Thank you very much!

The polymer caps ESR is normally given at 100 kHz right? At least I have not seen one given at 1 kHz. Would it trouble you too much to measure those 2 at 100 kHz? I guess it won't be drastically different?

Anyway, I've put together a short-term shopping list for the VRM caps as follows:

• VRM High - Big 8x20 - 5 - United Chemi-Con KZG - 1000uF 16V - 0.021 mOhm - 1870 mA
• VRM High - Big 8x20 - 1 - Ltec LMB - 1500uF 6.3V - 0.024 - 1640
  Replacement for both of VRM High sets of caps:
  8x20 - Chemi-Con - 1200uF 16V - 0.008 - 7500 - Polymer
  https://uk.rs-online.com/web/p/polymer-capaci … rs/2514823?gb=s
• VRM Low - Small 8x8 - 10 - Panasonic FL - 680uF 4V - ??
  Replacement:
  6.3x9 - Nichicon - 560uF 6.3V - 0.009 - 4300 - Polymer - https://uk.rs-online.com/web/p/polymer-capaci … rs/7149058?gb=s

I will retest the board after replacing the VRM Highs. If I succeed without too much struggle, I will continue.

My first and also last time doing this was on a socket 370 board, of course without much of the knowledge learned from this thread. Thankfully it was a success, but I struggled way too much with a cheap soldering station and a pump tool, took so many hours to clear the holes. This time I might try getting a desoldering gun.

randomdavid wrote on 2025-05-06, 20:53:

The polymer caps ESR is normally given at 100 kHz right? At least I have not seen one given at 1 kHz. Would it trouble you too much to measure those 2 at 100 kHz? I guess it won't be drastically different?

You have a keen sense of observation! Yes, correct, I did notice that too, but I can't test at 100Khz because my meter lacks that function. Yes, it will be drastically different if their 5mOhm claim on their initial datasheet was accurate. For example, tantalum-polymer manufacturers also specify the ESR at 100Khz and I have tested them at 1Khz at more than double the ESR!

But since this capacitor is not used for the CPU VRM, it is not filtering a high frequency circuit, so 13mOhm @ 1Khz is plenty low enough, if not complete overkill.

Anyway, I've put together a short-term shopping list for the VRM caps as follows: <snip> 8x20 - Chemi-Con - 1200uF 16V - 0.008 - […]
Show full quote

Anyway, I've put together a short-term shopping list for the VRM caps as follows:
<snip>
8x20 - Chemi-Con - 1200uF 16V - 0.008 - 7500 - Polymer
https://uk.rs-online.com/web/p/polymer-capaci … rs/2514823?gb=s

Looks good.

[*] VRM Low - Small 8x8 - 10 - Panasonic FL - 680uF 4V - ?? Replacement: 6.3x9 - Nichicon - 560uF 6.3V - 0.009 - 4300 - Polym […]
Show full quote

[*] VRM Low - Small 8x8 - 10 - Panasonic FL - 680uF 4V - ??
Replacement:
6.3x9 - Nichicon - 560uF 6.3V - 0.009 - 4300 - Polymer - https://uk.rs-online.com/web/p/polymer-capaci … rs/7149058?gb=s
[/list]

Note good. You need to match the original diameter of 8mm for the lead spacing. I think the lead spacing on the 6.3mm wide polymers will be too narrow.

but I struggled way too much with a cheap soldering station and a pump tool, took so many hours to clear the holes. This time I might try getting a desoldering gun.

I can tell you right off, it's not going to be easy. The VRM area has a large copper plane that's going to suck away any heat you put into it. Hot air used to heat up the area beforehand combined with your desoldering tool works wonders, just make sure you get the polarity correct. Asus reverses capacitor polarity opposite to the board marking, keep that in mind.

Try to order your caps from one place, I recommend Farnell if you're in the UK.

I see, thank you for pointing out! Yes I didn't realise that lead spacing for the VRM low caps would be a problem.

I used rs components UK website and couldn't find any suitable caps in this range. I then used farnell UK website, and found some suitable ones (all polymer with 5000 hours rating):
8x9 - Panasonic - 560uF 6.3V - 0.007 - 6100
https://uk.farnell.com/panasonic/6sepc560mx/c … v-20/dp/2354643
8x8 - Nichicon - 820uF 2.5V - 0.006 - 6100
https://uk.farnell.com/nichicon/rl80e821mdnas … 2-5v/dp/2841880
8x8 - WURTH ELEKTRONIK - 560uF 6.3V - 0.007 - 6100
https://uk.farnell.com/wurth-elektronik/87023 … v-20/dp/2466583
Edit: Weird that product direct links of Farnell website doesn't work (it gives 403 error), you have to go to the product page from their search page by manually searching...

