logo elektroda
logo elektroda
X
logo elektroda

Minitest: the cheapest UPS to maintain your computer after a power loss, thermal imaging, overheatin

p.kaczmarek2 1578 17

TL;DR

  • The GC PowerProof 600 is a budget UPS for keeping a desktop PC and monitor alive long enough to save work after a power cut.
  • It uses an offline design with a 50 Hz mains transformer, relays, an inverter, and a modified sine-wave output, plus USB monitoring software.
  • The unit tested is rated 600 VA / 360 W, includes a 12 V/7 Ah battery and LCD, and cost just under PLN200.
  • A desktop plus monitor stayed running for over 3 minutes at 200 W, but the transformer reached 100 °C and the app gave unreliable battery readings.
  • It also showed instability with a laptop power supply using active PFC, which may make the modified sine wave unsuitable for some devices.
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
📢 Listen (AI):
  • Cardboard box printed with GC PowerProof UPS 600 VA/800 VA and feature icons
    I'd like to invite you to a presentation and testing of the cheapest UPS bought with the idea of keeping a computer running for a few minutes after a power loss. Here I'll test it under various loads (including with a power supply with active PFC), I'll look at the output waveform, and finally I'll examine what heats up inside using thermal imaging. In addition, I will also test its PC-accessible application for monitoring load and battery status. Finally, I will find out where that characteristic overheating smell comes from when it is running....

    What is a UPS? A UPS (Uninterruptible Power Supply), as the name implies, is an "uninterruptible" power supply unit whose function is to ensure the continuous operation of electrical equipment (here a computer) in the event of a power cut. It works on the basis of a battery and gives you time to safely save your work and shut down your equipment in the face of a failure.

    We have different types of UPS:
    - Offline (standby) UPS - normally powers equipment directly from the mains and switches to the battery (with a short interruption) in the event of a power failure. The simplest and cheapest.
    - Line-interactive UPS - similar to offline, but has a circuit that stabilises the voltage without using batteries
    - Online UPS (double conversion) - supplies power to the equipment all the time through energy conversion, thus giving the highest quality and no interruptions

    In a similar way, power supplies can be divided according to the current waveform at the output. Here we have the cheaper power supplies, which generate the so-called "approximated sine wave", which I prefer to call simply "rectangular", and power supplies that generate a pure sine wave.
    - the approximated sine wave (seen here in the photos) - this is a cheaper and simpler solution by design. Such a waveform, although it works with basic equipment, can cause equipment to become unstable or shut down, especially with modern power supplies with active PFC, it also does not work with inductive loads, classic transformers.
    - pure sine wave - the waveform that is closest to that of the mains power supply. It ensures correct and stable operation of all equipment, including modern computer power supplies, audio equipment and equipment sensitive to power quality. Such UPSs are more expensive but much safer for equipment.

    NOTE: in descriptions of UPSs on vendor websites, you will often see a general statement that an approximated sine wave is sufficient to power a laptop/computer/TV. I disagree - at least my ROQ gaming laptop has a power supply with active PFC, and it effectively 'resets' the UPSs I've tested that give 'rectangular'.

    Here I will test the cheapest UPS purchased for office work with a desktop PC. We started the UPS selection by measuring the power required by the computer and monitor combo. I specifically overloaded the computer with a benchmark and added a considerable margin to the value, it came out to about 150 watts. I therefore marginally selected the model 600 VA 360 W , which I managed to buy for just under PLN200 , already with a 12 V/7 Ah battery and LCD display. UPS model GC PowerProof 600.

