Announcement

**Th3_uN1Qu3** · 11-18-2010, 10:36 PM

Re: DIY PSU load tester

Where did you get your resistors? Power resistors are darn expensive around here.

I went and bought some more miscellaneous stuff yesterday such as connectors and a couple extra opamps. Unfortunately the analog meters i've ordered still haven't arrived, so until they do the project is on hold. Reason being that i have to lay out the meters and knobs on the front panel before i do anything else.

**c_hegge** · 11-18-2010, 10:50 PM

Re: DIY PSU load tester

I got them from RS COmponents. You're right, they are expensive (about $5 each) so all up they cost me about $130AU or so, but that's OK as most of the other bits were just scavenged.

**Th3_uN1Qu3** · 11-19-2010, 10:23 AM

Re: DIY PSU load tester

With resistors you have an additional issue - their resistance increases as they warm up which is something you need to consider, if they get hot quickly chances are you're drawing a fair bit less power than you think you are. Hence why i decided to build a system with feedback in the first place, regardless of operating temperature the current draw will stay constant.

...Or just submerge the damn things.

**shovenose** · 11-19-2010, 10:26 AM

Re: DIY PSU load tester

If you cool them well, the resistors will be fine. Like use on of those 120v mains fans

**Th3_uN1Qu3** · 11-19-2010, 11:09 AM

Re: DIY PSU load tester

I'm not talking about whether they will be fine or not as resistors take more abuse than semiconductors anyway. I'm talking about whether they will keep their exact resistance, which most likely they will not.

**Th3_uN1Qu3** · 11-19-2010, 11:14 AM

Re: DIY PSU load tester

Originally posted by c_hegge View Post

I got them from RS COmponents. You're right, they are expensive (about $5 each)

More like $10 each here. So cost wise it was cheaper to do it all with semiconductors, even if it involved a significantly higher effort (designing the circuit to drive and meter them).

However in the long run this will be offset by the value, the circuit is very versatile, and will be able to test any type of power supply not just computer PSUs. If i need to test PSUs with more than 100v output i can simply buy higher voltage transistors on demand, since each of the 4 rails will have an independent monitoring circuit i can just upgrade them in groups of 4, 6 or 10, not all 30 at once.

**shovenose** · 11-19-2010, 12:44 PM

Re: DIY PSU load tester

um, ok...but i have given up trying to figure out all your complicated mosfet stuff and i may (if i do end up making any kind of psu tester), go the resistor route

**c_hegge** · 11-19-2010, 03:46 PM

Re: DIY PSU load tester

Originally posted by Th3_uN1Qu3 View Post

With resistors you have an additional issue - their resistance increases as they warm up which is something you need to consider, if they get hot quickly chances are you're drawing a fair bit less power than you think you are. Hence why i decided to build a system with feedback in the first place, regardless of operating temperature the current draw will stay constant.

...Or just submerge the damn things.

They only get too hot to touch with no fan. With 4 fans they get warm, but you can still touch them. And if I can't get them cool enough, there's nothing stopping me (other than the cost) than using these fans. 252CFM each!

**Agent24** · 11-19-2010, 05:45 PM

Re: DIY PSU load tester

I was going to go with resistors too when I first had the idea of making a load tester (mainly because I have no idea how I would design a MOSFET one)

Would love to build one of these MOSFET ones myself, but I have no idea where I would calibrate it..

I can keep dreaming.. haha

**Th3_uN1Qu3** · 11-20-2010, 06:29 AM

Re: DIY PSU load tester

Originally posted by c_hegge View Post

They only get too hot to touch with no fan. With 4 fans they get warm, but you can still touch them. And if I can't get them cool enough, there's nothing stopping me (other than the cost) than using these fans. 252CFM each!

That's good. But 40 bucks for a fan? Seriously.

Originally posted by Agent24 View Post

Would love to build one of these MOSFET ones myself, but I have no idea where I would calibrate it..

I can keep dreaming.. haha

Well my good meter only goes to 10A as well (the 20A one isn't that accurate), but since all PSU testing is done in 1A increments that wouldn't be a problem. At most, it'll be 1A off which on 12v that means 12W, on the other rails even less. Not a big deal in my book. So the calibration of the load tester itself won't be a problem. The ripple meters will be a bit more difficult to set up though, but you can simply use one of those USB oscilloscopes instead. Disadvantage - you would have to move the probes for each rail.

