Owon SPE6102 Power Supply Review & Teardown (SPE Series) | Voltlog 400

[@Lumi_nance]

I have the SPE6103 and ripple without load is 1.5mV rms not 28mV like shown. With same settings like in this video I only see a flat line. At 1A resistor load i have 16mV rms and 150mV Vpp. CC and CV are totally equal

[@dukeiiigmail]

How is the fan noise level? Just ordered one, finger crossed.

[@Lumi_nance]

@@dukeiiigmailThats my main criticism with this thing. its quite loud, even at no load. If you draw high current for several minutes, it sounds like a starting jet. I mainly use a fan less linear PSU unless I need high power.

[@pyromen321]

Owon is very good at making almost good equipment. That noise is crazy!

[@voltlog]

that pretty much describes it, although in the case of this particular unit, I would say "almost usable"

[@jstro-hobbytech]

You're right. Any owon stuff I've had, I've never been completely happy with. Even being a novice hobbyist.

[@m80116]

UNBEARABLE, returning mine. Screen UNVIEWABLE from my regular eye height when sitting to the bench. no idea where the selection is.

[@SidneyCritic]

18:02 The resistors between the transistors look heat affected, ie, brown. FFcossag modified a noisy Wanptek KPS305D PS years ago - kzbin.info/www/bejne/lZOTh5hjibJ4fJo

[@jaro6985]

Yeah that resistor is already overheating, it won't last too long.

[@de-bodgery]

Probably switching noise can be greatly improved with better output caps. I didn't see an ceramic caps and obviously the electrolytics are cheap.

[@MrGabrovche]

@VoltLog Thanks for your review! Allow me one correction though - at 22:30 you talk about current accuracy, but show again the voltage chart! I just got mine - will check the noise performance, based on your video!

[@avi-crakhome2524]

Upper left corner. Resistors in the hot part of the standby power supply are rated for 45V, but connected to a 310V circuit!!!

[@ericksonengineering7011]

Thanks for a great video. I like the high power, small size, and low cost. For that price and size you get a switching supply pretty raw output. No linear post-regulator which would help clean up the output noise.

[@andrewmorton9683]

What appears to be discolouration on R45 by the heatsink for the output devices makes it look like it is half-cooked already?

[@chrisg6597]

I agree. That resistor looks horrible.

[@johndevires5911]

Saw it immediately, and the other one to the left. Smells fishy??

[@andymouse]

Very nice review...cheers. (edit) I agree adding external smoothing is wrong its a deal breaker for me !

[@cryscrystiii7023]

The output noise reduction I did with 4 SMD capacitors X7R / 2,2uF/100V, replacing output electrolitic capacitor with good quality low ESR from Panasonic and large ferite core on the mains wires

[@aicisha]

thank you for sharing us this idea for improvement! I plan to buy the base model this month. Also ferrite cores on the output wires would help as well?

[@lplightportal]

Did you changed big electrolytic capacitor between positive and ground wires, with 4 SMD 2.2uF? How much noise did you reduced?

[@jonathanfulcher602]

The 5V DCDC also looks like it is layed out terribly. The inductor and diode loop area is huge with the feedback(&other) resistors between them and the chip.

[@de-bodgery]

The CN2 connector is the programming port for the micro controller. I have to wonder if there are any differences at all between the various 10a PSU's beyond firmware???

[@robbyxp1]

Great video. I picked up an old TTI EL302T for less than that a year ago, and i'm glad I did.

[@lukasgayer5393]

And now the question - is there any way to reduce the output noise? Better filtering capacitors perhaps? Maybe a little choke and ceramic caps?

[@batuhanyavasoglu5795]

I hope you continue the gophert panel project. Is there any update about it?

[@richaw42]

I second the question.

[@voltlog]

That project will likely never see the light 😐

[@superdau]

@@voltlog Asked that some time ago as well (with the same answer). Too bad, because the user interface of the gophert is really a little bit annoying.

[@lasersbee]

17:58... Is it just me or does the R45 power resistor look like it has already overheated and been discoloured ?

[@mareksvrcina5279]

Looks like a well engineered device to me.