    So let's see what the courier brought.
    Cardboard box for GC PowerProof 600/800 VA UPS with device line drawing and feature list. Label on a Green Cell PowerProof 600 UPS box with specification table and QR codes. Power cable in a plastic bag with a desiccant packet, inside a box with white foam inserts
    The power supply is securely packaged for travel. There is also a moisture absorber inside to keep the whole thing from getting soggy. We also get cables.
    Rear panel of a UPS with two Schuko sockets, a USB-B port, and two RJ11 ports
    The power supply has two sockets on the back whose power is backed up. In addition, there is a USB for communication with the computer (with an app), and RJ11 sockets (filter).
    Black GC UPS with an LCD display on a table, with cables, a box, and instruction sheet nearby
    Interestingly, I was supplied with a USB cable and a cloverleaf type power cable, I wonder what it's for? Probably with my devices in mind already.
    Instructions:
    Black Green Cell Power Proof 600VA–2000VA user manual cover on a white surface Open UPS user manual showing “Information on Safety” and “Description of Product” sections Open UPS manual showing a troubleshooting table and diagrams of the LCD display and rear-panel connectors. Open UPS manual: specification table on the left and German safety information on the right.
    The first start-up shows that the power supply works and the display is quite clear.
    Close-up of a UPS with a green LCD display and a lit light bulb next to it

    I started by checking the waveform. Indeed, it is 50 Hz. I also measured under load - the picture shows 30 Hz, but this is probably a measurement error. I'll repeat the test when I get a chance.
    Rigol DS1054Z oscilloscope screen showing a yellow stepped square waveform and frequency readout Rigol oscilloscope screen showing a yellow square-like waveform and measurement readouts.

    I then checked the overload behaviour (several hundred watts) - the power supply reacts quickly and shuts down:



    Second overload test - 350 W approximately:




    Separately, I tested a high load configuration, up to about 250 W, with the computer + my laptop with active PFC . You can immediately see great instability in operation and power spikes:




    Finally went the actual test - computer, monitor, additionally loaded with 20 W incandescent bulbs. 200 watts in total.



    There was no PFC laptop plugged in, so the whole thing held steady for over 3 minutes with a 200 W load. It probably would have lasted a bit more than that too, I only charged the UPS one night, so it may have run a bit short of full.

    Application capabilities
    The UPS is connected to the computer via a USB type B cable and works with the GC UPS application available for free download from the manufacturer's website.
    GC UPS app dashboard screenshot: input 0 V, battery 0%, output 229.9 V, load 57%
    The main panel shows us the voltages (input and output), frequency and load. The load is shown as a percentage, which surprises me somewhat.
    GC UPS app screenshot showing Remote connection settings and an auto-start toggle
    In the app we have support for remote connection and notifications to email address via SMTP. It can be started automatically after login.
    SMTP mail server settings window in an app, with fields for sender, host, port, user, and password
    In the control panel there is more information about the device and the possibility to perform a power supply test. Unfortunately, here too the load is given as a percentage.
    Screenshot of the GC UPS app showing the control panel and UPS test status
    Device parameters - here is the battery voltage:
    Screenshot of GC UPS app showing input/output readings and battery status
    You can also schedule the UPS and trigger certain actions based on the battery status:
    Screenshot of the GC UPS app showing action settings with toggles, including “Shut down computer”
    This charge level is not reliably shown either, here I show screenshots of 5 seconds, the panel changes its mind and first shows "Damaged battery" and then 100%.... i additionally thought that there should be over-discharge (and over-charging) protection so that such a situation could not occur.
    GC UPS app screenshot showing “Damaged battery” alert and dashboard with battery and load readings. GC UPS app dashboard screenshot with “Damaged battery” alert and power parameters. UPS app screenshot: input 240.1 V at 50.0 Hz, battery charging, output 239.8 V, load 8% GC UPS app dashboard showing 75% battery and 225.1 V output voltage GC UPS app dashboard showing 75% battery and 225.1 V output voltage