Originally posted by shovenose View Post

um, ok...but i have given up trying to figure out all your complicated mosfet stuff and i may (if i do end up making any kind of psu tester), go the resistor route

Well y'know, if ya don't try ya have no chance to succeed. When the PCB layouts will be posted i'm quite sure they'll shed more light on everything. In the meantime here's the schematic of the 12v load tester. The others are the same just with less MOSFETs and different R2 values. Btw, R2 is a resistor in series with a trimpot hence the funky value. R3/R4 simulate the 1k potentiometer.

Attached Files

**c_hegge** · 11-20-2010, 05:33 PM

Re: DIY PSU load tester

Originally posted by Th3_uN1Qu3 View Post

That's good. But 40 bucks for a fan? Seriously.

As I said, the cost would be the only problem. The freight to Australia would be about 30AUD, so they would really be about 70 each! so I won't use them unless I have no other option, but I think they might be OK with the fans I've got now.

**370forlife** · 11-20-2010, 05:55 PM

Re: DIY PSU load tester

Try some nidec ta600dc's. I have a few from a old Sun Sunfire 880 server. I think they are 300CFM each, and are really not that loud. They don't make that high pitch whine either. Just a bit big though, 172mm x 51mm.

**Th3_uN1Qu3** · 11-24-2010, 01:48 PM

Re: DIY PSU load tester

Mmm, maybe this thread should finally be moved to the "test equipment" forum? I'm still waiting on those meters, i'll go check the store tomorrow to see if they arrived.

I also found my old 20 to 24-pin adapter, came with a Thermaltake PSU AFAIK. So i now have a fully wired ATX connector to mount on the front panel, lazy me.

The 4-pin extension i can gut off some mobo so it can be full 24 pins. I have dozens of male molexes from busted HDDs and optical drives, so that won't be a problem either. I have an ATX 12v connector too... now i just have to find some PCI-E ones.

I also have enough thermistors to monitor each of the 4 rails separately. When i'll have the thing built i'll check for the hottest heatsinks in the bunch and mount the thermistors there. I don't want the tester to sound like taking off all the time y'know, so thermal control is welcome.

Same circuit will also implement overtemperature protection.

**shovenose** · 11-24-2010, 02:10 PM

Re: DIY PSU load tester

i thought you already were having marginal temperature control already. wouldnt fan speed control make it run even hotter?

**Th3_uN1Qu3** · 11-24-2010, 02:42 PM

Re: DIY PSU load tester

Well, with 2/3 more heatsink and roughly triple the CFM (gained from overvolting the fans), i doubt temperature will be that big of a problem anymore. These MOSFETs can take temperatures of 175C you know. The point of fan speed control is using no more fannage than you need at that point in time - running the fans at full blast when testing a 400W supply would be quite pointless.

I've tested the MOSFETs a while back (i think it's been posted in this thread) and they have no problem clearing the design specs. This thing will be capable of huge peak power - the design spec is 3A/device (power dissipation and temperature limited, i also applied temperature derating and with that taken into account i still have plenty of headroom) while the 'fets themselves are rated for a whopping 14 amps each. Just four of them had no trouble handling 20A for a couple minutes, and that's with just 3 of 8 fans running.

**c_hegge** · 11-29-2010, 04:05 AM

Re: DIY PSU load tester

Nearly finished building mine. I ran out of heatshrink tubing, though. I should be getting some more tomorrow. What do you think would be a good way of testing it to be sure it's pulling as much as I think it is? I'm thinking just load a gutless wonder to 350 watts. If it's still working 10 minutes later then there's a problem, but if it explodes, then it's a pass.

**c_hegge** · 11-30-2010, 11:05 PM

Re: DIY PSU load tester

FINISHED!

Got it all going and it seems to be working great. I loaded this unit to 277 watts. After a few minutes I got exactly what I expected, the PSU exploded, so it looks like it's doing its job.

**Th3_uN1Qu3** · 12-01-2010, 02:07 AM

Re: DIY PSU load tester

That's nice. I'm still waiting on those darn meters to arrive...

**goodpsusearch** · 12-01-2010, 05:45 AM

Re: DIY PSU load tester

Well done! We need some photos now

**c_hegge** · 12-03-2010, 07:46 PM

Re: DIY PSU load tester

Here. It looks horrible, I know, but it works.

Attached Files

Announcement

DIY PSU load tester

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment