EEVblog
ฝัง
- เผยแพร่เมื่อ 28 ก.ย. 2024
- Inside the Siglent SPD3303X Precision Programmable DC Lab Power Supply.
amzn.to/1m9Knp2
UPDATE: There is an even cheaper version, the SPD3303X-E which has 10mV/10mA resolution.
amzn.to/1jNsQRL
Atten PSU teardown: • EEVblog #440 - Atten P...
Rigol DP832 teardown: • EEVblog #511 - Rigol D...
Forum: www.eevblog.com...
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.c...
T-Shirts: teespring.com/s...
Donations:
www.eevblog.com...
Projects:
www.eevblog.com...
Electronics Info Wiki:
www.eevblog.com...
💗 Likecoin - Coins for Likes: likecoin.pro/@...
the 2020 version : banana connectors now with correct hole. the wires are now secured with cable ties, the too hot bridge rectifier is now 56C, no more screensaver, both setpoint and actual are displayed at same time. idle power usage 12W all outputs off.
Thanks for the info! I'm thinking to buy one
Thank you for the updated info!
There's newer firmware too, right? Maybe fixed some of the annoying UI details.
I've got yesterday (5/2024) the same model the X-E and the terminals distance are not ok. Do you think that I've got an old stock? But 4 years it's not reasonable.
@@netlogic1ify they are all like this, and it is not a huge problem, since today we all use standalone banana cables
I love the way Dave is so up-front with these reviews / teardowns, even after having the CEO of Siglent in the lab this week. lol
No holding back. I like it. :D
Message to Eric of Siglent, make all the improvements mentioned in the video and add $10 to the cost if that's what is required. It will benefit you in the long run!
Siglent is rubbish like all other Chinees electronic products. My 2 Siglent supplies (bought in 2015 having problems).
@@yoramstein bruh marco reps never had problems with his supplies. Maybe you received the dodgy ones.
@@DoctorThe113 Marco Reps will. Even their 4mm banana plugs and their Crocs are sh...t. Probably you work for the Chinees government . One thinks he got a deal with Chinees product . end story is that he paid more cause the rubbish will last 10% time of the American. Japanees products are different story. I make a video soon showing what wrong with my supply (encoder cause voltage jump to highers level, Transformer that zooms terrible all this after maybe 40 houre of use. DONT WASTE YOUR MONEY on Chinees equipment ! They are worthless ! NOT SIGLENT, NOT RIGOL amd sure not the small un known manufacturers.
@@yoramstein but you do have Chinese oscilloscope, don't you? And I don't work any country. What I do is just hate the hatred toward Chinese product. If you don't like Chinese equipments then don't buy it. You buy it because you can't afford the great sh*its. Like keysight (although they're truly trustworthy).
If you have problems with psus then why don't you upload your complaints at their review campaign. Maybe they will improve.
Siglent is relatively new company. They don't have the same experience as other companies. I admit that some siglent can be dodgy but keysight can be too. Keysight 3000x scopes used to waste 6 watts of power while it's turned off.
Instead of asking people not to buy it maybe you should head to their honest review campaign.
www.siglenteu.com/news-article/win-siglent-x-scope-or-x-generator-just-for-giving-us-a-review/
@@yoramstein fun fact: did you know that agilent rebadged rigol ds1052e as it's own? Dave jones said it not me!
th-cam.com/video/w6MdhoV03e4/w-d-xo.html
shorted turn on toroid transfomer! ooh, something else I wasn't aware of! Thanks +EEVBlog! learning more everyday!
the chassis is not laser cut, the deformation at the edges shows that it's stamped/shear-cut
Very helpful video, thanks. Sigmund has since changed the banana plugs and other than the spacing which has no meaning for me they are OK now. So I just ordered one of these. I also looked up the data sheet for the Lelon capacitors. The 6800/63 is rated 105°C, 3.6A ripple current, 3000 hour life at full load @ temp. That is spec for spec as good is the Nichicon, Panasonic, or CDE caps. I know you are not a fan Taiwanese capacitors but in the absence of any data to the contrary and if you can believe the specs, that’s about as good as it gets for a filter cap.