    Interior of power supply
    Brief this time, I didn't want to do excessive disassembly. The whole thing is based around a 50Hz mains transformer, relays to switch from charge to power and an inverter.
    Inside an open UPS case showing battery, transformer, wiring, and small circuit board Inside a UPS: PCB with wiring and a battery in a black housing Inside a UPS: circuit board with relays, capacitors, and many colorful wires connected to white plugs Inside a UPS: PCB with relays, connectors, and multicolored wiring Inside a UPS: PCB with relays, connectors, and multicolored wiring
    Anyway, we already have a detailed description on the Elektroda channel:




    Heat-up tests
    What heats up the most on the board, does the power supply overheat? Checked with
    InfiRay P2 PRO , for two load levels.
    Heating test - 60W load:
    UPS app screenshot with power-flow diagram, battery charging, and readings 240.7 V and 50.0 Hz Thermal view inside a UPS with cables; hottest spot around 45.2°C Thermal image of a device interior with a transformer and temperature readouts 16.4°C, 34.4°C, 53.5°C
    At 60 W it is acceptable, the transformer heats up a lot, up to 60 °C after a minute, the shunt resistor on the board also heats up, no excessive heating of the transistors can be seen.

    Now 180 W:
    Thermal image inside a UPS: transformer at 88.1°C, measurement point at 24.1°C
    Here, unfortunately, it is worse, the transformer heats up to 100 °C, there is a characteristic smell of overheated electronics.
    Thermal image of device interior with wires and PCB, markers show 56.5°C, 34.6°C and 17.6°C
    The shunt resistor and the transistor section are also heating up, to around 60 °C.

    Summary
    Indeed, this UPS is able to keep the computer + monitor combo running for at least a good few minutes, allowing us to save our work and save our data, but this is where the revelations end. The vendor honestly admits that there is a "modified sine wave" in the output that some devices won't work with, although, interestingly, right next to it he adds that laptops will work with it - my gaming ROQ with an active PFC power supply is an example of how this rule doesn't always work. However, this is an extreme case, so I won't focus on that here. A slightly bigger problem for me is the excessive heating, although here it could be argued that even this 100 °C on the transformer will not hurt, as this power supply does not work in this state for more than a few minutes, and we get what we deserve for such a low price....
    Additionally, I am thrown by the unreliable readings from the app.
    In a nutshell, the pros:
    + the power supply passed the test with my desktop + monitor set-up and actually allowed work to be saved, additionally warning when the battery started to run out of power
    + dedicated application with data reading via USB, hassle-free installation, remote testing possible
    Minuses:
    - already at less than 200 W load the transformer heats up almost to 100 °C, even though nominally this UPS has 360 W
    - the dedicated app cannot meaningfully determine the % of battery charge, it literally manages to show "battery voltage too low, consider replacing" and 5 seconds after connecting the power supply "100% charge", and this on a brand new configuration
    Neutral comments:
    -> there is a "modified sine wave" in the output, so the power supply does not work with some devices, including devices where the power supply has APFC
    I will leave the final verdict to you however - is it worth it? Do you use UPSs?

    Cool? Ranking DIY
    Helpful post? Buy me a coffee.
    About Author
    p.kaczmarek2
    Moderator Smart Home
    Offline 
    p.kaczmarek2 wrote 14393 posts with rating 12314, helped 650 times. Been with us since 2014 year.
  • ADVERTISEMENT
  • #2 21864797
    TechEkspert
    Editor
    Posts: 7028
    Help: 16
    Rate: 5451
    Board Language: polish
    In this price shelf and in this type of UPS it seems OK, and you can also see a different PCB design to the model we used to test.

    I'm concerned about this change in output frequency, I don't quite understand it, after all it's not a generator set where the RPM decreases at load :) and the regulator has to stabilise it.

    With the so-called approximated sine wave there are some limitations:
    -voltmeters without RMS show wrong rms voltage readings
    -trying to connect a transformer-based unstabilised power supply will increase its output voltage (such power supplies are still present in some telecommunications equipment)
    -in SMPS they can increase the voltage on the capacitor behind the input bridge rectifier

    and what you wrote APFC in SMPS goes crazy....