[@lasersbee]

@@mareksvrcina5279 That doesn't answer my concerns about the R45 resistor possibly overheating.

[@mareksvrcina5279]

@@lasersbee a VERY well engineered device. Sorry.

[@mareksvrcina5279]

@@lasersbee You seem to have misunderstood. It was just a joke - It certainly isn't very well engineered if the resistor is already discolouring.

[@lasersbee]

@@mareksvrcina5279 My bad... Thought you were serious...

[@ACoTam2]

22:36 You are talking about current, but the table is about voltage accuracy. Can you give us an image of the current accuracy table?

[@voltlog]

Thank you for pointing to this editing error. Here are the results of the current accuracy test: voltlog.com/pub/owon_current_accuracy.png

[@drruncmd]

I love how they skimp cost on the BOM capacitor brands but use belt and braces in the input protection and input mains filter! Great video! Maybe show us some schematics for parts of the circuit like input protection and filtering section??

[@jaro6985]

Input protection is required to pass EMC testing, expensive branded caps are not.

[@beardedlion]

Hmm... Just took a look at a Velleman DSP-3005V (which is 30V/5A/150W) and most of the PCB indeed looks the same as yours. The main difference is that the two large caps (C6, C7) and some resistors are missing. I was hoping for an easy way to hack the firmware to at least go up to 12V @ 7A on this unit, but after watching this teardown I think I'm going to return it and look for a used alternative (or a better brand). Thanks for the video!

[@allenlutins]

I work in a lab with a pile of high-end power supplies, and it's unbelievable how many stick a chassis ground between the positive and negative terminals, prohibiting the use of standard 19mm adapters :(

[@guillep2k]

It's time for a new standard! 😂

[@Rapid-eraser]

Actually the note in the asterisk was mentioned in the manual in the ripple section :P I think the switch in the back side is a total deal breaker for a crammed desk like mine :(

[@zaprodk]

R18/R19 is not only discharge resistors, they also balance the voltage across the caps. Nice touch.

[@DamirUlovec]

22:36 Got an "error in reading" - while watching this part about measuring current (Voltage Accuracy, No Load). Goofs section.

[@steubens7]

I've got the SPE3103 and the current offset with an open output is .03

[@voltlog]

Uh, nasty! I feel like there has to be a hidden calibration menu that would ideally allow us to null those errors.

[@steubens7]

@@voltlog emailed them back in November about it and never heard back. haven't got to reverse engineering it yet

[@guidosawatzky116]

@@voltlog I think there is a calibration menu. Try to press : Display -> Memory -> "Knob" -> V -> I

[@voltlog]

@@guidosawatzky116 You are correct, that brings up a menu. I might have to do a follow up video on this.

[@guidosawatzky116]

@@voltlog That would be great. I've done a calibration&justage of the output and measured values and everything is now in the specs.

[@danstiurca7963]

For the price point, this thing is amazing

[@aicisha]

9:30 is there any power-up feature like holding down a button while turning the unit on? I mean to change the default display view...

[@gamerpaddy]

i have had a owon odp 3ch psu and it had a massive delay before cc kicks in, unuseable around sensitive electronics. i never seen this behavior on any of my psu's.

[@voltlog]

that is typically the case for these cheap PSU because they use a lot of capacitance on the output to compensate for the noise.

[@gamerpaddy]

@@voltlog it wasnt the capacitance, it also happened when starting it up from zero. the cc part was just made in software not hardware, which causes a massive processing delay.

[@frankhovis]

Check the temperature the mosfets get to - The one I just tested got to 85C before the fan even started.