Mark, is there a date code on the supply you purchased that has the banana plugs updated?
Great vid, as usual. The poor quality of the front panel connectors is very surprising. And failing to use the standard spacing between them is nuts. I wouldn't buy this thing, just because of those problems.
The whole digitization of the interface is a mistake on a PSU. It should remain simple and 99% analog. An ARM microprocessor in a freaking PSU give me a break.
2:08 "I hate it when the hole is too small to poke your stuff through"
Yeah, it's really annoying trying to get your *stuff* in small holes ain't it Dave?
I don't get it
Ichabaud Craine Think of it like this, what happens when you plug a male banana plug into a female plug?
I thought small holes were best holes?
Dave at his finest... objective with good valid critisme.
* lack of bind post standard
* Way to loose fitting
* Display not incl. helpfull adjustment info & limits.
Though the trend / grpahing on the units display, looks nice..
and the colored graphical display, do like that.
Those banana plugs are inexcusable, the rest of the instrument is irrelevant.
+Andy Lawrence Well those 90+°C diodes are not so irrelevant I guess...
+Krisztián Szirtes While not exactly good design practice, a lot of power semiconductors are rated for a junction temp of 150C instead of the usual 100C that you see in ICs. Is it going to last for 30 years and still be just fine like an old HP power supply? Nope. Is it gonna die in a month? probably not.
+CKOD I am not concerned about the diodes, even if I had baked a few but more about the PCB itself. The epoxy or phenol based or whatever PCB that is just burns and bends, that way ruins the traces and might even crack or completely cut them. The diodes are fine, but the PCB manufacturer probably didn't use "/godmode 1" when made the boards.
+InXLsisDeo As I look back at the video, at 25:58 it's more like the redesign is a must thee.The electolitics will dry up in no time considering what the full lifetime of the thing should be.
+InXLsisDeo Ha, as I see I'm not the only one that gets the """ and similar shit after editing the comments :D
Reviews like this make me smile when I recently picked up a BK Precision 1660 (yup old but in great condition) for $100 Cdn :)
Those insulators don't break the shorted turn. One would need to be behind the bolt head to break the circuit, not between the transformer and the mounting flange.
Also no way they're running Linux on the STM32F207VE with its 512k of flash and 128k of SRAM. This is an M3 microcontroller, not a big ARM SoC.
+ollopa1 Dave said that he could see an insulating washer under the bolt. I couldn't see it but my eyes have already seen their best days.
+ollopa1
Maybe it's a chopped up version of Linux?
+Benjamin “Ozias” Esposti It's not running Linux.
ollopa1
Okay if you say so :P
actually there IS linux kernel (uCLinux) for Cortex-M3 chips and it may run on STM32, but still it requires lots of program and data memory, can't run on standalone STM32.
Perfect timing Dave, was just choosing between Rigol and this Siglent. Ordered Rigol for it's display and regulated ch3
Feel-o-vision, love it.
Love your work, man.
You pronounced "Lelon" incorrectly.
It's actually L'Elon, Elon Musk's newest high-fashion, high-capacitance offering from his newest company, based in Paris.
I'm looking forward to the upcoming release of their new line-up , including "Eau de Fumée Magique" ("Eau de Magic Smoke") and his clothing line-up "Vapeur Vêtements" ("Vapour Wear") but if industry rumours are to be believed, the latter one might not materialize.
However, I doubt this will be too much of a loss, considering his flagship product, the eponymous fragrance "Elon Musk" is still selling well.
Dave, you should put it in graph mode when you connect it to the PSU killer, would be awesome to see what it shows when it goes full panic.
If they consider repairing the banana plugs issue, screen saver and out of range message, and also that possible hotspot on that rectifier. I would buy it.
I would'nt be let down by an out of range message or a screensaver that can be turned off in the settings ;-) But the binding posts are indeed very annoying.
This is an old video but I think it's worth giving an update on the SPD3303X-E that I just bought.
