DIY Electronic Load Project Thread :D

[R.Naresh]

Hello,
I am planing to build a DIY Electronic Load so i searched alot of youtube videos kits and premade kits and so and so on

And i just want a Constant Current DC Load nothing more complex atleast for now...... (150W i think.....)
And there is 2 ways to do it
1. Analog (with op-amp and some simple circuits)

2.Digital (with a Microcontroller probably a Arduino)

well heres what i found

Analog

75W Electronic Load Module (Pre-Soldered) (XHDZ-FZ-2D)

Links to buy:
https://www.banggood.com/DC-1V-100V-Constant-Current-Source-Electronic-Load-Board-75W-0-10A-Power-Tester-p-1148591.html?rmmds=search&cur_warehouse=CN
https://pt.aliexpress.com/item/75W-Constant-Current-Electronic-Load-100V-6-6A-Battery-Discharge-Capacity-Tester/32707669154.html?spm=a2g0s.13010208.99999999.262.6kVhrE
https://www.ebay.com/sch/items/?_nkw=dc+1v100v+constant+current+source+electronic+load&_sacat=&_ex_kw=&_mPrRngCbx=1&_udlo=&_udhi=&_sop=12&_fpos=&_fspt=1&_sadis=&LH_CAds=&rmvSB=true

Links to Videos:
(Volt Log)
(Volt Log) (From 3:15)

Schematic:
https://drive.google.com/file/d/0B4zALdueXAxdcm1jSWt6Tm9TZXc/view (Credits to Adrian)

150W Electronic Load Module (UN-Soldered) (XHDZ-FZ-2D)

Links to buy:
https://pt.aliexpress.com/item/DIY-LM324-Electronic-Load-Power-150W-Simple-Electronic-Load-Kit-72V2A-15V10A/32822340743.html?spm=a2g0s.13010208.99999999.270.6kVhrE
https://www.ebay.com/itm/NEW-150W-72V-10A-Constant-Current-Electronic-Load-Battery-Discharge-Capacity/122835967837?_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D49452%26meid%3Daa7aa749a4f24e55b69ad0f470bc4379%26pid%3D100675%26rk%3D2%26rkt%3D15%26sd%3D152799349972&_trksid=p2481888.c100675.m4236&_trkparms=pageci%253Adef320d9-dee2-11e7-9ba5-74dbd180fb11%257Cparentrq%253A488500ad1600a688ace48449ffffeb3a%257Ciid%253A1

Links to Videos:
(Electronic Trivia) (Mailbag) (From 11:00)
(Electronic Trivia) (Build Part 1)
(Electronic Trivia) (Build Part 2)

Schematic:
On Ebay Seller's Listing


EEVBlog Dave's DIY Electronic Load
(EEV Blog)

Mjlorton's take on EEV Blog Daves design
(Playlist)

Marco Reaps Quick And POWERFUL Electronic Load


Digital

Skullcom Hobby Electronics DC Load (Arduino)


Adam Welch Electronic DC Load (Arduino)
(Part 1)
(Part 2)
(Part 3)
(Part 4)
(Part 5)
(Part 6)

my requirements are about 150W...
can the secound kit do 150W with those transistors ? is it even worth it 11$ish ?
and can i replace the puny T0-220 Transistors with a bigger one as they have genorusly provided pad for it
and can i buy the 1st 75W Kit and probally prallel the transistors with balancing resistors (All Mosfets are not the same )
And wat is the main diffrence between pralleling more mosfets with balanceing resistors other than having each opamp to control the mostfet throught sepreate feeeback (Current Series Resistors)
and use a single pot to set the VRef for all the opamps this is how the 150W Model does

if i would go with the 150W Model wat cheap and readily aviable transistors will fit the bigger pads (Bigger Package more heat capacity i belive  )
and i could probally do 150w With some active cooling ? (if i go for a huge headsink and bigger transistors than the kit provides + Fans )
i also though i could use like 1 small intel cpu cooler (65w rated i think....) for each transistor (the kit uses a quad opamp and 4 transistors...)

Well i will probally keep EDITing this to add new stuff so every thing is documented well here but ill post to notify below the thread

Thank You for your Contrubution!

[jaycee]

First things first - if you want to dissipate 150W of power, you need a GOOD heatsink for this, and this is often where the money goes. Forced air cooling is an absolute must.
If you can get scrap PC's, then PC coolers are a good way to do this.. especially the kind with 80mm fans on them. Bolt 2 or 3 of them to an aluminium L angle or plate, and you have a great heatsink. I wouldn't just use a single PC heatsink.