[@nagasainathkoduru9995]

Imagin using a HF radio with this psu

[@reedreamer9518]

Yes I can imagine that. I live in an RV and power my ham radios off the internal 12V supply and as you may know feeding noise directly into the power circuit of a receiver can be more devastating than noise coming in through the antenna. The OEM power supply of my RV was extremely noisy (5x small cheep switch mode supplies in parallel), so I disconnected it and built my own; starting with a toroidal transformer, rectified and feeding a large capacitor bank, then regulated using a $50 20A synchronous buck converter, which gave me a very efficient DC source for running the 12V systems, and maintaining the batteries which also powers my ham radios. Now you might say, this is just another HF noise generator - and you'd be right. But since I built the power supply myself I left enough room inside the enclosure for a 20A second stage filter. Using my oscilloscope to measure the exact frequency of the noise and it's harmonics, I wound 2 series 50 uH toroidal chokes (31500) that feed a 50 uH common mode choke (FT-240-31) using 10 AWG magnet wire, which flattened the ripple and most of the RFI. Then to get rid of the remaining HF noise I used 2 capacitors in parallel (10nF & 100nF) across the +/- terminals, and another 2 sets from positive to Earth ground and from negative to Earth ground - with the Earth ground leads running on the outside of the steel enclosure. This results in zero noise on any band (AM to 10 meters) as seen on an SDR band scope.

[@nameone7183]

I bought one of these. After I found it having 1.5v noise spikes on 12v at 1A, I stopped using it. Probably ok as battery charger.

[@anermly1729]

I have the same PSU and measured with my Rigol MSO5354 oscilloscope. Couldn’t find as much noise as you had but still a little. Don’t know why you had this much noise but anyways. For that price it’s almost acceptable. Maybe more of their budget could have been going in some less noise instead of the quality of the screen.

[@steubens7]

had it powering something for a couple days and the on/off button wouldn't respond when it was done :O

[@aicisha]

22:37 you showed us the same voltage comparison table instead of the current table :(

[@woodrustandthings]

I only use it for electrolysis of rusty parts, worked ok for 3 times and now I don't get any V or A, is there anything that I can troubleshoot or try? It does not seem to be broken but just something that is not active... The fuse is ok also. Tnx

[@Djmatrane]

You have v1.3 and CPU_v1.1! I have v1.5 and CPU_v1.2.2 And on my version when I turn on the power supply the FAN is not running. Or if I have small load the FAN is running very slow. It seems that FAN has now Temperature VS PWM function. No noisy FAN at all. They solved this.

[@rokiedecentra9656]

price gets close to the Riden Power Supply Kits which have the P series claming low noise

[@theprepperfrog167]

Thank you, sir.

[@dpyles9396]

Front power switch is essential on a benchtop power supply or any benchtop test gear.

[@Todestelzer]

I would suggest to look into riden PSUs. I would never buy one without numerical key pads anymore too.

[@voltlog]

I have one, I don't like it. Partly because of it's output noise which is higher than that of Gopher power supplies and partly because I find it very inefficient to use that separate industrial PSU which comes with a high standby current.

[@piotrchec8156]

23:33 I found calibration procedure here: kzbin.info/www/bejne/gXfNn3uGns2Sfqc Thanks for the review!

[@antibrevity]

Adding capacitors to the output to quote a noise spec is simply cheating and voids the entire purpose of the specifications. This certainly doesn't help Owon's brand image. This ia a bizarre product as there's obviously been some effort made to add quality that might be comparable to RDTech's top model, yet someone dropped the ball entirely with this output noise. It's laughable to display 3 digits of measurement precision when the measurement itself is not accurate *and* there's up to half a volt of noise. Many Chinese companies follow this same pattern of designing and building an "almost good" product. Perhaps someone can figure out why the noise levels are so high or Owon can do something to improve it in a revision or firmware update.

[@shayan8040]

Hello, thank you for your complete and beautiful video.I have the same power supply and unfortunately the calibration is broken and I sent an email to the support. Send me a video for calibration is messed up,I send an email to support team and thay send to me a video for calibration I went through the steps correctly, but at the end of the step, it gets stuck on the last step and it doesn't work.must be show calibration is finnished but didnt please help me to fix it Thanks alot With regards

[@UpLateGeek]

Uhh, are those input capacitors in parallel or series? 1.41 * 220V = 310V, so unless they're in series that's above the 250V rating of those caps.

[@jaro6985]

If it weren't in series they would have blown up.

[@voltlog]

yes, series connected.

[@UpLateGeek]

@@voltlog Well that's a relief.

[@nidusa1]

is the software open source?

