EEVblog #1023 - Rigol DL3021 Electronic Load Teardown

Dave tears down the new Rigol DL3021 Electronic load and has a play around with it.
emona.com.au/products/electronic-test-measure/electronic-loads/dl-3021.html#.Wb7yY7Kg9hE
DIY Electronic Load: kzbin.info/www/bejne/bqm7Y4aMmK6nhaM
Forum: www.eevblog.com/forum/blog/eevblog-1023-rigol-dl3021-electronic-load-teardown/
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: kzbin.info
Support the EEVblog through Patreon!
www.patreon.com/eevblog
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.com/crypto-currency/
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.com/eevblogstore-20
T-Shirts: teespring.com/stores/eevblog
💗 Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3

[@MitsumaYT]

22:00 Seeing how there is a green underline it feels like you can change the digit and then use the wheel to increment in those steps.

[@SeptrothFFXI]

Mitsuma's Animation and Stuff yep that's what ya do

[@jort93z]

you might be able to press the button down to change the digit. it works that way on a couple of similar buttons.

[@FlyingShotsman]

Yeah, I wish Dave would spend ten minutes with the manual before he turns on the camera. We'd all spend a lot less time shouting at our screens...

[@thenaimis]

"REVERSE ITALICS" ... c'mon Dave you're overreacting... "LOOK AT ALL THESE FONTS" ok maybe you're not overreacting...

[@ws_stelzi79]

The firsttime I saw the "reverse italics" I thought Dave had put on a funny lense which distorted the shot.

[@rwbishop]

With all new 'Ransom Note' styling!

[@mikeselectricstuff]

No velocity control but you can probably move the cursor position for coarser adjust. Unexpected jumps due to velocity sensing can be be issue on power stuff, though more so on PSUs than loads, so cursor+know probably not too stupid. Apart from the fonts, the different button shapes look pretty silly. As regards regulator heat, maybe they only tested at 220v mains.

[@tmmtmm]

Other Rigol products (scopes, function gens) use a smattering of different button shapes for visual effect and are still ergonomic. Assuming the physical button design was set in stone all they needed to do was use a cohesive set of fonts, make the decision to put the text next to the button rather than on it in some cases and use consistent margins. Either the industrial designer person was drunk this time or the work experience kid did it.

[@RedRider2001]

@@tmmtmm Why would you let an intern determine something like that? Makes no sense, unless he was the only one who speaks Chengrish.

[@convolvulaceae123]

My company uses a Rigol DL3031 load. I took it apart and looked at it too. It is no different from Rigol DL3021. MOSFET quantity 10 pcs. The same IRFP250N. Maximum current limitation is performed by the program. These transistors are controlled by the MC33078 chip without marking. The shunt has a resistance of 0.0005 ohms. From this shunt amplifies the signal, the ADA4522-1 chip is unmarked. Still without marking, 3 chip OP27G. I haven't identified the DAC chip yet. If anyone knows what it's called, write.

[@ericjohanson6423]

Probably everyone already knows this by the time I'm writing this in 2023, but nowadays many people have upgraded the DL3021 to a DL3031a with a simple software change.

[@radarmusen]

I think that the little green cursor under the value can be moved to next digits so increments or decrement will go faster when using scroll wheel.

[@vladimiralena9129]

Úžasný komentár. Koľko emócií. Cítil som sa ako pri nejakom športovom prenose. Vďaka. :)

[@Sjaak_PD1SRM]

I use this with much pleasure. of course are there small things that are different from what you expect but overall i'm happy with it

[@phpn99]

Pricing is not linearly proportional to the BOM. The cheaper device could be priced as a loss-leader, and the marging recuperated on the pricier items. You cost your engineering, manufacturing, inventory and support, SG/A, marketing and sales, and estimate the addressable market and the competition's prices. You then price items based on customer threshold points, not necessarily such that each item has the same margins. But granted, manufacturers have a tendency to unnecessarily multiply versions of their products, making purchasing choices that much more complicated.

[@dardosordi]

"It's just creepy" hahaha, that one made my day, thanks Dave!

[@rot_studios]

"Innovation or nothing" Seems the graphics designer took that slogan very seriously.