I cant really comment on kits as I designed my own.. but.. you definitely want a few pass devices, and TO-247 packages are the best option to dissipate heat. I'd probably use at least two IRFP240 for the job.

So far most of the designs i've seen have crap compensation of the opamp...usually just a big capacitor which slugs the opamp bandwidth heavily so the transient response is poor.. or they oscillate under certain conditions. For more info about this, Jay_Diddy_B did a great series of articles about it.

[Freelander]

Just get one of these -
Excellent bit of kit. Fully programmable. Uses Modbus RTU so easy to use.
Stacks of features, you wont need anything else really.
They also have a higher power version available for about 35 dollars.

Mine was about 23 dollars IIRK.
https://item.taobao.com/item.htm?spm=a312a.7700824.w4004-1173757744.11.21c386dcaHJ2Lb&id=550339476787
(right click and translate in Chrome)
It offers loads more features than listed.

I have enclosed a few bits of info.
I can give you a link to a page to buy one (the link supplied is the manufacturer) - you can try him but may have to go to a supplier.
I have translated the modbus info already (they dont supply it in English as their market is for the internal Chinese mainland for load testing and burn in.

The newer '300W' (tongue in cheek but - it can handle it for short times - well mine can handle 150 for short periods - no worries)
is also listed on their site. This has only just come out. I will be ordering one soon. Mine is the 150W 60v 10A unit.
Main site is -
https://shop103127720.world.taobao.com/?spm=a312a.7728556.2015080705.3.3463f1b30tPIIt

Beauty of it is, just hook it up to an arduino or whatever, a small display and a keypad and you have full programmability.
Yay !! 

(had to zip the product overview file as it was over the 1 meg limit) -
CC CV C Resistance check it out. Not worth buggering about with other really. I was going to build or mod something and went for this. It does what it says on the tin - VERY well made as well - real quality even if they may be a tad optimistic on the total power (for any length of time)
If you need any more info on the Modbus stuff let me know.
 

Enjoy...


edit - the modbus info shown are from the parameter files loaded into 'Modbus Poll' (commercial modbus testing) - but you dont need that at all - just hook it to an Arduino  .
All the data is in Chinese only, but I have translated all that matters.
Joe.
If you need 'more power Egor' then PM me .......................

[Freelander]

some more piccies - great quality - even fully isolated RS485 - not the usual crap.
crazy low price as well. !

[Freelander]

https://item.taobao.com/item.htm?spm=2013.1.20141002.2.35ca2fc2cb4tjH&scm=1007.10009.70205.100200300000001&id=554016021922&pvid=7e7b4c3b-0e4c-464f-be8f-4c82188eef1d

If you want really simple and just be able to set it with a pot (CC or CV) then the above is less than 12 dollars.
Might as well get the modbus version though as you can add buttons displays and bells and flashing gizmos to impress the neighbours. 

[R.Naresh]

@jaycee
well i am willing to go the route of analog just CC and which kit should i use or should i wire up my own ? and if i use the IRFP240 4 of those and 4 cpu heat sinks with forced cooling will it work ?

can i use the 150w kit with the bigger pads and do it or can i also get the small one and paralle the mosfets with balancing resistors(i geuss)
?

[R.Naresh]

https://pt.aliexpress.com/item/IRFP240-TO-247-20A200V-5pcs-lot/32377452008.html?ws_ab_test=searchweb0_0,searchweb201602_4_10152_10065_10151_10344_10068_10345_10342_10343_51102_10340_10341_10609_10541_5130013_5000013_10084_10083_10307_5080013_10539_10312_10059_10313_10314_10534_100031_10604_10603_10103_10605_5060013_10594_10596_10142_10107,searchweb201603_12,ppcSwitch_5&algo_expid=e76f1f78-ad9d-4f12-b4a9-c57b9236be6d-1&algo_pvid=e76f1f78-ad9d-4f12-b4a9-c57b9236be6d&rmStoreLevelAB=0 (IRFP240 5pcs) (One is around )