First of all, I am not seeing any of the Siglent rust in there, it's all nice and clean steel sheet in there.
Second thing is about the mains wires. They clearly have seen this video because the main wires are no longer ty-wrapped together with other wires, the no longer run along the sharp edges of the steel sheet and they now also have a very rigid extra plastic sleeve over the wires for added protection.
edit: Third is the UI, the new UI still shows the timer but in a small font and the display now shows both the settings for voltage /current and the actual voltage/current. Best of all: no screen saver anymore!
The only thing (which I don't remember you mentioning) is the lack of a sense input for both channels. It's a bit strange to have a power supply that can supply 32V @ 6A with a 1 mV adjustable voltage level without a sense input.
The binding posts quality and non-standard separation is a serious issue. It will stop me from buying it for sure.
Good video. CH!
It seems like decent unit, but few small little touches and it could be a quality unit.
I hope Siglent is watching :)
Those fixes wouldn't break a bank for them to do.
Non standard distance for leads? YOLO!! Cheap caps? YOLO!! Scraped mains lead? YOLO!!... Non twisted-pair ethernet bridging?? YOLO!!!!!!
+Diggnuts Yolo, you get what you pay for......
+Diggnuts YOLO!
I know it's a little off the cuff, but I'd like to send you a product for review that is a bit of a departure from your normal reviews. It's an open source BBQ temperature controller called the Heatermeter. It monitors temps, and controls the temperature of a CHARCOAL smoker/grill with a fan and automatic damper system.
Digital knob is really poorly sensing the up/down when quickly scrolling the knob.
I find that a deal breaker and I also prefer instant voltage change.
Why would it be better to wait a second to change it!?
This only makes it even more annoying to set the voltage/amps!!!
And thank you very much for a really good review!
Couple of notes:
1. STM32F2 series is absolutely not running Linux. It's a standard Cortex M3 microcontroller.
2. You can't hear it switch when you go through the voltage range because the default current limit is set to 0. You have to increase it before the output voltage actually increases.
Love that relay oscillation...
Dave, you constantly talk about the capacitor brand, and how one brand is "better" than the other. What do you mean by that? Are you talking about electrical specs, or the lifetime (or both)? Is it something measurable? Can you come up with an experiment that clearly shows that Rubicon caps are better than Lelon caps? How would you test a claim like that?
Both,yes, failure rates are recorded and you can see them, use some caps hard for years and examine them electrically and look at the failure rates and they are better lost of the times even in specs.
It's not the brand that matters, that's just personal bias (which Dave does very well).
The specs of the capacitor is what is more important. Rubycon makes cheap low-end short-life capacitors, and Lelon makes really high quality capacitors. Download each brand's "Series Chart" and you will see the vast range of characteristics that each brand supplies.
That's punched sheet metal that leaves a burr on one edge. One is supposed to run a deburrer down the cut edges but it's often neglected. Also I've seen some lovely colour lcd drivers for the STM32 that wouldn't need any fancy OS. I guess they're probably only running a basic RTOS in that thing. Bit worried about all that clicking though.
Hi, Dave. Love the great work. Bought the siglent PSU because of the bang-for-the-buck. Happy with it except for the non-std spacing. Don't get that at all. But, when I got it, one side of the units paint was delam. I took pix to share if you'd like.
KJM
Agree on the binding posts. Wow, unreal!
This ist an STM32F2, an Cortex-M3 i think. Nothing that could run a linux, maybe some low-end realtime OS like FreeRTOS.
+Torben Hellige heard of ucLinux?
Thanks for the useful review - it looks really similar internally to GPD-3303S from GW-Instek
I don't understand how it's physical possible to be smart enough to design/build a precision power supply and fuck up the FRONT PANEL POWER CONNECTORS!
+ElectricGears i think, they don´t *design* it, they just copied an already existing thing, like that on this little inboxed link, in this video....and made some changes due to competition...and off they go....it´s like as the audience already said...YOLO...
I would rather go for : RipOff or Fast money makers....
Dave made a good video, but from my point of view (as it would be *my* money as well) This thing is a true RipOff!!