The Panasonic seems to be the cheapest, it is 1mm taller than the original but I think it should be ok.
I have never heard of Wurth Elektronik, but then I also have never researched this much into capacitors. Maybe they are a good brand?

There's also a brand called KEMET, which is significantly cheaper than other brands. Seems to be an American company? Due to the much cheaper prices, my gut feeling is to stay away, but please tell me if I'm wrong!

randomdavid wrote on 2025-05-09, 19:39:

8x8 - Nichicon - 820uF 2.5V - 0.006 - 6100
https://uk.farnell.com/nichicon/rl80e821mdnas … 2-5v/dp/2841880

Go with this one. Higher than 8mm and you might have an issue with heatsink clearance on the VRM low.

I have never heard of Wurth Elektronik, but then I also have never researched this much into capacitors. Maybe they are a good brand?

Würth is a European re-brander of electronics components. We don't know where their caps come from -- but they don't make them themselves. Probably Taiwanese or Chinese caps.

There's also a brand called KEMET, which is significantly cheaper than other brands. Seems to be an American company? Due to the much cheaper prices, my gut feeling is to stay away, but please tell me if I'm wrong!

It's impossible to know for certain what Kemet capacitors actually are -- they are very tight-lipped and their capacitors should be considered unproven.

I've got some complications...there are no polymer 16V capacitors on Farnell UK for the VRM High caps. Unlike RS Components, where I originally found the replacements for those

8x20 - Chemi-Con - 1200uF 16V - 0.008 - 7500 - Polymer
https://uk.rs-online.com/web/p/polymer-capaci … rs/2514823?gb=s

On RS Components there are these:
8x9 - Nichicon - 820uF 2.5V - 0.007 - 5200
and nothing similar to these

8x8 - Nichicon - 820uF 2.5V - 0.006 - 6100

There are some others with a more suitable size, but much higher ESR and less Ripple rating
8x7 - Nichicon - 680uF 2.5V - 0.015 - 3900
https://uk.rs-online.com/web/p/polymer-capacitors/7149004
8x7 - Nichicon - 560uF 4V - 0.015 - 3900
https://uk.rs-online.com/web/p/polymer-capacitors/7149039
15 mOhms sounds a bit too close for comfort.

At least, with both the stock Intel cooler and the Gelid cooler that I have, the 1mm extra height of those caps should still be fine, but I'm not sure about all the other coolers out there.

Looks like I'll have to calculate more before pulling the trigger. Perhaps look for full replacements for all caps on the board on both websites for the cost to add up to free shipping. Or add a Hakko FR301 desoldering gun to the cart xD

randomdavid wrote on 2025-05-10, 12:50:

I've got some complications...there are no polymer 16V capacitors on Farnell UK for the VRM High caps. Unlike RS Components, where I originally found the replacements for those

No, you're mistaken.

Something like this fits the bill:
https://uk.farnell.com/chemi-con/apsg160etd10 … dial/dp/4630196

You also don't need that much capacitance on the VRM high... You can go down quite a bit here if you're so inclined: example.

Also, don't waste money on expensive desoldering guns before you have the technique down to a pat.

mockingbird wrote on 2025-04-29, 13:27:

What electrolytics lack for in ripple suppression capability, they make up for in endurance rating and price. Why use a 2000-hour polymer cap which runs for $1-$2/piece to filter a USB port when you can get a 5000-10,000 hour electrolytic cap for 30 cents.

Chemi-Con PSG series would like to have a word with you. 😁
Other than that, I agree with everything you said.

mockingbird wrote on 2025-05-11, 02:47:

Something like this fits the bill:
https://uk.farnell.com/chemi-con/apsg160etd10 … dial/dp/4630196

You also don't need that much capacitance on the VRM high... You can go down quite a bit here if you're so inclined: example.

THANK YOU!!!
How come when I looked for 16V polymer caps on Farnell (with no restriction on capacitance), I only found 2, 150uF and 100uF. Also my product links would always give a 403 error but yours don't. Very strange indeed. Anyway, the search is over thanks to you!

mockingbird wrote on 2025-05-11, 02:47:

Also, don't waste money on expensive desoldering guns before you have the technique down to a pat.

Yes, the Hakko comment was a joke 😁, I intend to join a Hackspace UK since there's one in my city, and hopefully I can use the equipment there for this job.

Just to be 1000% sure I won't buy the wrong thing and break the board along with the CPU, I've made an annotated shot of the caps. I don't trust my ability to pin down which caps are VRM ones and which aren't.

Here is what I believe:

• Red - VRM Lows
  Original:

  8x8 Panasonic FL 680uF 4V. No datasheet.