[@BrunoPOWEEER]

Thaaaaanks for showing this new Rigol electronic load... I was super curious to see someone playing with it, in action, specs and prices were a bit confusing to me as well, the price jumps between the 4 models are just craaaazy... they could've killed the market but seems like they rushed a bit with this one. The navigation between functions, setting up a test seems a bit difficult and not very obvious at all. Anyway thaaaanks for this first impressions video!!!! You ROCK the electronics industry mate... ohhh yeaahhh =]

[@yangyeshuai3686]

Do you think theres any reason to get this over some ZKE tech battery dischargers other than the accuracy of the equipment ?

[@Fake_Blood]

What a rollercoaster this teardown was.

[@ElmerFuddGun]

1:51 - No, the CR mode has the same ranges for the DL3021 and DL3021 *A* models. At least according to that graphic as there is no line separating the range values except to the DL30 *31* models and even they are the same.

[@KFuzz001]

Dave, you're killin' me! Its so entertaining to listen to you i can't seem to stop .. i have no idea what you're talking about but you say it with so much conviction i'm totally sold. I'm glad you don't sell used cars near me!

[@mikeselectricstuff]

I wonder what happens if you put it in short mode with a supply capable of a huge current - e.g. car battery

[@simontay4851]

Either the wires melt or the MOSFETs go bang.

[@mikeissweet]

Or hopefully it has some sort of protection. Should give this a try

[@CliveChamberlain946]

That sense resistor would die very quickly... ha!

[@paulvoid3247]

The whole thing becomes a fuse... unless there is a controlled transition into short mode.

[@nathanhaltman7235]

I want to watch.

[@hlokkAus]

According to the specs, the difference between the standard and the A is 0-40A (vs 0-4A) CC, 0-150V (vs 0-15V) CV, 2-15k ohm (vs 0.08-15 ohm) in CR mode, up to 30kHz (vs 15kHz). The Readbacks are similar on both except for the resolution. But the constant loads seem to be 10 times the range! Although the non-A version seems to do a lot lower constant resistance which is interesting

[@FLyyyT_]

What's a good word for "Graph" Plot? No... Graph? NO!! A picture of an eyeball? oh...yes!!

[@DoRC]

The binding posts aren't 4mm standard because the 4mm ones aren't rated to handle the juice that this is capable of sinking. I think their choice was good as it forces people to use actual lugs for their wire

[@DoRC]

Dick Fageroni right but I would imagine they couldn't use a connection that was rated lower than the device. Probably some liability issue.

[@carstenellwart3315]

They cannot leave that deciscon to the user. They have to built a device that has connectors thay will not fail at the rated current - otherwise they will have a lot of devices to repair within warranty.

[@DoRC]

Carsten Ellwart or worse.

[@JubulburbWuldrd]

nice, love teardowns!

[@Sjaak_PD1SRM]

A great piece of gear. work a lot with it

[@alimmi9]

10:54 "Rigol - Innovation Or Nothing" That's hillarious xD

[@CliveChamberlain946]

I'll take nothing.

[@rasimbot]

Mercedes-Benz "The best or nothing"

[@mikeselectricstuff]

GPIB on the menu, but no GPIB socket....

[@MichelPASTOR]

Looks like the people who designed the front panel also made the GUI

[@stargazer7644]

It has a USB-GPIB converter option for $346

[@mcuembedded]

Maybe the front panel designers never saw the back panel.

[@yaghiyahbrenner8902]

I was really looking forward to the ADC, DAC / Analog stuff but they rubbed of the markings. buggers! The IC's closet to the Mosfet's are just servo analog op-amps probably NE553N like the new BK electronic load.

[@FlyingShotsman]

Whenever I see "serious" instruments like this that contain ICs with obliterated markings, I immediately think that they've probably sourced counterfeit ICs and they don't want anyone to be able to tell. By removing the markings, they greatly reduce the chances that anyone can identify the parts as fake. Even if you de-encapsulate the die, you won't know what to compare it against. Really good ADCs and DACs from the likes of Texas Instruments or Analog Devices aren't cheap, after all!

[@yaghiyahbrenner8902]

maybe they copied some of the sub circuits from the BK electronic load. That's my take on this.

[@supernova86]

The 'A' versions are the colorful ones...

[@Daniel-hd7gq]

Has anyone found out why? 🤣 Maybe the colors are quite expensive.