Design origin is well in the end of the last century made up with a nice DAC and a newer µC and there you go....
My BOM would be: 40 for the toroid, extra 20 for these Control ReDesign, 50 for all semicon, 30 for PCBs and mounting, 20 for case, another 20 for buildup and test....rest is true : YOLO....
The "precision" relais only on and whithin this ADC, all the rest is "howYaDoin" no excuses here....
+ElectricGears
BTW, the EE's don't usually do the case design or parts sourcing, and in the case of the caps, the engineer might have used good caps, but the person who sourced the parts used the cheapest ones there are...
The is a thing called "white Booke" in where all the specs are nailed into....and if i´ve chosen a specific component to be a needed one, i usually wrote it into as well, just to be save that this thing would not be repalced by a cheaper one....;) We EE desingers know that as well ;)
TubiCal
Yeah, unfortunately it seems that not all EE's do that. Either that, or they are bad at picking good parts! O.o
ElectricGears because proper bananajack connections would add 30-40 cents to BOM which means less profits.
Dave, from the picture the bolt through the toroidal transformer unless it's made of plastic (doesn't seem so) IT IS CONNECTING through the chassis, so there seems to be a shorted turn there!
+Kris TheTrader Not if it has an insulated washer, which it does appear to have.
+EEVblog How can a washer break the loop? To me it looks like if you cracked the threadlock it's connected. I couldn't let it go and lost attention after that. On the outside of the bracket where the nut is?
+Brek Martin Same thought over here... I guess we will never know now... :-)
@@BrekMartin - It's VERY simple... the washer electrically insulates the bolt from the metal. Think about how you connect a power transistor to a heatsink.... there is an insulating washer so that the bolt doesn't electrically connect the case and the metal tab of the device.
The washer is clear plastic, and has a shoulder than fits into the hole in the sandwich plates. There is no electrical connection between the bolt and the 2 side plates. Dave has already explained that there is washers, both in the video and in the comment... why can't you comprehends such a simple thing?
@@johncoops6897Brain damage, and a poor attitude probably.
I watch these videos even though I don't have the slightest clue about what it is or what anything on it does. Make more!
lmfao " i hate it when the hole is too small to poke your stuff into" epic
30:56 Default screensaver are great. Once on the used market, you won't end up with tons of burned-in displays. Usually if you go away, you won't be watching it anyhow and if you set a reasonable time-out it's just fine.
TFTs does not burn-in.
@@gabiold - No, they just get dimmer and dimmer as the backlight fades out..
Just pointing out that it's not Linux or any "high level" OS running on the STM microcontroller. It hasn't got the memory, nor the processing speed; it's a Cortex-M3. It might do the display rendering (drawing fonts and such), but there might be another actual controller. As for the JTAG header, the STM also supports JTAG-SWD, so it might be using the 20 pin connector, but 3 pins would be enough too:)
I've seen cat 6 fail completely due to a centimeter or two of untwisted pair where the installation tech has unwound it to make it easier to punch down. A long run of untwisted pairs, inside a high-emf 50.60hz environment, surprises the pants off me. I'm guessing it's a pretty slow connection.
The layout of PCB boards, heatsinks, mains transformer in this power supply is very similar to GW Instek GPD-3303S. Even the chassis is the same type. What a bummer!
Siglent Listened to Daves advice @ 25:35 the modified the display just as he suggested
Those binding posts are a real killer for me ...
Siglent is the producer of ATTEN's Oscilloscopes, power supplies and more as it is the manufacturer of BKprecision and LeCroy
Siglent YOLO! Power Supply.
+Gabriel Fuentes LOL
i would say it would be a powered mic test where you have +/- 5 volts leg 1 and the same with the 2nd leg with neutral on the green, remember this what works on 1 set of equipment may never work on the next set you place it on
With the exception of board technology, this looks like my Instek GPC-3020 from 1980 something. The old toroidal is a hell of a lot better but I imagine they don't need great mains gear with digital control.