https://pt.aliexpress.com/item/5-PCS-IRFP250NPBF-IRFP250-IRFP250N-TO-3P-TO247-Novo-original-frete-gr-tis/32843556937.html?s=p&ws_ab_test=searchweb0_0%2Csearchweb201602_4_10152_10065_10151_10344_10068_10345_10342_10343_51102_10340_10341_10609_10541_5130013_5000013_10084_10083_10307_5080013_10539_10312_10059_10313_10314_10534_100031_10604_10603_10103_10605_5060013_10594_10596_10142_10107%2Csearchweb201603_12%2CppcSwitch_5&algo_expid=516ad0e3-23bd-46f3-ab92-fafc6646ad69-0&algo_pvid=516ad0e3-23bd-46f3-ab92-fafc6646ad69&rmStoreLevelAB=0(IRFP250) (One is around )


which to buy i am trying for 150w but i would like to get 300w (Greedy me :LOL:)
and i think if i prallel the mosfets in the 75w kit the current shunts may get a bit hotter..?
very simple yet diffucult to deside.

EDIT: whats the diffrence between china post air mail and parcel ? :LOL: (Pic Below)

[jaycee]

That 150W kit doesnt look very good to me. I've never understood why people think there is a need to have one opamp per MOSFET - that is partly the function of the resistors in the MOSFET source. The LM324 isn't a particularly good opamp either, though it does have an output that goes all the way down to ground. There is no gate protection on the MOSFETs. No regulation of the power input fed to the opamp or the voltage reference. The tracking on the PCB around the MOSFETs doesnt look particularly heavy for carrying high current.

[duak]

Jaycee, I found two reasons for using one op-amp per FET:

1.) current sharing between the FETs is improved because the op-amp compensates for each FET's transfer function,  A simple source ballast resistor approximates this too but is not as effective.
2.) bandwidth and stability are improved because a number of op-amps now share driving a capacitive load.  This is especially important if the load is to be used to simulate transient loading.

I designed an electronic load 30 odd years ago using one op-amp and four FETs and had no end of trouble stabilizing it.  Going to one op-amp per FET made it far easier.

Cheers,

[R.Naresh]

@jaycee it does have an voltage reference if 2.5v in a transistor package.....

@dusk so what is the way to go?

[jaycee]

Jaycee, I found two reasons for using one op-amp per FET:

1.) current sharing between the FETs is improved because the op-amp compensates for each FET's transfer function,  A simple source ballast resistor approximates this too but is not as effective.
2.) bandwidth and stability are improved because a number of op-amps now share driving a capacitive load.  This is especially important if the load is to be used to simulate transient loading.

I designed an electronic load 30 odd years ago using one op-amp and four FETs and had no end of trouble stabilizing it.  Going to one op-amp per FET made it far easier.

Hmm point 2 is a good one. I went with a two opamp approach - one to sense the currents across the source resistors, and one to do the actual control. Both have R-C networks across them to stabilise the feedback loop, as well as snubber networks across the load.

Point 1 sounds good as well but makes it more difficult to do constant voltage mode as well...

@jaycee it does have an voltage reference if 2.5v in a transistor package.....

Yes, a TL431, which is a shunt regulator. However the voltage being fed to it, and to the opamp is not regulated, so the opamp has to work harder to reject supply noise, and the reference has to work harder to smooth it out. Using a 7812 regulator would have been cheap and effective.

[R.Naresh]

well ok you mean the main supply for everthing ??

[Bootalito]

Would anyone be interested in a full collaboration effort to design a fully automated battery characteriser from the ground up? What I envision would be a 4-6 month collaboration using appropriate  online collaboration tools such as Google Docs to Define requirements, easyeda.com for schematics and PCB layout, and get Github.com for the firmware.

I have the rough outline of a project that I've been wanting to complete for a couple years, and doing it collaboratively is the push I need finally get started on it. A brief description of the device I'm envisioning is as follows:
Automatically cycle any chemistry battery to determine capacity and life. Basically you would need to be able to charge and discharge a battery based on appropriate battery chemistry and number of cells until some threshold is reached; number of Cycles or % capacity left.
Do a two-stage internal resistance test after every cycle to be able to Trend the increase in internal resistance
Record all data for user extraction after the test is completed. Probably accomplished via an SD card.
Use hobby level components or breakout boards. probably Arduino or esp32 depending on the size of the firmware and interface requirements. Probably use an external ADC like an ads1115 to get higher precision, fully differential inputs and a built-in  PGA.
At the end of the collaboration produce a design spec with theory of operation, full schematics, bill of materials, programming and operating instructions, Etc.

If there is enough interest in this project I'll set up a different thread on this forum for discussion, and create a collaborative Google doc with the full user requirements as I currently Envision them just to see if it gains any momentum. PM me, or email me at TerryJMyers AT gmail,