  Replacement:

  8x8 Nichicon 820uF 2.5V 0.006 6100 Polymer
  https://uk.farnell.com/nichicon/rl80e821mdnas … 2-5v/dp/2841880

• Green - VRM Highs
  Original:

  8x20 United Chemi-Con KZG 1000uF 16V 0.021 1870
  https://www.digikey.co.uk/en/htmldatasheets/p … /0/1/kzg-series

  Replacement:

  8x20 Chemi-Con 1000uF 16V 0.008 7500 Polymer
  https://uk.farnell.com/chemi-con/apsg160etd10 … dial/dp/4630196

• Yellow - VRM High
  Original:

  8x20 Ltec LMB 1500uF 6.3V 0.024 1640 https://paullinebarger.net/DS/Ltec/Ltec%20%5B … MB%20series.pdf

  Replacement:

  Ether the same as above (the 1000uF 16V), or:
  8x8 Nichicon 1000uF 6.3V 0.009 5700
  https://uk.farnell.com/nichicon/rl80j102mdn1k … 6-3v/dp/2841881

  I could not find a 8mm 6.3V Polymer cap with more than 1000uF

Since we've got this far, can you pleaseeee help me one more time to confirm if this is OK?

Forgot to mention that I'm fairly confident the red highlighted ones are VRM lows and green ones are VRM highs, but the yellow one I'm not so sure about.
If it is VRM high, then I can use a replacement with less capacitance, but if it's not, I think I'd better match it.

No, the LMB is not part of the VRM high if it's 6.3V... VRM high is 12V, so only 16V caps are on it. It's part of the VRM low and I can prove it to you -- take a multimeter and check continuity between its positive lead and the positive lead of one of the short Panasonic FL VRM low caps. You will see continuity because they are all in series.

The reason Asus went with a taller part is because this cap doesn't sit under the heatsink. By the way, you have the spec wrong, it's 0.020 ESR and 1910 ripple. Spec on the datasheet goes according to cap dimensions, and that's the spec for 8x20mm. The spec you cited is for a 10x12.5mm part.

Use the nichicon 8x8, or if you feel you must have an equivalent, use this instead.

momaka wrote on 2025-04-30, 15:01:

Seems a good deal of former BCN members have "flocked" to this place. Not sure if it's because of interest in retro PCs (which doesn't seem to get that much attention there anymore) or the forum update. For me it's a little bit of both, I suppose. Though I only discovered VOGONS a little over a year ago. Still active on BCN too, just not as much. The new forum update is PITA on many levels.

The forum update doesn't bother me, but the computer side of BCN is a lot less active. This forum has the opposite problem, though. It's too busy for me.

mockingbird wrote on 2025-05-09, 21:33:

There's also a brand called KEMET, which is significantly cheaper than other brands. Seems to be an American company? Due to the much cheaper prices, my gut feeling is to stay away, but please tell me if I'm wrong!

It's impossible to know for certain what Kemet capacitors actually are -- they are very tight-lipped and their capacitors should be considered unproven.

Kemet is owned by Yageo, so I've been assuming that they're the same as Teapo (formerly owned by Yageo, now part of Jamicon). I don't trust Yageo for critical stuff, but I have bought Kemet caps for test boards and low-stress applications (like audio).

This thread makes me wonder if I've been getting sloppy with my cap choices.

lti wrote on 2025-05-11, 23:29:

Kemet is now owned by Yageo

Fixed. They bought them not too long ago. Shortly thereafter, they started selling Kemet-branded polymer capacitors.... One could draw the inference that Kemet polymers are Taiwanese -- that would mean decent quality.

I don't trust Yageo for critical stuff, but I have bought Kemet caps for test boards and low-stress applications (like audio).

I don't touch the old American brands when it comes to electrolytic caps-- Illinois Capacitor, Cornell Dubilier, Epcos -- these are artifacts of long-gone corporations that were swallowed up by multinationals. You want to be a multinational? Fine. You don't want to be transparent about where the product is from and who made it? aint touching your stuff - don't trust your quality.

This thread makes me wonder if I've been getting sloppy with my cap choices.

I only buy Yageo resistors personally... Teapo has some decent stuff out there.

mockingbird wrote on 2025-05-11, 21:15:

No, the LMB is not part of the VRM high if it's 6.3V... VRM high is 12V, so only 16V caps are on it. It's part of the VRM low and I can prove it to you -- take a multimeter and check continuity between its positive lead and the positive lead of one of the short Panasonic FL VRM low caps. You will see continuity because they are all in series.