    Here is a link to a piece where there are several such attempts:


  • ADVERTISEMENT
  • #3 21864858
    p.kaczmarek2
    Moderator Smart Home
    Posts: 14393
    Help: 650
    Rate: 12314
    I was also surprised by this frequency change, I already presented an old UPS on the forum once and when I compared the oscillograms for the loads there, only the voltage dropped. In this thread it is, at the end:
    Interior of an old computer UPS, running as a solar inverter
    Helpful post? Buy me a coffee.
  • ADVERTISEMENT
  • #4 21864866
    BANANvanDYK
    Level 42  
    Posts: 7618
    Help: 1148
    Rate: 2533
    Board Language: polish
    In this UPS, the output transformer is both a power transformer and a battery charging transformer.
    I suggest you do a test with two wattmeters, or at least one with the UPS running at idle (no load).
    I know from experience that transformers with the UPS plugged in 24/7 can heat up quite a bit, which means wasted electricity.
    In addition, the transformer is close to the lead-acid battery, which negatively affects its durability.

    In the old APCs, which had a separate transformer for power and charging, this problem was not noticeable. The UPS itself drew a few watts depending on the battery charge.
  • ADVERTISEMENT
  • #5 21864885
    zgierzman
    Level 31  
    Posts: 1776
    Help: 108
    Rate: 1537
    Board Language: polish
    >>21864858

    Maybe it's an oscilloscope error, because the cursors (although set incorrectly) show the same thing - about 15 ms. You can't see the scale in the second photo, but it looks like the period is about 20 ms (four bars) in both photos. If the period were to be 30 ms, there would be six grids, even a blind person would notice without a scale.
    The run is jagged, it's Rigol that got lost, the cursors are more reliable.
  • #6 21865148
    mkpl
    Level 37  
    Posts: 5637
    Help: 306
    Rate: 839
    Board Language: polish
    I will defend the manufacturer a little.
    In the inverter, the output frequency has no way of changing.
    The change in frequency is due to an incorrect oscilloscope measurement (automatic measurement and cursors). The cursors are misaligned and do not cover the entire period of the waveform.

    The active PFC module should be capable of operating with DC voltage. It then operates at full power (no mains voltage modulation) as a simple boost.

    Loading such a UPS with full power, well, it should be able to withstand it, but here the matter is a bit confusing. It will probably handle a resistive load, but a capacitive load, including several tens of var of reactive power (filter from the power supply) or inductive power is a different matter.
    Power is usually given in VA, but the fact that it is written in W is another matter.

    The power consumption of a computer also varies and the average value of the power is much lower than declared.

    Where might this work nicely? -> 3d printers. I myself use with an old APC printer. Data loss on a computer is rare these days because hardly anyone works without auto write. With 3d printing, however, a power outage can put a few hours of printing (and sometimes even a few days) :)
  • #7 21865154
    adambyw
    Moderator of Acoustics
    Posts: 6694
    Help: 1120
    Rate: 828
    Board Language: polish
    UPS is cool, but the app gives a mega bummer. It's terribly resource intensive, can take up over 2gb of ram and up to 30% of CPU resources. A bit much for a background application.

    Windows Task Manager snippet showing “gcups (4)” using about 23.9% CPU and 2822.4 MB RAM

    bdw, maybe you guys can suggest how to de-gear it.
  • #8 21865160
    p.kaczmarek2
    Moderator Smart Home
    Posts: 14393
    Help: 650
    Rate: 12314
    @mkpl Useful idea with that 3D printer.

    @adambyw What are they counting that much RAM in there? Maybe there's an open source alternative? And if not... i might think about making some. I haven't checked what it "talks" over, but it's possible it's over HID. I wonder how complex a protocol they have.