[@uzaiyaro]

I said this on one of Mike's videos as well, but one of the nice things about Chinese companies is that if they're serious about carving into the market, they do tend to listen to what people like you have to say. I can't quite remember the name of it but there is a very interesting documentary here on KZbin about companies based in Shenzhen, which is basically the hardware capital of the world now - it is to hardware what Silicon Valley is to software, more or less. Either way, if you come to one of these companies with a problem, and offer them your feedback/advice/a solution, these Chinese companies have the unique ability to respin things very, very quickly, even if its a hardware change this can be implemented next week in some cases, Xiaomi is renowned for this. If Rigol is serious about their products, you might just find that if people like yourself find problems and offer up solutions, fixes may just be implemented, maybe even quicker than you think! You did a teardown of a Siglent spectrum analyzer a while ago that I just watched before this, the way they designed that, is a prime example of this attitude in action that some of the larger Chinese companies have. They clearly designed that thing with modularity in mind so if problems are found, they can implement fixes more easily in the factory and the field. Just my 2c, which is probably all it's worth.

[@hslim814]

you can use the left right arrow button to change the digit the wheel control

[@Boffin55]

Not italic, but oblique (or slanted). To be a true italic font the characters are actually different (compare a lower case a in Times Roman with it's italic cousin to see what I mean)

[@avejst]

Thanks for sharing 👍😀

[@gile849]

Nice video, man. Elektro guru :D

[@mikeselectricstuff]

Wh but no watts display in battery mode ?

[@Chupacabras222]

They used "wh" not "Wh". Upper case "W" stands for Watts. I do not know what lower case "w" stands for.

[@iwtommo]

Weighing the electrons as they flow through. Clever stuff.

[@stargazer7644]

Dynamic mode goes to 30kHz in the A version vs 15kHz in this one. And I think you have a few of the A series options installed as well.

[@kalhana1]

Even the BKprecision E-load has the "short" feature. I use it to test fault protection on DC/DC converters.

[@kentvandervelden]

Great video, looking forward to watching it completely. Have been wondering how this load compares to the B&K 8601.

[@kentvandervelden]

Think I'll keep my B&K 8601. Rigol unit looks fancy, but that's that.

[@martinsalko1]

Costs twice the price, and has one new feature. Is this made by apple?

[@martinsalko1]

The longer i wach this video the moore it looks like it was made by apple. XD

[@carstenellwart3315]

It has more than one new feature: - Frequency range in list mode is doubled - better slew rate - better readback resolution - Software features included in the a models that must be bought in the non-a models for extra cost: -- LAN Interface -- Digital I/O Option -- Readback Resolution -- High Frequency Option -- Hight Slew Rate Option i am assuming that dave got a non-a model with all this options preinstalled. So his device was not half of the price than the a-model.

[@Todestelzer]

Apple wouldn’t have released the load with such a front panel.... But for the price point I agree ;)

[@Sjaak_PD1SRM]

here is this DL3021 in use for two years now. it's a great piece of gear. happy with it. PD1SRM

[@widecast]

Push the knob in to move the green cursor to select which digit is scrolled with the knob.

[@ZylonFPV]

It seems really strange that they would have that two piece enormous heatsink with a fan to cool some components.. but they failed to provide sufficient cooling for others which end up at 80c.

[@thinktronics675]

Cool to see inside one of these puppies

[@phuang3]

On my dummy load, I can push the nob to change the cursor position. So, the lack of velocity control is not an issue. I think this rigol should do the same.

[@ZEROSTATIC72]

That's the CBS logo. Maybe Rigol is going into broadcasting.

[@SeptrothFFXI]

ZEROSTATIC72 it's to broadcast your current readings to the interwebs! 😂

[@Maskddingo]

That eye-ball button at the top tunes in a vertically squished version CBS when hooked up in The States

[@rocketman221projects]

If they release a firmware update and put a bigger heatsink on the regulators, it could be a pretty good electronic load. It seems like they rushed the firmware and left out a lot of features.

[@PeregrineBF]

The A version lists a faster current slew rate / Con Mode frequency range is greater. I think I'll stick with my homebuilt one for now. Got a huge heat sink off Ebay (120mmx120mm computer fans fit on each end, it's about 50cm long) & built one for 1kW sustained dissipation.

[@simontay4851]

Water cooling is probably best for 1Kw. How about using a kettle element?

[@tubical71]

when i was a repairTech back in the ProAudio rental biz, i built an 20kW dummy load with a bunch of washing machine heaters boiling oil...;)

[@caddyguy5369]

Hung, drawn, and quartered...There's a punishment we don't see often these days.