Almost 40 years and only one recap and a hacked in digital display and a little trickery to get 4-wire on the master side, not sure I have better than 4-1/2 digits because that's all I can measure.
I'll keep my old one thanks. Cost me $64 US.
Yep you're old one will still work when this think will fail if used regularly.
Dave it feel like you recieved a pre-production unit? Cant just understand why they would miss on so many important things like the binding posts that is no good at all hope they take your
advices and change to better quality parts Lelon caps common go for something better like Rubycon or NipponChemicon?!
STM32 running Linux with tens of kb of RAM.. probably not.
+donpalmera agreed, 0% chance there is a linux in that.
+donpalmera FreeRTOS all the way
Dave doesn't know shit about STM32F2s I suppose. Can't judge him, there are so many.
+billigerfusel I didn't bother too look it up.
+donpalmera µCLinux can actually be run on an STM32F2 series chip. even on the lower F1 series.
I've got yesterday (5/2024) the same model the X-E and the terminals distance are not ok.
Consider that you had the case open, so the through put of the fan was not cooling at it would if the case were closed. I agree about the display, they may have used a cheap display that could not get the resolution for the small details of having more on the screen would need. The Rigol, today, costs way more than this Siglent, if you took that into consideration, maybe the Siglent is a good deal for a hobbyist. The banana plug spacing has no room for forgiveness, I would like to know what they would say if you asked them about some of the more obvious issues, like the jack spacing, and looseness.
Yes, the banana plugs' spacing and looseness was not our finest hour! We did switch vendors on the jacks after this video came out and those were replaced by a tighter version in 2016. We were stuck with the non-standard spacing on this model but we have rectified that on the new SPD1168X models.
Yes please on the shootout with DP832, thanks!
I agree with your fear of shorted turns on toroids, but on a hefty transformer like this I would expect the bolt to melt the first time it was turned on, so there has to be an insulator somewhere.
BTW, I would have been interested in buying one until after I saw your review. The banana sockets would be the biggest downer. Pity, it looks quite good otherwise. Imagine, if they invested some $5 more they would have a really good piece of gear. instead they have something that I wouldn't buy for half the price.
Every time dave makes a power supply oscillate an engineer gets their degree.
street price... i've seen a lot of people trying to sell me something on the street, but usually nothing with a half life as long as a bench power supply. also holy shit Dave figured out what a spudger is! :D it looks like at the right load the life of the psu is limited to how many times the relay can trigger, which is sad because if it is linux like you said... they could have detected that, had it pick one or the other then pop up one of those warning boxes to tell the user you arnt getting quite what was requested for xyz reason
What a Siglent fail, thanks. I got the same crapping banana plugs on the cheaply switch mode variable supplies. I had to make some brass extenders/adapters to screw on the terminals and accept the full length of a standard banana plug.
Dave - Thanks for the video. I am a "bit pusher" EE so don't know squat about transformer design. Question about your comments about "shorted turn" at 5:05 and the bolt through the middle of the torroidal transform. Is this an issue because it could change the magnetic fields/properties of the transformer? I assume so as the insulating washer under the nut (very hard to see - not the big purple thingy) breaks the continuity.
Cemx86 Bolt & sheet metal bracket(s) create a magbetic current path paralleling the windings. Only it is a closed path, hence shorted. Putting a gasket between the toroid & bracket spaces them apart & reduces the effect.
+davecc0000 That purple gasket doesn't reduce the effect at all. As Dave mentioned, there is an insulating washer under the nut which eliminates the short.
Not standard dual banana distance is a deal breaker, what the?
so siglent watched this and made all the UI changes dave wanted about 1 month after this video was released... it now displays ouput and set voltage at all times etc
Seems so. Search for a comment by "rumblfmp" (around year 2000, perhaps 2 years ago) who describes the other improvements.
The mains input board was missing a screw! It was even labeled to have one!
YOLO, being a power supply afficionado, this thing has more similarities with the cheapies I tore down...
Dave - short out the 5v supply and see what happens! Does it blow the internal fuse? That's the killer for me, it should be current controlled.