    As for power backup, I've still been testing something like this - a low voltage DC UPS under a router and a Raspberry for example. Details will be in a separate topic.
    Hand holding a black IPS device with power button, battery indicator, and a USB port.
    Helpful post? Buy me a coffee.
  • #9 21865225
    puchalak
    Level 19  
    Posts: 1275
    Help: 11
    Rate: 206
    Board Language: polish
    p.kaczmarek2 wrote:
    ...- the dedicated app cannot meaningfully determine the % charge of the battery, literally manages to show "battery voltage too low, consider replacing" and 5 seconds after powering up "100% charge", and this on a brand new configuration...


    Generally, it is very difficult to determine the charge of this type of battery, in this type of device, especially since, in fact, by the charge level, one expects to know how long the ups will sustain power at a given load. And it usually ends the same way - after a few years of blissful certainty that we have emergency power, in the "w" hour the ups runs on battery for 15 seconds, or sometimes not at all.
    I remember the old, very expensive ups, probably APC smart series, which once in a while carried out battle tests on the battery, updating themselves with information about its wear and tear and the expected duration of operation. The problem was with replacing them, because some kind of memory erase procedure had to be carried out afterwards.
  • #10 21865289
    Mateusz_konstruktor
    Level 37  
    Posts: 4150
    Help: 266
    Rate: 1099
    Board Language: polish
    puchalak wrote:
    I remember the old, very expensive ups, probably the APC smart series, which once in a while carried out battle tests on the battery, updating itself with information about the battery's consumption and the time it was expected to last. The problem was with replacing them, because some kind of memory erase procedure had to be carried out afterwards.

    Right.
    @p.kaczmarek2
    Or did you have to perform a battery "touch-up" operation at the outset?
    Does the instruction manual say anything in this regard?
  • #11 21865422
    p.kaczmarek2
    Moderator Smart Home
    Posts: 14393
    Help: 650
    Rate: 12314
    @Mateusz_konstruktor acct I always take care to read (and post) the instructions - you have clear pictures in the subject line showing what was included. Point 2 talks about the first charge.

    @puchalak I don't have that much experience with UPSes, just my subjective impression was that when a new one (bought to plug in at a small business) alternately shows "100% charge" and "consider replacing" I think there's something wrong. But in the end it does work.... so far.



    And with this app and USB, I'll think about it.
    Helpful post? Buy me a coffee.
  • #12 21865430
    Mateusz_konstruktor
    Level 37  
    Posts: 4150
    Help: 266
    Rate: 1099
    Board Language: polish
    p.kaczmarek2 wrote:
    my subjective impression was that when a brand new unit (bought to plug in a small business) alternately shows "100% charge" and "consider replacing" then there is probably something wrong

    This is reminiscent of the situation found in mobile phones.
    By this I mean shutting down at the moment of higher load, despite even a fairly high percentage charge level presented on the camera display.
    It is a question of too much internal resistance of the battery, and this can occur quite independently of the correct capacity.
    The device presented is made with an emphasis on a low end price, so a poor quality battery may be present here.
    In my opinion, it would be worth repeating this check after installing a different and certainly working battery.
  • #13 21865636
    Olkus
    Level 32  
    Posts: 3908
    Help: 132
    Rate: 909
    Board Language: polish
    Thanks for the test.
    I was thinking of buying a similar one but with more power just mainly for the PC. Maybe worth the price though, because something too cheap seemed to me to be able to work well with a "more difficult" load, that is, especially APFC.

    Regards,
    A.
  • #14 21865689
    puchalak
    Level 19  
    Posts: 1275
    Help: 11
    Rate: 206
    Board Language: polish
    The battery charge indicator in such a simple ups, the only thing it can show is the voltage on the battery, so it is just an unnecessary gadget. The important information that should be signalled, and with normal mains operation, is that the battery needs to be replaced. This was usually realised by detecting a drop in charging current resulting from an increase in the battery's internal resistance. In the past, simple UPS's usually had 3-4 LEDs - mains operation/battery operation/replace battery and sometimes overload, although this was a bit nonsensical as in the case of an overload, the UPS usually switched off immediately. Now everything has to have a display, so it shows something there, no matter if it is relevant or not.
    The UPS is well worth buying, especially as at least 25% of the price is the cost of the battery. Just remember to check once a year how much it holds.
    On the other hand, on auction portals you can buy very good company ups for pennies, sometimes even with a sine wave, for which the only thing to do is to buy batteries.
    First from the shore
    https://www.olx.pl/d/oferta/zasilacz-ups-apc-...-ID1928Dn.html?search_reason=search%7Corganic

