Is it a good idea to connect a 9-10 yeard old PSU in a new build?

[Ankerson]

4 minutes ago, akio123008 said:

Everything degrades. That's not the point. The point is, is it likely to have degraded so much that there is significant risk to using it short term; I don't think so. 

 

That is the whole point, is the point and the MOST IMPORTANT POINT.

[akio123008]

Just now, Ankerson said:

 

That is the whole point, is the point and the MOST IMPORTANT POINT.

The point is HOW MUCH something degrades, not IF it does so. And in general, power supplies don't degrade enough over 9 years to form significant risk for short term use in a new system, especially if they were still in use all this time. I wouldn't trust it if it would have been on a shelf for years.

[Ankerson]

3 minutes ago, akio123008 said:

The point is HOW MUCH something degrades, not IF it does so. And in general, power supplies don't degrade enough over 9 years to form significant risk for short term use in a new system, especially if they were still in use all this time. I wouldn't trust it if it would have been on a shelf for years.

 

I will say again...

 

Until they are hooked up to PSU testing equipment NOBODY KNOWS.

 

I think people would be surprised how much they do degrade over time from what I have been told by people who actually tested them.

 

I think there needs to be things actually published on this so we can put this freaking BS to rest that PSUs last forever as long as they still turn on.

[akio123008]

2 minutes ago, Ankerson said:

Until they are hooked up to PSU testing equipment NOBODY KNOWS.

Nobody knows 100%, but usually they're fine.

 

We can go back and forth for a couple more hours but what I think it boils down to is this:

 

 - In your opinion you can't be sure if it's ok without testing, which is technically true, so why take the risk to run it short term, if you're going to buy a new one long term anyway?

 - In my opinion the risk is small, there's no direct reason to assume it's harmful, so go ahead and run it, don't worry too much about it (for short term use!) 

 

Which option one would agree with I guess depends on how much risk one considers acceptable, which is personal preference.

 

 

[Ankerson]

11 minutes ago, akio123008 said:

Nobody knows 100%, but usually they're fine.

 

We can go back and forth for a couple more hours but what I think it boils down to is this:

 

 - In your opinion you can't be sure if it's ok without testing, which is technically true, so why take the risk to run it short term, if you're going to buy a new one long term anyway?

 - In my opinion the risk is small, there's no direct reason to assume it's harmful, so go ahead and run it, don't worry too much about it (for short term use!) 

 

Which option one would agree with I guess depends on how much risk one considers acceptable, which is personal preference.

 

 

 

 

Would you hook up an old PSU to a $3000 or more machine that you paid for?

 

Yes I am talking about real money here.

 

Let me know because I am sure I can find one that old for you. 

 

And those high end 3080's and 3090's series GPUS that are hard to find, and expensive if you can even find one...

 

 

[akio123008]

12 minutes ago, Ankerson said:

Would you hook up an old PSU to a $3000 or more machine that you paid for?

If I bought a PC like that, a new PSU would be on the order list, so it wouldn't be an issue.

 

But let's assume I had new hardware right now, but no new PSU, and for some reason the new PSU was out of stock of whatever, and I'd have to wait for 2 weeks before it arrives.

 

In that case, I'd be fine taking a 9 year old power supply from another rig that's running fine to do a test run of my new hardware. I wouldn't put it in for good, but just to test stuff out, or while I'm waiting for the new one to arrive, I wouldn't mind that really.

 

I'm not even entirely sure I'd do it, but that's not because of the old power supply, but because I'd have to take stuff apart again and reassemble when the new one comes in, so it'd be sort of a pain. But if that wasn't a thing and I had a good reason to get the new rig up and running, I'd go for it.

 

What it comes down to is, I can hardly imagine a situation where I'd do that for practical reasons, but from a "fear of destroying the hardware" perspective, I'd have no problem doing it.

 

12 minutes ago, Ankerson said:

Let me know because I am sure I can find one that old for you. 

Problem is I don't have a 3000USD machine. I do have plenty of old power supplies though, the one in my main rig is about 7 or 8 years old at this point. (and that's a very crappy unit btw, without 80PULS certification even, but I keep putting off replacing it)

[Ankerson]

10 minutes ago, akio123008 said:

If I bought a PC like that, a new PSU would be on the order list, so it wouldn't be an issue.