Rob
Two noob questions:
1) So linear power supply is a good thing... Why? (I thought switching ones were the best thing since sliced bread since it wastes less energy, less heat etc..)
2) And what's the point of supplying power so precise down to single digit milliamps?
Nobody replied, so I will try to cover at least some issues (some 5 years later):
1. Linear power supplies are heavier and have some other limitations, however the power output is CLEAN. There is no ripple or electrical noise from the switching operation. Ripple on the power input can really upset sensitive electronics, and cause misdiagnosis (eg: where is that harmonic frequency coming from?).
2. The ability to regulate power *accurately* is important for troubleshooting and designing. It becomes more important the lower the power requirements of the connected device. 1mA is irrelevant to a 12V50W light bulb, however it is vitally important for a SMD LED that is designed to operate at 2mA maximum.
3. It's not just the precision, but also the ACCURACY, the resolution drift, the stability and the repeatability of Voltage and Current measurements. It's no point having 1mV precision if the accuracy is +/-5mV
4. Generally a Linear power supply is ALWAYS considered better than a Switching power supply. However they cost more and are large and heavy. You get better "bang for bucks" from a Switching power supply, but they are always an inferior solution. That said, for the same $$ you will probably get better features and raw "on-paper" performance from a Switching type power supply.
100°C Power transistor case temp... wow ! and I was panicking when I saw 51°C in my own design ;)
Dave, in the followup video U need to do a output measurements as for noise, impulse response, overshoot and so on !
Too small things to poke your stuff into? Like Mr. Siglent said: YOLO!
Looks like they forget to install a screw at 14:09. Overall Siglent quality is very questionable and price is premium. I don't think I would buy any of their gear.
Good to see a real transformer inside, and not a switch mode converter. This tells a lot about the quality of the instrument. Which in this case is very good with respect to the toriod transformer. Thanks for your review.
Replying to a very old comment.... It's not an indication of "quality" - it's simply because this is a LINEAR power supply. Most manufacturers make both Linear and Switching. They both have advantages and disadvantages. The Linear ones are inefficient and get hot, and will be massively large and heavy for even moderate power outputs. Switching type is smaller, lighter, less expensive and not get as hot, plus much easier to have large output current. Switching type also allows for a wider voltage range without adding massive complexity to the circuit.
However Switching type is electrically noisy, with high frequency ripple. Linear type can provide "cleaner" output, but may have other issues (like Dave showed) due to the complexity of optimizing the transformer tappings to ensure proper transformer loading, eliminate voltage drop, etc.
Hi Dave, I always love ur vids. Just wondering, and maybe it doesn't matter to you coz you can fix it yourself or whatever, but when you break the security seal on a brand new appliance, do you care about the voiding of your warranty? I know you joke about it on older gear, but... Cheers mate, and Merry Christmas to you and the fam.
Hey Dave
nice video. Can you do a video on the differences of the different Capacitor Brands? You always tell that one brand is the best but why and can u prove it ;)
Greets from Germany
Reply to an old post... Capacitor brand is not as much importance as the SERIES of capacitor. There are cheap and nasty Rubycon models, and there are good quality Rubycon. Same for Chemicon and Nichicon,w ho also make very high quality and also make low cost short-life models. Same for Lelon, and even for brands like Capxon who DO have good quality capacitors in their ranges. There are many many different types within every brand, each with various characteristics and compromises.
The main thing about the brand is "reputation" rather than anything else - this is more about personal bias, as well as knowledge of whether rated specs (eg: Lifetime) will be met.... those things cannot be tested or proven.
A good analogy is for cars - is an Audi a "better" car than a BMW or a Ford? Which is the "best" brand of vehicle? So that is the type of question that you are asking about Capacitor brands... nobody can answer that.
I don't like the power switch. It was designed to be mounted with screws but they just soldered it on the pcb. I don't think those joints will last very long.
The adjustment box on the display looks really bad on camera, Very hard to read.