    https://www.olx.pl/d/oferta/zasilacz-ups-apc-...-ID14IzIM.html?search_reason=search%7Corganic
  • #15 21865974
    sq3evp
    Level 39  
    Posts: 6334
    Help: 210
    Rate: 833
    Board Language: polish
    puchalak wrote:

    On the other hand on auction portals for a penny you can buy very good company ups, sometimes even with sine wave, for which the only thing to do is to buy batteries.

    Back in the days of the fledgling Allegro, a colleague bought 2 APC UPS for a rack, something like 2kW each at scrap metal prices. It turned out that it was enough to buy batteries. Smaller versions of APC sometimes required someone to upload the settings with the parameters of new batteries. Generally a decent UPS and worth playing with. A colleague once had a UPS from a company that makes equipment for the power industry - battery bank outside, but it worked fine. The new one cost quite a lot but the used one was very good.
  • #16 21866073
    Mateusz_konstruktor
    Level 37  
    Posts: 4150
    Help: 266
    Rate: 1099
    Board Language: polish
    sq3evp wrote:
    Lesser versions of the APC required someone to time the settings with the parameters of the new batteries.

    Many UPSs require a process called calibration or touch-up.
    This is most often done through a programme from the respective manufacturer, and installed on a computer.
    A cable is required for this. Nowadays a USB like for a printer, previously an RS232 type.
    Replacing the battery itself can bring only a small improvement, and this is due to operating on previously stored data of an already used battery.
    Special cases are situations where the UPS has stored the status of the battery as 'defective' in its internal memory.
    In such cases, the undertaking of purchasing a used UPS with a view to merely replacing the battery and calibrating it, often becomes the reason for misdiagnosis of the fault.
    At the same time, in most cases the manufacturers' software does not make it possible to remove the status of a defective battery and calibration cannot be carried out at all.
    It is then necessary to act with a so-called service programme.
  • #17 21866176
    __Kai__
    Level 12  
    Posts: 17
    Rate: 5
    Board Language: polish
    >>21864295
    A piece of good work!

    My purely practical opinion is: stay away from this manufacturer.
    Justification:
    I have had five GC UPS of various wattages bought at a similar date (around the end of 2023). NONE of them survived the warranty period without repair. All were sent back, some came back "repaired" and others were replaced with new ones. A couple were repaired twice during the warranty period. The general disadvantage is that they cook the batteries, finishing them thermally. The fact of the warranty repair unfortunately didn't solve the problems, at the moment I have one such UPS left, the rest went to the electro-waste because they went bad again shortly after the warranty period. The fault manifested itself either in dead silence or in endless charging of the battery. On the plus side, the only thing is that the guarantor did a quote for a potential repair for free, but it turned out to be economically unreasonable because I had a new UPS from another company for the price of the repair.
  • #18 21884200
    sq3evp
    Level 39  
    Posts: 6334
    Help: 210
    Rate: 833
    Board Language: polish
    >>21866073
    This is how we used to repair APC UPS's - the service had a soft and to the cost of the battery was the cost of the repair.
    The total was 30-60% of the price of a new one because sometimes something more than the battery died.

    Simple EVER brands were enough battery - the repair was too expensive, and sometimesme the electronics died, sometimesme the trafo punctured on the primary winding, the price low and the quality too.
📢 Listen (AI):
ADVERTISEMENT