[@Bushougoma]

31:57 maybe the stick wasn't formatted with the proper file system you'd be surprised how many devices still require FAT32.

[@Firecul]

The majority of the time I use FAT32 on USB sticks just for the compatibility. I don't know anyone who uses anything else other than in special use cases. Maybe I'm rather isolated.

[@rasimbot]

I had Nikon camera which worked only if memory card is formatted with specific cluster size -- unlike pc which works fine with any cluster size

[@jackwhite3820]

+Firecul42 Just today I had to move files bigger than 4 GB with a USB stick. Doesn't work with FAT32.

[@Firecul]

Jack White anything bigger that the limit I'd first try to compress. If that didn't work enough it would be split into 4GB archives. As I said, the compatibility is the main thing I need from USB sticks. It's no good being able to fit an 8GB single file if the device reading it just can't handle it. And that's coming from the guy who has several 100+GB files he has had to move in the past.

[@mq1506]

I'm going to guess the plastic cover is for cases where the load is used for assembly line QC with a test fixture.

[@captiveimage]

Looked like quite a cool box for the dosh, the Std version anyway. With so many interface issues, weird feature anomalies and lack of accuracy I've been put right off. :-(

[@alessandrosaporetti9778]

As we are on the subject... Dave do you think that it can stand reverse polarity ??? my HPs E-Loads blows up like fireworks and in a lab specialised in testing batteries it happens ever too often.

[@xDevscom_EE]

There is GPIB option in menu. Where is GPIB connector? :D

[@ghislaingroleau7708]

Hi Dave, I was wondering if one can connect a negative voltage to a load: example a PC power supply has both a +5 and -5 Volts as well as a +12 and -12 Volts. How can one test the negative voltages?

[@ChristopherJohnsons]

@ 11:09 the silk screen next to one of the "linear jobbies" even says DP821A_3_V01.x (can not read the x) so Dave was correct to be suspicious

[@ElmerFuddGun]

Poor GUI programming. 22:18 - the pop up error message is way too fast. Even at .25x speed you can't read it! 24:00 - lower case "w" should be upper case for watts... at least they got "mAh" correct, I guess. 30:42 - 111683.055 ohms? Ya, sure it is accurate to 9 digits... WTF?!

[@TomStorey96]

ElmerFuddGun not to mention "? Battery test complete" Question mark???

[@0x8badf00d]

And saving an image is a warning message. And don't you just love the errors with exclamation marks but no reason. Failed to post this comment!

[@JimGriffOne]

Those reverse italics hurt my brain so much that it burped out some IQ. Now I can't brain electronics properly any more.

[@alimmi9]

I recommend shaking some electrons from upside-down electronics into your coffe. You'll feel much better.

[@ccmc919]

I feel dizzy when I look at the buttons.

[@ZacherYT]

Noticed the green cursor... is it movable with the directional arrows to adjust what the knob adjusts?

[@PapasDino]

Hmmm, think I'll stick with my ancient BK 8540...I was hot to get this but think I'll wait until they work out some of these ergonomic flaws.

[@willhelmx8388]

Hi Dave, regarding this load, is it possible to use this as a simplifed power analyzer ie: will it measure Amps in any mode going thru the thing, like hooking it up as a dmm.. BR / A

[@Chris-hy6jy]

Dave, have you seen the Tenma 72-13210 electronic load? 120v / 30A / 300w. It's basically a rebadged KORAD KEL-103 but these units are selling very cheaply at the moment in the UK from Farnell and CPC. RS have also just released their own rebranded version of the exact same load (part no RS-KEL-103). I've just bought one and it seems very good for the price. Would be good to see a review if you can get your hands on one :-)

[@tablatronix]

Wow that display is so much nicer than the DP832 PS

[@MatthewSuffidy]

The eye is like the cbs logo. Each button group must have had its own team.

[@monchiabbad]

I think those four so8 packages are covered with some black carbon like stuff please verify if they are actually lasered off and not simply covered with something to dissipate the heat.

[@DevilByte]

Hello! I always watch your channel! I recently purchased a rigol dl3031a. She seems to be quite bearable. How can it be calibrated?

[@Yamahapsr200]

35:57 AAA?? I don't think that's right ;) 40:03 so you'd use a AAA or even a AAAA Cell for 1A Loads, but not a AA? Dave? Did the Illuminati-Button do bad stuff to you ?