[@m80116]

I've bought Owon SPE3102 thinking it could have been enough for my 1.5 to 4.5V 1A devices... turned out the worst purchase ever. The noise floor (ambient) is the same as a small hair dryer at first speed. I spent 15 minutes trying to program it as the manual mistakenly said to push to confirm from list but you had to push AND HOLD. The unit blows dead cold air from the sides, my hands over the bench went numb in no-time. The rear button I suspect it's not even on the mains input as the unit continues to work for a good 1 second after it's been switched off. The visibility of the screen is appalling, the small numbers are unreadable and the selection from eye height impossible to tell because of the poor viewing angle. This is one of the poorest power supply I've ever tried. I was wrong thinking I could get one cheap... I also bought an Owon multimeter at the same time to give this company a chance, all it was required to do was to measure Voltage and Ohm with two decimal precision. It whines like crazy in any mode and beeps like a lion in continuity with no way of turning it down, you can only disable the beep which renders absolutely USELESS the function. This company is a JOKE.

[@RixtronixLAB]

Nice info, thanks for sharing :)

[@ianlouden7939]

Having purchased Owon products in the past (2 x oscilloscopes and a multimeter) I would not touch another, they all have software bugs and are not supported by Owon.

[@voltlog]

I did own an owon scope a good while ago and I can totally agree with you. I just thought things have progressed and a PSU couldn't be as bad. I guess I was wrong..

[@lmaoroflcopter]

Hmm. I have an owon p4305 here and it works great. Real shame this one doesn't live up to the brand really.

[@voltlog]

How is the output noise on that model?

[@michaelpalmgren3606]

R45 looks like it has been heated up a lot..

[@dl8cy]

I like my Riden 6006P with higher accuracy for the same price

[@therealb888]

And a little cap across the OP solves the noise issue

[@jstro-hobbytech]

I guess I'm lucky mine got lost in the mail lol

[@voltlog]

Oh, I didn't mention this in the video, but that happened to me too with the first seller I tried getting it from. I don't think it got lost though, he just provided a fake tracking number to eventually find out 2 months later that he never shipped the item. I suspect in the beginning it was difficult to source these or they were not yet ready for sale.

[@hyperbyte2]

That resistor is a bit tosted

[@y_x2]

For 120/240V input they simply use a very old technology that may save few cents in China only. For the rest it is only bad choice of technology.

[@SonicWawes]

Hello, I have noticed a small issue with your test setup. But please don't be bothered. Even Dave Jones from EEV Blog screwed this up. Can you please try to match the impedance of you test setup for voltage ripple and verify your results? I also add the Dave jones screw up for reference: kzbin.info/www/bejne/naHanJyAeJqbnLc

[@voltlog]

That's actually one of my favorite @eevblog videos but if you look at my video 24:42 my test setup is described and you can see there is no impedance mismatch happening.

[@SonicWawes]

@@voltlog Sorry, I did not notticed. But you might be correctly using the probe with 10x settings to match it. I just did not see it from the video.

[@voltlog]

@@SonicWawes I always measure with probe set to X1, that's my standard setup. The reason for using X1 is because I'm measuring low signals, I don't want that attenuated by the X10 setting. If any change, it should be in favor of the PSU by measuring at X1, because the bandwidth is lower for that setting.

[@mariolondyn50]

Welcome in world of chinese cheap pulse power suppliers , ha ha ha 🙃.

[@Gengh13]

A switching PSU that does not have universal power input? Pretty shitty design.

[@NiHaoMike64]

They were too cheap to put in the selector switch.

[@voltlog]

still trying to figure out the reasoning on that.

[@ianlouden7939]

@@voltlog the PSU variable output section is of a push pull design to maximise the efficiency of the transformer (much like a PC power supply), these require a centre taped point derived from centre of the 2 bulk caps in 230V mode this will be mid point (115 and 115V) but in 110v mode the PSU is configured to double the input voltage (2 x 1/2 wave rectifiers supplying each of the bulk caps) +110 and -110v. some PC PSU's have a switch on the back to select the voltage but if set wrong they would destroy the PSU. its either cost or to prevent damage they dont use a switch?

[@Chris-hy6jy]

Awful! And no mains switch on the front panel!! FAIL