Upgraded LM324 Based 300W 72V 20A Electronic Load

[enut11]

Upgraded LM324 Based 300W 72V 20A Electronic Load
The origins of this project goes back almost 6 years.
https://www.eevblog.com/forum/projects/upgraded-lm324-based-150w-72v-10a-electronic-load/msg1506610/#msg1506610
My Take#2 150W/10A/30V DIY unit has served me well. However, when it comes to Electronic Loads, big is often not big enough! We want more voltage and/or more current and more power.

If you want to extend the capability of an existing unit, upgrading the voltage rating can be relatively simple. Just insert a suitable value high wattage resistor in series with the DUT. Eg, my original 30v limit could be extended to, say 60v with a series 6R/150W. It may be difficult finding the right resistor but easy to implement for testing.

Increasing the current is more difficult and only really solvable with an uprated set of MOSFETs. This means more MOSFETs or maybe the use of linear MOSFETs. The latter can be expensive (see Reply #13 from above) so I opted for the former.

Electronic DC Load Take#3
It is hard to go past the LM324 kit that I used previously so I bought 2 more. The aim is to achieve at least the full 72v rating by using high voltage MOSFETs. The current will double to 20 amps as will the total power to 300 watts. The 2 kits will be wired in parallel and controlled by one pot.

[enut11]

Well, I opened one of the kits and found a problem already. Notice the 2W metal film resistors? Well, these should be 1/2W types to fit the PCB spacing! I messaged the supplier but not expecting much.

This was the listing on eBay (AU)
https://www.ebay.com.au/itm/134811253388

[Dan123456]

Is the issue that the holes in the PCB are too small for the leads? Or just the overall length of the resistor is the issue?

If it is the latter, would bending the leads back on themselves to make them flit like the below be an option (not drawn to scale but hope it makes sense  )?

They may sit a little higher off the board and not look super pretty but could that work?

Edit: or even have them standing upright rather than horizontal. I have seen resisters like that on a few ewaste boards 

[enut11]

Hi @Dan123456. The resistors are too long for the hole spacing and too thick to fit next to each other. Even though they might be squeezed in, it would look ugly as some would be upright and at odd angles. No, I am going to use the right sized components...

[enut11]

If you want to build a room heater, you need the right parts. I was lucky to find a couple of large heatsinks at our local metals recycle center. The fans I had are 120mm and will be used in push-pull fashion. 100mm fans would be a better fit but they seem to be hard to find.

I also had a temperature alarm kit from Jaycar Electronics (AU). I modified it to drive a relay to activate the fans at a set heatsink temperature. There are lots of similar devices available on  eBay.

[enut11]

Happy New Year everyone from Aus

Not much progress here due to festivities.

Just bolstered some of the PCB heavy current tracks with copper wire.

[enut11]

PCB populated with the smaller components.
Off to the side is a sample of one of the resistors supplied with the kit that the seller claims is a good fit for the board, LOL.

[enut11]

One half of the new load almost ready for testing. Aim for this module is 70v max and 10A max with an overriding 150W max power level.

MOSFETs are bolted directly to the heatsink for maximum heat transfer. Later I will install a watchdog temperature monitor firstly to switch on the fans at around 40C and to activate a buzzer at around 60C.

Source resistors are 0.1R (0.22R were in the kit) to allow high current testing at low volts, eg, a single LiPo cell. The 5W resistors are mounted 5mm above the PCB to aid cooling.

[enut11]

The 'wind tunnel' is almost complete. Twin DC Load modules will run in parallel controlled by one pot. Total MOSFETs = 8. Push pull fans have been mounted.

I tested the individual modules to 80v and 10A (not at the same time!). Each module has also been tested to 120W.

[enut11]

Temporary control panel until I organise a suitable box.

A series connected 10R+2uF snubber will be used across the input terminals to mitigate potential oscillations as per @Kleinstein suggestion.
A 30A automotive fuse will be in series with the +terminal.

One of the modules will be designated CONTROL and the other SLAVE. Interconnected power and control wiring will allow full operation with only one pot.

With a 100v input rating, the live heatsink will need to be electrically isolated.

[enut11]

Modules wired in parallel with individual high current cables back to the input terminals. Slave control signal was taken back to the wiper on the AMPS pot.

View of back of temporary control panel. Snubber is a pair of 1uF 600v caps plus 10R resistor. DUT voltage is monitored via front panel. Not fitted yet is an ammeter or a 30A fuse. As MOSFETs die with a short circuit, a fuse is highly recommended.

Tested to just over 20A at 12v as monitored  by a clamp meter.

[enut11]

The control signal for the SLAVE module was fed to where a pot would normally go. The socket labelled 'RP 4K7' was wired with center pin and ground only. I used a shielded cable back to the same location on the master CONTROL board.

[Vovk_Z]

Temporary control panel until I organise a suitable box.

It may be more convenient to have a range switch, like 0-3 A and 0-30 A. Or even three position switch - "0/3/30A". 
And possibly, you want powerful protective Shottky diode in series with modules (or in series with MOSFETs) to protect your load under test from occasional reverse.

[enut11]

Thanks @Vovk_Z. A timely reminder to install a switch to be able to instantly switch off the load if needed. This would simply ground the pot wiper signal.

As for an additional current range, not sure that is needed here but could simply be another resistor in series with the top end of the control pot. The value would need to be such that the full range of the pot gives a chosen lower amps limit.

Input reverse protection would require a big diode. The resulting voltage drop could compromise low voltage high current tests.

[enut11]

Learnings from this project:
1) Electronic Load modules can be paralleled for additional amps and power capability.
2) Additional voltage capability needs higher rated MOSFETs.

Changes to Chinese load kits
1) The 5W SOURCE resistors reduced from 0.22R to 0.1R. Improves  loss at high currents at the cost of reduced sensitivity at very low currents. The resistors were mounted off board to improve cooling.
2) Substituted output devices with IRFP260N MOSFETs. These are rated up to 200v working.
3) 7812 voltage regulator was used to stabilise the TL431 voltage standard.
4) 15v supply was used to drive 3). The fans power was came from before the regulator.
5) All the metal film resistors were discarded! They were the wrong size and rating to properly fit the PCB.

A note on MOSFETs
I ordered 8x  IRFP260N devices. Two different types were supplied in the batch. Five weighed around 5.5g each and the rest around 4.5g each. Clearly they were not the same and this was confirmed when I compared the labels. The lighter devices also had a smaller label area. I purchased 3 more of the heavier devices for this project.

A better way would be to purchase 20x MOSFETs and 20x 0.1R 5W resistors and set about matching them in batches of 4. Matching MOSFETs was covered in Reply #33:
https://www.eevblog.com/forum/projects/upgraded-lm324-based-150w-72v-10a-electronic-load/25/

[adinsen]

Thanks very much for sharing your project. It's been helpful for me researching ways to build a high power electronic load. I have gone by a different path (2N3055's controlled by a single LM723), which I'll share in a new thread.

However, I got my own kit of the LM324 based load in the mail a few days ago. Like you, I was puzzled by the incorrect resistors and fitted my own 1/4W's instead. It's a good kit, though. I think it is probably best upgraded to higher amps by running several of them in parallel (with the same reference voltage).

/Anders