[@3ffrige]

What would be cool is if Rigol would just populate everything and the end user can purchase a software license later that would simply just enable the additional transistors. What would the added BOM cost be? Perhaps 20 or 30 bucks? My company (Keysight) does that with the various (Ixia) L2/L7 network traffic generator modules.

[@tomgeorge3726]

Hi Dave, can't see the NATA traceable cal sticker on your reference Keysight Electronic Load.

[@toddlaplanteVT85]

Dave, I would love to see a review of the DP711 Rigol Power Supply.. @ $300 US it's not a whole lot more than the cheap knock-offs you find on Amazon and e-Bay.. any thoughts?

[@gearstil]

Even if the battery mode don`t allow CV, CP I think you can test the battery in those modes, by using the general testing CP, CV modes available, am I right?

[@mikereuter9911]

Wonder if you add the two IRFP250NPbF if you get it up to 60AMP and 350 watts. Would be a great update

[@AmRadPodcast]

Got the willies on that one. Love the Illuminati mode! Head over to eBay and get an old Kikusui (unless you need all the fancy graphing stuff)

[@lawrencebillson6224]

I'm down with the OPP functionality

[@TheExileFox]

Next to the fan on the main board, there is a ton of via's - i am curious to why they do this

[@AGeekNamedRoss]

Dave's vocal pitch is inversely proportional to the quality of the part he's reviewing. Maybe we get that plotted over future videos.

[@SteelBlueVision]

The irfp250n MOSFET is not specified to DC, so, they cheaped out

[@DaveMcAnulty]

Looks like the x/y capacitors are unpopulated on the board??

[@teemoinvietnam1356]

23:03 it says "FERQ" instead of "FREQ"

[@friedmule5403]

I do not know if et works, but to read the "secret" chips text, would it be possible to turn the system on and when it gets "warm" then try to read via a thermal camera?

[@samarsaifi8766]

Please suggest right choice between rigol DL3021 & siglent 1021X-E

[@UnreasonableSteve]

So many of these gripes would be solved by open sourcing the software / allowing unsigned flashes so that the community could make changes to the functionality of the hardware.

[@stargazer7644]

But then how would they sell license key upgrades for $500?

[@bloodyl_uk]

Great to see that reverse italics trigger you EEV!!! :p

[@sanityd1]

Rigol engineers watch the Eevblog, the rest of the company doesn't though.

[@DavidLeeMenefee]

I have reconsidered purchasing this product. I just wonder what would be a good alternative? hmmm thumbs up... PEACE

[@chriskalkman3815]

That PCI-e board has 64 MB of RAM and 128 MB of Flash.

[@SuperZylar]

The a model has all the options, LAN interface, Digital I/O interface, High resolution readback, High frequency option (30Khz) and high slew rate option. www.batronix.com/shop/load/DL3021A.html

[@ats89117]

Looks like the reason the A model is $300 more is because it has a colorful front panel!

[@rubusroo68]

it's hard to tell if the deficiencies are due to your unfamiliarity with the device or its shortcomings tbh but great teardown nonetheless.

[@artelse]

Display seems skewed in its frame. I had a similar problem (and banana plugs that didn't fit in sockets) with a DP832. Send it back and got it replaced. Quality you pay for, but to call it innovation..

[@Jimmeh_B]

Gotta love a good tear down and review video. Recently I pretty much submitted to never being able to justify/afford the cost of Rygol equip. Having seen this I now don't care. Been doing a lot of research into data acquisition systems, USB based DSO's, things of that nature (Hantek's all the other chinesium cheapies) and it's a mine field! I'm getting to the point where I'll just end up building an acquisition system from an Atmel because the cost is prohibitive for "real" gear and the cheap shit doesn't have the resolution/accuracy/speed. Again, now that I've seen this... Rygol can jam it! I can't believe, given the number of fun vouchers Rygol want for one of these units, that what we just saw is what you get. Most of it's functionality seems as if it was either an after thought, or they did initially only have one model, but later decided that there should be an "A" model, and just cut out the functionality (why they would?? Make more money??). If I had committed to purchasing one of these, which at the moment is a fair whack of money for me, and ended up with that sitting on my bench! I'd be PISSED OFF. Actually, I'd be spewing forth fury of the like that Rygol customer support had never experienced. It REALLY isn't good enough! This in my opinion, is a real quandary for the hobbyist. There's a lot of equipment out there and we'd all love to own it all but again, the price is prohibitive for most, certainly is for me, and apparently, there's no guarantee that you'll get any value for the money spent. So what are you meant to do? Just very VERY grateful that good folk like young Dave here and many other authoritative figures across all disciplines take the time to create and publish vids like this one. Dave mate, you and other's like you really do a service for hobbyists all over and I would like to take this opportunity to thank you for that service for if I'd not seen several of your vid's and didn't find the forum you have created when I did (was about to purchase a Hantek), I would have already burnt a lot more of my hard earned cash, been completely pissed off, disenfranchised and possibly even turned right off electronics all together (though I doubt that will ever happen, once your bitten....). Cheers mate, keep up the good work!