Where did you got this nice blue "Fail" button? With a very attractive sound :)
I just wanted to scream about the Analog Devices ICs there but the next moment you just mentioned it :D
@eevblog im sure in setup menu u have option for screen saver and desactivate that
WOW, 100C on the outside case of that transistor?
I know transistors can handle that, but I'd rather see it at a lower temp than 100C ... maybe like 70C or so.
Dave, stock up on transparency sheets (a la overhead projectors) and craft a transparent cover for every equip you test. Tape on the clear cover, power on, and shoot the Flir.
Would you please do a video and show us low quality caps to high quality caps and their names as the graduate upwards in quality. Id like to be able to identify them by name.. Thumbs up!
The "quality" of capacitors relates to the SERIES of capacitors, not the brand. Search for another comment of mine here somewhere that covers the basics of it. The bottom line is that "low-grade" Rubycon capacitors are cheap and nasty. High-end Lelon capacitors are excellent. It's not about the brand, it's about the characteristics and most importantly what is the best product for your $ manufacturing budget. Personal BIAS makes for poor decision making.
Toroidal transformers also can make sound, they do not use solid core but laminated cores to and if they are poorly made hey will buzz.
+Miloš Lazović
Yes, that's right. Hum through a transformer is capacitive not inductive and interlacing makes them especially bad. The best transformer for noise is a split bobbin. It's the jerky audio guys who think equipment must have these transformers.
Rob
+Miloš Lazović there is a lot of holywar on what type of transformers is better... And for me it's not obvious that toroidal is better under all conditions. I heard they quite bad when there is a DC offset because the core is not grounded (although technically it's not an issue this rarely done). Hope Dave will make a video on this.
Alexandre Kandalintsev
All have its good and bad side, in most case I will agree that toroidal transformer is better, but they are more expensive (which is no big problem) and high magnetizing current (inrush current) which sometimes can be problem (you need to use soft start circuit for high efficient /power toroidal transformers), you can't implement magnetic shunt on them (now it is useless method, but was popular in transformers for welders)...
For me they are best for medium power/voltage, for small classical transformer is better choice (1000 KVA and voltage is >10KV).
What...??? 539 USD and those Banana-Plugs even *don´t* take ordinary bananas 02:31 ????
Go SIGLENT, go quick...hurry, do your homework!!!
This is it, for me, an absolute #NoGo !!! Regardless how cool the electrons flying around inside, i won´t even spend $53,90 for a LabSupply *!!* which does *not* take ordinary bananas!!
You are, by far, to generous with them....I would had stopped that video, and return that bloody thing to siglent!!!!
Hi, great video thanks, enjoyed it! Even decided to get one of these albeit the X-E version. A question have you tried to connect this power supply to Labview NXG or Labview? I am trying to use it with Labview NXG, but cant get it too match the driver to the instrument.
With a shorted output at a 3.198A limit, the feedback circuit is probably keeping the voltage drop across the power device the same no matter which voltage setting is used right? Unless it doesn't automatically switch taps as the current limit is reached.
Could the non-standard jack spacing be an attempt to prevent people forgetting about the internal relay output configuration?
Does Siglent have Hantek manufacture these for them. A lot of things point to that like the buttons, LCD and that red sticker on the LCD and the UI.
What's the advantage of a linear power supply? Just cost? Or does a SMPS inject noise into the output?
+TheTrueM3ga Certainly easier to design for.. Only line frequency*2 to deal with. Cost is not an advantage at all.. the transformer weighs more, and it’s a chunk of costly iron & copper!
+TheTrueM3ga definitely lower noise and perhaps a better transient response as a linear supply might have a wider bandwidth feedback loop. It's also easier to design a linear supply for a wide range of constant current and voltage, especially if you want to get down to very low voltages.
+TheTrueM3ga
Well, there are a couple of low ripple/noise ATX PSUs on the market (with an LLC resonance mode topology) - but they are judged with just 20MHz Bandwith...
So low ripple/noise isn't neccessarily true anymore.
But there are other disadvantages an SMPS has.
For example the EMV - due to the pretty high operating frequencys of those things.
