The cheap $10 tester from China is useful - it does have an LCD screen. It shows LL on some voltages when you first plug in the main motherboard psu plug, but that doesn't mean a failure - it means that the other voltage it needs is missing. When you plug something else in, the LL goes away and it shows the voltage. Other PSU lines light up LEDs. I found the tool useful, and since I had several PSU failures, I decided to upgrade and purchase the Dr. Power II tester for about $35 - relatively insignificant in regard to the cost of a PSU, and in regard to the damage a failing PSU can cause. I had one PSU failure take out the motherboard - but not the cpu - and also it took out the full-size graphics card - if those items had been new, we could have been talking about $1000 worth of components. I have only had one other power supply failure in my 35 years of personal computer experience, and this was a one-year-old Corsair 450 running in a Pentium 5 4 ghz single core gaming computer. I called Corsair, as it was under warranty, and they told me to short the green and black wires - I did that and nothing happened. The power supply did not turn on. So they sent me another Corsair 450 and it was great. But as I say, I just now experienced the failure of 3 power supplies. #1 One was a Thermaltake Toughpower 750 that had been in service for 10 years - so okay, I got my money's worth. That one failed and took out either the 9450 quad-core cpu and/or the socket 775 motherboard (probably the motherboard) and since it was an old system, it was time to upgrade. My newer $500 GTX 1070 graphics card had not been damaged by that PSU failure. #2 Another was a no-name 550 watt power supply that came with a 2-core old computer - it actually passed both testers, and it would cause the ready light to come on with the bad smaller asus p5e slot 775 motherboard, but it wouldn't light up the ready light on the bad 6-core sabertooth X79 motherboard, which is now my main "psu load test." So the PSU is bad - it passes the testers, but not the actual load test of the sabertooth motherboard. That's why people will tell you - "a PSU tester will tell you if a PSU is bad, but it won't tell you if it is good." A load test goes beyond by actually putting a real load on the power supply. #3. The third PSU, a Corsair AV 850, came with a computer that I was able to get from a family member who works as a video game animator - he had one, a 2011 slot (not the year 2011, actually it has 2,011 contact pins) 6-core 12-thread that had been in his garage for 3 years, a big step up for me as my old 10-year rig was a slot 775 quad core 4-thread 9450 machine with a 30% overclock up to just under the standard clock of this 6-core. So I started testing his 6-core computer - he had pulled all the hard drives - I could boot to bios, and I saw the nice bios monitor showing the cpu temp climbing steadily, until it hit 80 within about two minutes, and then the 80 number turned red, and then the system rebooted. I was initially a little unaware of what was going on, I was busy checking out the bios – but I realized later that I certainly hadn’t done the cpu any favors by letting it overheat like that. When I “woke up” I realized that the water pump must have failed inside the Intel waterblock. I called up my close relative to chat - he tried to remember back 3 years, and he said "Yes, I seem to recall that one computer kept rebooting for some reason - so they quickly sent me another one." I recalled later that we had been on the phone 3 years before and I had advised him to go to the bios and reduce his core count from 6 down to 4, and also to turn off hyper threading (dropping from the reduced 8 threads down to only 4 threads) to try to get the heat under control. That actually did help quite a bit - he was able to get about 30 minutes of work done each time before the next auto-reboot while he waited for the replacement computer to be flown in. So, now, sitting there with that hot-running cpu, I figured that there must not be any water circulation carrying the heat away to the large radiator. So I pulled off the Intel waterblock, laid it to the side with the water lines still connected, and fetched my massive Silver Arrow air cooler, still brand new in the box from my purchase 6 years before when I had considered upgrading at that time, installed it with paste, and then I watched the cpu maintain a nice cool 27 degrees for about 5 minutes. “Awesome!” I took a picture of the system and the monitor display – you can see the “CPU Temperature 27” - thinking “this will be a beautiful upgrade for me” and I sent a nice text to my relative. But the next morning – the system started acting up – the memory was rejected – hitting the “Mem ok” light caused rebooting as the motherboard tested different memory settings, then finally that wouldn’t happen anymore, and the cpu LED light went on and stayed on, and the motherboard fans stopped spinning – one of them jerked in time to a clicking noise from the power supply. The PSU, a Corsair AV 850, after sitting in the garage for 3 years, and after working for 5 perfect minutes the night before, decided to fail - taking out the motherboard, and as I discovered later, taking out the graphics card also. You heard me, motherboard and graphics card. (I liked the motherboard so much I ended up buying a used sabertooth x79 motherboard/cpu combo for about $300 which works fine, but initially I had no monitor display, which was deeply puzzling to me, not knowing that the GTX 680 graphics card had just been ruined.) My much newer $500 GTX 1070 had been put safely away. When I finally decided that, as much as I could not believe it, my video card must be bad, I was reluctant to pull my recently purchased 1070 out of safe-keeping, but I remembered that I had an old known-good GTX 8800 ultra on hand to test the new motherboard/cpu combo. I also had a couple of known-good AMD 7950s that I could have used, which had been in crossfire service until I picked up the GTX 1070 since “sli is dead” as they say – many newer games are not coding for sli or crossfire, because one really good card modern card can usually handle the load. (I now clearly see the value of keeping old known-good components around, rather than selling everything off - in failure situations substituting a known-good component is absolutely the surest way to isolate and diagnose the failure and end all the guess-work.) Of the 7 power supplies that I ended up testing, 3 of them were bad as I mentioned. In regard to using the testers on the bad power supplies, there was one PSU that appeared to be good as far as the cheap tester was concerned. All the voltages were there. But the Dr. Power II clearly indicated that the PSU was bad because its display flickered. That power supply also would not pass the load test - the green ready light on the sabertooth motherboard flickered instead of maintaining a nice steady green. Another bad power supply displayed nothing at all on the cheap $10 tester, but the Dr. Power tester turned red - and indicated "F" for failure on multiple voltages. I think a red display with a bunch of “Fs” is a much more definitive “Failure” than nothing at all. I could toss out my cheap tester, and it wouldn’t bother me. Why even keep it around? I probably WILL toss it out. The Dr. Power II is far superior. By the way, the Dr. Power tester says in the manual something to the effect that "the tester may turn warm, as it tries to put a small load on the power supply as it tests the voltages." Yes I have felt it warm up a bit - that is truly great – when I read that it gave me even more respect for the Dr. Power tester. But as I have said, nothing beats having more of a load, and my old bad sabertooth motherboard, which has two built-in tiny 35mm fans cooling two sections on the motherboard, is just a great load test. Why is it a great load test? PART A: First of all, if the re