 

But let's assume I had new hardware right now, but no new PSU, and for some reason the new PSU was out of stock of whatever, and I'd have to wait for 2 weeks before it arrives.

 

In that case, I'd be fine taking a 9 year old power supply from another rig that's running fine to do a test run of my new hardware. I wouldn't put it in for good, but just to test stuff out, or while I'm waiting for the new one to arrive, I wouldn't mind that really.

 

I'm not even entirely sure I'd do it, but that's not because of the old power supply, but because I'd have to take stuff apart again and reassemble when the new one comes in, so it'd be sort of a pain. But if that wasn't a thing and I had a good reason to get the new rig up and running, I'd go for it.

 

What it comes down to is, I can hardly imagine a situation where I'd do that for practical reasons, but from a "fear of destroying the hardware" perspective, I'd have no problem doing it.

 

Problem is I don't have a 3000USD machine. I do have plenty of old power supplies though, the one in my main rig is about 7 or 8 years old at this point. (and that's a very crappy unit btw, without 80PULS certification even, but I keep putting off replacing it)

 

And you hook it up and boom it fried the $2000 GPU and motherboard.

 

So you are out $2000 plus the motherboard just like that.

 

I would wait until the new PSU came in while using the old machine.

 

Look at the machines I have in my sig, do you really think I would even think about running an old PSU in any of them?

[akio123008]

9 minutes ago, Ankerson said:

And you hook it up and boom it fried the $2000 GPU and motherboard.

Well you know that's where the risk factor comes in. In my opinion the likelyhood of that happening is pretty darn small. This isn't some old thing off a shelf, this thing's been in a running rig, and still is, with very similar hardware in it.

 

Once again, I probably wouldn't bother even touching the new hardware until the actual PSU arrives, because for me it would be a waste of time anyway, but given there was some good (unknown) reason to get the new thing running ASAP, I would not be worried about that worst case scenario too much. 

 

You also have to keep in mind the spectacular failure you mentioned could only take place if the PSU was either operating very out of spec, but that wouldn't be the case since the other PC was fine, or fried the components slowly over time, but that also wouldn't matter, as it'd be very short term use anyway.

[Beerzerker]

On 3/23/2021 at 4:16 PM, Rabolisk said:

I built a PC in late 2011/ early 2012 at the time with a Sandy Bridge Intel (Still running great!). I still have a PSU, a cooler master 650W ATX Bronze, and I believe it has a 2 x 6 pin connector for my very old GPU. Would it be a bad idea to connect it to a modern motherboard? If I connect it to a GPU would have to use an adapter.

The above is the original question as asked.
The OP is the only one that can answer their intended useage, that is IF they intend to run it period with the board or just to test it and nothing else.

I've said if simply wanting to test I don't see any real harm in doing it. I never said it's a good idea to run it in the sense of it being a "Permenant" source of power, that's why I said earlier:

On 3/24/2021 at 7:35 AM, Beerzerker said:

I will also say if doing a new build, you'd best get a new PSU because that just makes sense.
Cheaping out on the one thing that's at the heart of the system is just asking for it.

What I referred to earlier is that to state ALL older PSU's are bad period is a mis-statement, context of what is said is very important here.
Just because it's old does NOT automatically mean it's bad period.

The insistence of hooking it up to test equipment, in itself isn't bad advice but in the real world, I'll say it again - Not everyone has access to that kind of equipment or the knowledge to use it. 
If you have it that's great - Seriously, use it!
However that does not apply to many if not most here.

So, if just wanting to do a simple test of the board AND the PSU has shown no signs of ill-behaviour, chances are it's OK as long as it's done as a test-only hookup.
I've done that literally thousands of times along the way and never had a problem once. Any failures I've ever had were caused my me directly in doing something wrong on my part, never from a PSU that was just "Old".

I'm also not arguing the point of things failing - They all do at some point but the risk of possible failure has to be recognized and stated because that too is real.
No guarantees of anything, even one brandnew right out of the box can pop and that's it.

To me, the risk of failure because everything IS subject to failure is more or less the same as far as the end result goes. 
The thing about age does apply because that is true, nearly all PSU's are used from being new until they reach a certain age but at the same time, the thing about one automatically being "Bad" just because it's of a certain age or older isn't right either.