You need way more manhours to design a SMPS than a Linear Power Supply.
You need a PFC circuit.
And you need better (and more) caps for them. Low ESR and even polymer capacitors are pretty common in high end ATX SMPS these days.
Besides that we are talking about a pretty low current power supply...
So you can possibly live with the losses of a linear power supply...
Dave had a SMPS 'bench' Multimeter a couple of days ago - that 600V 3A one I mean...
Those shitty banana plugs (can't believe someone passed through them on $500 PSU) and overheating (100 C) makes it a regular chinese box of surprises.
The relays probably don't switch until you activate the channel. Try that next time.
I get the feeling that the ethernet connecter
enables a lot more flexibility than the older
GPIB interconnects did. What kind of uses
would stretch the ethernet stuff and to what ends?
From the net to slave the PS and to the web to
provide graphs or status? How many outfits
actually use that stuff? Do server farms use these?
I hope Eric Qin isn't in any way offended by your fair and honest review.
Programmable PSU's are typically used for characterization and factory chain testing at all spec min max of supply and temp.
Perhaps a minor thing to some, but the banana jack issue is a bit of a let down. Companies have been building equipment like this for years and everything including the subtleties (like banana jacks) should be worked out by now. It's not like a new company doesn't have past designs features / techniques from other manufacturers to draw from. Yes it may be minor, but still frustrating. I see this all to often, even in other markets. A business will do something great for a decade and then suddenly it's like everything that it and others have learned up to that point get tossed out the window and some half-assed design comes along that makes you wonder. I'm sure people have their reasons for such apparent mishaps, but I can usually find none that are sensible. Put some thought in it and get it right the first time. It can't be that difficult.
One thing I would want a big obtrusive message for would be for when the thing goes from constant voltage to constant current. Sometimes a short in your device under test is survivable for an instant, but other parts in the device under test may overheat and fail if left that way for too long. For example:
Motor controlled by MOSFET fails with a shorted winding
Power supply beeps and says "CURRENT OVERLOAD!"
Technician hits the "output off"
MOSFET survives
2:12 What is the standard width?
The standard spacing is 3/4 inch. The standard ID of the jack is 0.161+/-.005 inch.
It seems like those binding posts would prevent me from looking any further at this unit.
Hey Dave. I'm wondering if you would recommend that to a hobbyist who needs a good psu with rail to rail so I can do light audio and opamp stuff. I'd really appreciate your opinion
That sheet metal isn't laser cut, that's punched in some way. Either in a punch or an eccentric press, I guess the later based on the fact that I not able to see any "repunch marks".
Why? because laser cutting doesn't make burs unless the operator is completely useless.
Look at the outer edge and you can see that it is bent a little, laser don't do that either.
Why do you like being able to turn off the screen on the 34461A et al, but dislike it on this unit? I'm sure the saved power is minuscule, but never the less it would be equal to the power of the keysight screen, if not higher given its larger size.
Because Dave absolutely hates brands that he considers "sub-standard" to his elite requirements. His oscilloscope and multimeter reviews are even more biased and petty.
Hi Dave. Can you PLEASE REVIEW QUALITY measuring and power supply devices!!!!????? What's with all this one hung low junk???
Most of the caps in that are Rubycon . Shame about the big filter caps though bummer .
Most of the caps are LELON, not Rubycon. The Lelons have Y on the top, the Rubycons have K on the top.
However the brand is largely irrelevant, since both brands make both low-quality and high-quality capacitors.
is the edge rust caused by it being cut by a water jet
I have had toroids HUM, for certain!
the best fisher-price powersupplu
Are 110V with 120V and 220V with 230V really that matter? my house mains voltage fluctuates between 215V and 230V.
Why do you say switch mode rubbish dave ? Are they that bad?
+Alperen Akküncü
Well, they can be.
Switch mode power supplies have very noisy outputs, and the noise is at a high frequency.
But I do like how they are smaller and more efficient! :)
thank you for your responses guys
have you done a teardown on owon odp3032 power supply ? I searched but maybe I missed the keyword, if not please do! :)