bank puzzle

Yeah, and... for an inspection-quality load, you might want a barcode for each of a dozen variants. At inspection time, if the right barcode isn't scanned, the inspection is invalid.

Adjustable boxes aren't good "standard" loads for a formal inspection process, IMHO. At a minimum, you'd want a calibration date, and complete record of the settings, printed onto an attachable sticker label, for the final report.

How does one calibrate eight different adjustable loads, each having dozens of settings?

Vitreous enameled resistors would be best for not outgassing. Silicone or whatever would be worse.

Wirewound resistors tend to run at crazy temps, like 270C, at full power. I wouldn't want to scorch the PCB.

There are some nice DPAK power resistors that could be clamp mounted on heat sinks, which would provide a lot of surface area.

There are also TO-220 power resistors, which could be attached to heat sinks with nuts and bolts.

lørdag den 18. marts 2023 kl. 23.57.33 UTC+1 skrev John Larkin:

Overv time you've talked about all sorts of low voltage loads that swotted the heat in mosfets and a heatsink.

Doesn't that have a placxe here?

RL

I'm planning a separate big single-channel electronic load board that uses mosfets on a CPU cooler and can do programmed constant-current and constant-resistance ac or dc over a huge range. That topology would be complex for a small 8-channel load that has onboard inductors. The mosfet thing needs isolated fast DACs and ADCs per channel and closed loop control in an FPGA. Resistors and SSR are sure easy.

A CPU cooler is a possibility on the 8-channel board. A K199 type would fit on my board and can easily dissipate 150 watts. The bottom copper plate would need to be tapped to screw down a bunch of (expensive, unavailable) TO-220 resistors or something.

High thermal conductivity (ie, fairly pure) copper is gummy and a nuisance to tap, but it can be done. That would roughly double the cost of the cooler. At least TO-220 resistors don't need insulators like mosfets would.

If I suggest a product with even three variants my boss will attack me. Production hates that. Hence programmability. Customers would like programmability too.

I wouldn't calibrate the resistors; just specify their tolerance. We would calibrate the ADCs, one per channel to report actual voltage and current.

I don't think that something like 5% tolerance would be a problem for people simulating solenoids or motors. The measurements could be calibrated to 1%.

There are some very cool 2 and 3-channel isolated ADCs intended for use in electric meters. Cheap but not super fast; good for average measurements but not for closing fast control loops.

I plan to digitize actual load voltage and current on every channel. An FPGA can calculate power dissipation and protect the parts. A polyfuse might be prudent too.

So use the cool ADCs for metrology and do the control loop with op-amp?

piglet

how much inductance do you need? that could take up quite a bit of space too

I'm estimating that 5 or 10 mH would keep their PWM driver happy. A real solenoid or torque motor could be henries. In that case, they'd have to add it externally.

Simulating a resistor needs multiplication. That could be done with an MDAC but gets complicated.

Digital pot or PWM multiplier could do I guess, I have never used FPGAs so still go first to analog ways!

piglet

A dpot could do the multiply well enough, at least over a moderate signal range. A dpot as the actual load resistor wouldn't work well.

FPGAs are liberating in that you can do very complex signal processing and control loops basically for free. As long as you get someone else to grunt out the code.

Could be done with a digital pot, ah yeah, that's a type of MDAC I guess.

In my experience the biggest issue with the versatility of even the simplest load is connecting to the DUT.

Every sodding device uses a different connector.

RL

Yes. If we do a D25, we will also offer a small termination board, D25 to barrier strip.

My big single-channel load board will use a Phoenix connector where the male has wire screws.

For car batteries light bulbs are convenient. They can be constantly on. I have several emergency kits around for old cars, with wattage ranging from 5 Watt to more than a 100 Watts.

Groetjes Albert