Testing is a smart thing to do IF you can but if not and the unit hasn't shown any signs of trouble, it's fairly (Not completely mind you) to assume it's fine as long as you're not wanting to run it full time with it.
Agree or disagree, feel free to and do as you will - I do.

That's my take on it and that's it.

[Ankerson]

Just now, akio123008 said:

Well you know that's where the risk factor comes in. In my opinion the likelyhood of that happening is pretty darn small. This isn't some thing off a shelf, this thing's been in a running rig, and still is, with very similar hardware in it.

 

Once again, I probably wouldn't bother even touching the new hardware until the actual PSU arrives, because for me it would be a waste of time anyway, but given there was some good (unknown) reason to get the new thing running ASAP, I would not be worried about that worst case scenario too much. 

 

You also have to keep in mind the spectacular failure you mentioned could only take place if the PSU was either operating very out of spec, but that wouldn't be the case as the other PC was fine, or fried the components slowly over time, but that also wouldn't matter, as it'd be very short term use anyway.

 

You think?

 

It has happened before.

 

Taking on old PSU out of a machine might pull what 300W or maybe 400W and put it into one that pulls 500W to 800W and you really think the risk is low that something will happen?

[akio123008]

3 minutes ago, Ankerson said:

Taking on old PSU out of a machine might pull what 300W or maybe 400W and put it into one that pulls 500W to 800W and you really think the risk is low that something will happen?

I did mention the old PC had the same power consumption didn't I?

 

18 minutes ago, akio123008 said:

In that case, I'd be fine taking a 9 year old power supply from another rig that's running fine to do a test run of my new hardware.

Oh, right it was supposed to be in this sentence, but I rephrased that and forgot. But yeah, I'd be fine with it if the power output was roughly correct, I wouldn't put a severely underpowered unit in there, but neither would I do that if it were new.

 

4 minutes ago, Ankerson said:

It has happened before.

I know, it's just that to me personally it's not common enough to be worried about. But that may very well be different for you or other people and I can totally understand that. You asked me for what I'd do though, so there it is I guess.

[Beerzerker]

4 minutes ago, Ankerson said:

 

You think?

 

It has happened before.

 

Taking on old PSU out of a machine might pull what 300W or maybe 400W and put it into one that pulls 500W to 800W and you really think the risk is low that something will happen?

If someone can't read a label and figure that out BEFORE hooking it up, they asked for it and got it in that case.
BTW if they are that clueless, the age of the PSU is a non-issue, if it's too much, it's too much - End of story. 

In short, they would be the problem, not the PSU itself regardless of age or wattage capability.
 

[Ankerson]

2 minutes ago, Beerzerker said:

If someone can't read a label and figure that out BEFORE hooking it up, they asked for it and got it in that case.
BTW if they are that clueless, the age of the PSU is a non-issue, if it's too much, it's too much - End of story. 
 

 

 

If it's the correct wattage for the machine?

 

Taking a 9 or 10 YO 1000W PSU as an example.

 

Who knows how much it's degraded over that time period.

[akio123008]

5 minutes ago, akio123008 said:

But yeah, I'd be fine with it if the power output was roughly correct,

By this I mean that I know the power supply has the required output, by being from my old rig, which I assumed in my example, has roughly the same power consumption.

[Beerzerker]

You do bring up a good point but the crux of it is:

1 minute ago, Ankerson said:

If it's the correct wattage for the machine?

That is the responsibility of the one doing the testing to figure this out and know beforehand.
I'll say again, if a person can't figure all that out that's all on them and so are the consequenses of it.
You screw up and (Hopefully) Learn from it in that case.
 

[Ankerson]

Just now, Beerzerker said:

You do bring up a good point but the crux of it is:

That is the responsibility of the one doing the testing to figure this out and know beforehand.
I'll say again, if a person can't figure all that out that's all on them and so are the consequenses of it.
You screw up and (Hopefully) Learn from it in that case.
 

 

 

Yeah, exactly. 

 

Hopefully not the hard and expensive way.

[Beerzerker]

I certainly don't wish it upon anyone myself.

[Ankerson]

Just now, Beerzerker said:

I certainly don't wish it upon anyone myself.

 

Me either, which is why I normally recommend replacing them, just incase.