[@Novous]

32:00 My guess from experience would be the reason your USB drives cannot be saved to, is they're formatted in NTFS (windows default) and most embedded devices can't use that because you have to pay Microsoft a fee so everyone just uses FAT32 up to (IIRC) 64GB, the max size for that partition format. Which is what MP3 players/cameras/etc do when you use the "FORMAT" option in them. Though sometimes, some chips just... don't "like" certain USB drives. One other quirk that freaks many out is drives that have multiple "drives" (1 is the main, and other drive letter shows up like RECOVERY/BACKUP/SYSTEM or some other name). Raspberry Pi's have all kinds of quirks with USB drives like that to the point they actually maintain "whistelists" of drives that actually work. Either way, a proper developer would have written a real message so you could diagnose WHY the image cannot be written. Invalid mode/No graph data? Invalid partition type? Cannot access USB drive firmware at all? Disk is full? (etc etc etc) Even better would follow up those with an easy error code number so the user can check the manual for common troubleshooting, or google it, or ask tech support with the clear error code. But, it's not like this thing is an expensive product made by a large professional company, so why expect the bare minimum in software quality.... ........................

[@romanhar4151]

80 Grad and this beside of the Elco !!!! - So is this the lifetime Counter?

[@moshly64]

Looks like Rigol has all the Gear and no f^&%in Idea ! Hey Dave, you missed the green under bar cursor under the digits, probably use the cursor buttons to move it to the higher digits and then try the big knob.

[@jnygaarddk]

Yeah, really disappointing that he makes an almost 1 hour video and don't spend more time thinking while making it and then misses an important point like this. I get that the first impression is good to have on camera, but doing this one take shot makes it stressfull and don't give proper time to think... Maybe wouldn't hurt mixing in some afterthoughts once you have looked into the main complaints.

[@eko653]

I am late a few years but can this load measure a batteries internal resistance accurately?

[@BMRStudio]

Bit complicated operation for such a basic testing. I hope they will do some fixes on the next firmware. But why they not ask the EEVBLOG community before the development is done? For real. Let say, drop the job to the forum, and then apply some discount on the pricing for members. Rigol?

[@SilverGreen93]

I really can't look at those reverse italics... My eyes are hurting.

[@NathonDalton]

Anyone know where I can get the cable he used (time: 28:25 - 28:46, ...) to connect the load to the power supply?

[@ciprianciprian5228]

Did it record anything on the USB thumb drive?

[@DAVIDGREGORYKERR]

Does the unit use a 16 or 24bit R2R network that operates as the load.

[@RandornCanis]

Those power MOSFETs with compound painted over, is there any compound under? You know, where it belongs.

[@davidwesterfeld82]

Great review, but I think Dave blew it with the accuracy spec. Rigol quotes +/- (0.05% + 0.05% FS) for the A models (which this unit seems to be). The full scale current is 40 A, so at an actual current of 1.0044A, the display should read 1.0044 +/- (0.0005 + 0.0005 (40)) = 1.0044 +/- 0.0205, or between 0.9839 A and 1.0249 A. It looks like they met their accuracy specification. The various usability factors uncovered here are more troubling, as is the excessive linear regulator temperature. And, of course, the reverse italic font.

[@brettski74]

Maybe they do just make one model and when they calibrate it, they decide whether to call it a DL3021A or a DL3021 based on the measured accuracy during factory calibration. Just a thought.

[@PlasmaHH]

Hm, looks like the 60A version has nicer coloured buttons, surely worth the price to pay ^^ At first I thought that the slew rate for the A version was better, but then I saw that you get the non-A version with an option for the same slew rate. Also the I/O stuff seems to be sold as an option for the non-A things here... o_O