EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: FAQBytes on August 06, 2016, 09:49:16 am
-
Hello all,
Been a while since I've been on the forum, so I hope all is well.
I've been running into a few problems with my dual socket LGA2011 motherboard. After buying a new PSU (Old one only supported 1 CPU) I've contued to run into a power issue. It seems that the side that is working now previously ran into the same problem, as the photo I will post will suggest, so I am now assuming that the side that is not working had the same problem.
If you guys can possibly break down what they say to tell me the voltages, that would be greatly appreciated, as the voltages can't be measured as the motherboard goes into self-preservation mode (at least that saves my 8-core Xeon processors).
With that said, should they not be figured out, I have a few questions. As the input voltage is DC12v from the rail, and the CPU uses .85-1.15v, I think it would be relatively safe to assume the voltage across the capacitor would be somewhere between there (If you want to have some fun I can post a picture of the entire circuit between the ATX EPS connector (8 pin input, 4 ground and 4 12v pins) and the CPU itself if you wish to figure it out and fact check me. (I'm in school for Mechanical and Aerospace engineering, not EE).
Should it be closer to the 1-2v mark, would it still be safe to put a 25v rated 470uF? My grandfather, an electrical engineer, suggests 3-4 times working load, (general discussion, not this in particular) but 25v would be far outside that.
Thank you for your time and reading this far too wordy post.
Tl;Dr- Would a 25v rated Tantalum Capacitor cause problems on a 1-2v circuit?
-
My understanding is that capacitors with a higher voltage rating tend to, all else being equal, have a higher ESR than those with a lower voltage rating, so it's at least possible to derate too much - which is not to say that using a 25V cap would necessarily cause any problems in your particular situation.
-
Well, for one, those aren't tantalums...
You need a low-ESR capacitor since it's a CPU rail. Go for a 4V or 6.3V cap; anything more just gives you worse performance.
-
My understanding is that capacitors with a higher voltage rating tend to, all else being equal, have a higher ESR than those with a lower voltage rating
Exactly opposite if we are talking about same capacitance. 25V, 16V, 10V, probably even 6.3V capacitors just won't fit (if you can find 470 uf at such voltage rating at all), and those caps on the photo likely are 2.5V rated.
-
Whelp, I guess I'm wrong about them being tantalum. Would anyone know where I could source them without spending an arm and half my soul?
Preferably shipping from the USA as I'd like to get them this century.
-
They are either LOW ESR polymer tantalum (most likely) or tantalum.
-
I'm most likely going to be buying from mouser unless anyone else knows a better place.
-
Eh, here you go (http://www.digikey.com/product-detail/en/panasonic-electronic-components/EEF-SX0E471E4/PCE5042CT-ND/3511317). That should be a replacement that's as good as, or better, than the existing components.
There are slightly cheaper models (http://www.digikey.com/product-search/en/capacitors/aluminum-polymer-capacitors/132402?k=&pkeyword=&pv1500=346&pv1500=394&FV=1c0002%2C1c0003%2C3400b7%2Cb836d9%2C1140003%2Cfff40002%2Cfff80532%2Cfc0009&mnonly=0&newproducts=0&ColumnSort=724&page=1&stock=1&quantity=0&ptm=0&fid=0&pageSize=25), but I'd rather spend a bit more cash to get something which is not going to be insufficient.
-
Exactly opposite if we are talking about same capacitance.
Indeed. Just checked some datasheets and yes, dissipation factor gets lower as voltage rating increases - which translates into lower ESR for the same capacitance. I stand corrected.
Cheers,
Karl.
-
Exactly opposite if we are talking about same capacitance.
Indeed. Just checked some datasheets and yes, dissipation factor gets lower as voltage rating increases - which translates into lower ESR for the same capacitance. I stand corrected.
Aluminum electrolytic capacitors have a step increase in dissipation factor at about 160 volts which I suspect is caused by a change in electrolyte composition.
Another reason to use a significantly greater voltage derating is physically larger capacitors have larger ripple current ratings.
-
Eh, here you go (http://www.digikey.com/product-detail/en/panasonic-electronic-components/EEF-SX0E471E4/PCE5042CT-ND/3511317). That should be a replacement that's as good as, or better, than the existing components.
Recommend to go for that one as well.
-
Alright, I'd like to thank all of you for your help, especially those who provided links to replacements. I dug a little deeper and found some things out. Those capacitors are connected directly to the vcore of the CPU, and as I mentioned previously, the absolute Max is 1.35v vcore, as it's a non-overclockable chip and motherboard. With that said, the 2.5v capacitors that were linked will have over a volt headroom, and if that isn't enough, I have a bigger problem than the capacitors.
I'll most likely be ordering them and getting new flux after the school year starts. I'll update you all on the progress.
Thanks again.
-
The only time I've had too much derating causing an issue was when a batch was built with tant caps the wrong way round, but so overated that they didn't blow on production test, or during burn in with the customer, so ended up with hundreds of units spread all round Germany that died after a couple of weeks.
I've had happier customers.
-
In the past when I have wanted to replace capacitors in an application like this where the highest possible performance was desired at any cost, I found the appropriate type but increased the value of capacitance or voltage or both to the largest which would fit in the package size.