Looks like I was right not to trust myself in identifying these caps!

mockingbird wrote on 2025-05-11, 21:15:

The reason Asus went with a taller part is because this cap doesn't sit under the heatsink. By the way, you have the spec wrong, it's 0.020 ESR and 1910 ripple. Spec on the datasheet goes according to cap dimensions, and that's the spec for 8x20mm. The spec you cited is for a 10x12.5mm part.

Thank you! Yes I didn't realise that I needed to match dimensions too

mockingbird wrote on 2025-05-11, 21:15:

Use the nichicon 8x8, or if you feel you must have an equivalent, use this instead.

I will trust what you taught me and go with what the VRM lows need, rather than just matching the original and picking a lesser (I think) brand. Can I use:

8x8 Nichicon 820uF 2.5V 0.006 6100 Polymer
https://uk.farnell.com/nichicon/rl80e821mdnas … 2-5v/dp/2841880

or

8x8 Nichicon 1000uF 6.3V 0.009 5700
https://uk.farnell.com/nichicon/rl80j102mdn1k … 6-3v/dp/2841881

Since the LMB is VRM low, I think I can use the 2.5V Nichicon since it's not going to see voltage higher than around 1.4V, so going with the 2.5V will give me slightly better ESR and Ripple specs, right?

I guess Asus didn't go with the Panasonic FL for this is for cost saving, not because they needed the 1500uF? Perhaps the Ltec LMB 1500uF was simply cheaper

randomdavid wrote on 2025-05-12, 09:55:

<snip>Can I use:

8x8 Nichicon 820uF 2.5V 0.006 6100 Polymer
https://uk.farnell.com/nichicon/rl80e821mdnas … 2-5v/dp/2841880

Yes, use this -- it's 5000 hour as opposed to 2000 hour rated.

mockingbird wrote on 2025-05-12, 10:48:

Yes, use this -- it's 5000 hour as opposed to 2000 hour rated.

It should be noted that even the 2000 hour caps will most likely outlive us all, so... no need to be very picky. 😅

Hi,

I have an update for this saga. I'm not sure if I should comment on this thread or create a new one, mods please let me know!

I have purchased the caps and a Proskit S-331-H desoldering station after seeing good reviews.

Today I finally put the plan into motion. The good news is: I managed to replace the 5 VRM High KZG caps and the board still works! I played CS 1.0 on it for 30 mins and it was rock solid!

Also, I'm not sure if this is due to the new caps, but the VRMs seem to be running a lot cooler now, because my unscientific measurement of finger touching the VRM heatsink shows that it's a lot cooler than before.

Here's some pics of the result:

Now, the bad news, and I seriously need help on this one:

Even with the desoldering gun, it was still a struggle! Especially for the ground holes (i think they are ground holes, they are the holes with 2 or 3 smaller holes surrounding them)

Today for nearly 4 hours, I did this on repeat:

Add flux, add 60-40 leaded solder, use gun, see that the hole is still not clear.

I started at 350C and went all the way to 450C, no dice. Held the gun at the hole, 90 degrees angle, for 10 seconds all the way to a minute, no dice.
Because of this, I damaged the board a little! A friend told me the lacquer is burned?
Thankfully the board still works!

What ended up working was to first remove the cap by heating the joint, then sandwich the hole between the soldering iron at 390 C and the gun at 450 C for 10 seconds, then suck. I feel that this is way too extreme and there should be a better way.

Note that the gun works fine, because the other non-ground hole was super easy, just add flux, new solder, suck, and it's all out.

What am I doing wrong? Should I try to preheat with a hair dryer? Do you guys have any advice for me?

I heard that the VRM low caps are supposed to be even more stubborn, so I will need all the advice I can get!

For the large ground / power planes, I recommend helping your soldering iron with a hot air station or preheater, otherwise all that copper will quickly absorb the heat, even if you go above 400C.

bloodem wrote on 2025-05-25, 20:30:

For the large ground / power planes, I recommend helping your soldering iron with a hot air station or preheater, otherwise all that copper will quickly absorb the heat, even if you go above 400C.

Yes, I think I have seen others mentioning preheating, but I haven't found one that directly talked about desoldering lead-free solder in ground planes...

Do you think I can use a hair dryer for this? This doesn't feel like it would require the precision of a proper hot air station, so maybe I can get away with a hair dryer and a narrow nozzle?
Plus a hot air station would be another £60-80...

I got some from here https://vi.aliexpress.com/item/10050030202345 … ayAdapt=glo2vnm
6.3V 1000uF - I also got a LCR-T7 and they tested pretty good on it - and yes, if you plan on getting aliexpress caps, definitely get one of those testers.
Also, if you can remove the old caps without distressing them too badly, the tester can tell you how bad they were