Which caps to replace: ECS P4M800Pro-M V1.0A

[Mattz_GT]

Getting some Nichicon polymer 1000uF 6.3V for the RAM VRM, so the CPU VRM out and in are getting changed, and the RAM VRM, I don't think I'll bother with the general decoupling ones all over the board unless it's still not stable

[momaka]

^ Flux is a must for me.
Yes, you can (and should) use fresh solder on top of the old leads to improve the heat transfer from your soldering iron to the board/joint. But once you have the cap pulled out, you might find that clearing the hole can become a bit of a "mess" if you just keep adding solder. At some point, just adding flux and heating with your iron is more helpful than adding fresh solder.
As for clearing the holes, I prefer doing it the "manual" way, with a (stainless steel) sewing needle - basically heat each hole with the soldering iron and then push through the hole with the sewing needle from the other side to push the solder out. For boards with really thick copper planes, I find neither desoldering braid nor a vacuum pump / solder sucker to be of any use - some leftover solder always tends to stay in the hole.

yes i think that was the problem. iirc gp caps arent meant for smps use. the high ripple generated by the switching frequency of the smps will easily fry any gp cap. for smps, caps which are said to be low-esr or better should be used.

Exactly.
OST RLG tend to be pretty stable when used in low-stress circuits... at least that's been my experience with them so far.

well, u could swap between the two gpus as needed... depending on if u wanna play on win98 or winxp.

The only problem with that is if you do it often enough, it will eventually wear the contacts in the AGP slot and it might stop detecting cards properly after a while. I'm running into this issue already with the PCI-E slot on my video card test PC - it has about only 10-15% chance it will detect a GPU now... and quite often, not always in full 16x mode. So I've started using PCI-E 1x-to-16x crypto miner risers to test GPUs. Unfortunately, this has the side effect of the card not being able to reach its full potential - even under a single core Athlon 64 4000+. And I can't know if a GPU has problematic PCI-E lanes or not. But at least this I can still test if the rest of the GPU portion (and connected memory) is OK.

[Mattz_GT]

the pentium 4 and pentium d cpu design called the netburst microarchitecture is designed to scale with more mhz or ghz, so the faster ram speed is more helpful than tighter timings. try doing a mem bandwidth test or benchmark like in sisoft sandra, aida64 or maxxmem2. u should have overall higher memory reads, writes and copy speeds at 667 5-5-5 than u do at 533 4-4-4.

or u can just try to get low latency ddr2-667 at cl4. most of not all of the low latency ddr2-1066 and ddr2-800 from the good brands like kingston, g-skill and mushkin etc. seem to be all but sold out on junkbay. but i noticed plenty of unwanted low latency ddr2-533 and 667 available on junkbay. if u're lucky, it may even do ddr2-667 speeds at 3-3-3-9 or 3-4-4-10. the tighter timings also result in slightly higher mem bandwidth which results in better performance. hope this helps in squeezing out every inch of performance out of your system which is the point of any overclocker. hehehe

Double trouble with the RAM, I struggled to find the required low density 2R8, anything faster is probably going to be high density, anyway, kit at the ready (hmm, maybe I ought to get some flux... otherwise I know that putting a bit of fresh solder on can help melt the old solder, hopefully this predates RoHS anyway.
Wish me luck, I'm going in, well, tomorrow anyway

[ChaosLegionnaire]

Funny enough, the only OST cap I've seen fail was an RLG series - their general purpose 105C caps. It was in a Bestec power supply though (in a Dell Vostro 200), so it probably wasn't set up for a long life.

yes i think that was the problem. iirc gp caps arent meant for smps use. the high ripple generated by the switching frequency of the smps will easily fry any gp cap. for smps, caps which are said to be low-esr or better should be used.

Can't make it a dual boot if I put my best AGP card (Hd3650) in what is my best AGP host, as there is no Win9x driver... Looks like 9x support ended with the 9800 (my second best AGP card).

well, u could swap between the two gpus as needed... depending on if u wanna play on win98 or winxp. but anyway, since u put a pentium d in there, the second core of the pd wont work on win98... so yea, like u said, maybe build another agp system for a single core cpu on win98 and with the radeon 9800 card instead.

Well, no tears shed here on my part here. Winblows 98 is maybe truly only useful for a handful of games. Most everything from the late 90's will run OK on 2k and XP... with some patches, at worst. W98 is mostly just for re-living the 90's retro experience... and maybe some older DOS games, I suppose.

just some games that wont work right on winxp have to be played in win98. like for example, in descent freespace, the afterburner doesnt seem to work on winxp. the controls are also glitchy on some machines on xp. have to play it in its native win9x environment.

but if the game works fine on both 98 and xp, better to play it on xp because while even though xp has the slightly higher mem footprint, it has better cpu and gpu performance so games run faster on xp than they do on 98. so that should clarify things for those who arent sure of whether if "should i play it on 9x or xp?"

The chipset... VIA P4M800Pro is single channel (and could support Core2 but only in the V2.0 of this board). The VIA PT880Pro is dual channel and still hybrid DDR/DDR2

yea the p4m800 is just the earlier chipset that came out. the pt880 is the later version that came out so better/more features for the mem subsystem.

I'll benchmark 533 4-4-4 vs 667 5-5-5 - so the latencies are actually the same, but the 667 can burst a little higher, but I wonder if the 3/2 FSB to RAM ratio at 533 might be "cleaner" than the 6/5 ratio at 667.

the pentium 4 and pentium d cpu design called the netburst microarchitecture is designed to scale with more mhz or ghz, so the faster ram speed is more helpful than tighter timings. try doing a mem bandwidth test or benchmark like in sisoft sandra, aida64 or maxxmem2. u should have overall higher memory reads, writes and copy speeds at 667 5-5-5 than u do at 533 4-4-4.

or u can just try to get low latency ddr2-667 at cl4. most of not all of the low latency ddr2-1066 and ddr2-800 from the good brands like kingston, g-skill and mushkin etc. seem to be all but sold out on junkbay. but i noticed plenty of unwanted low latency ddr2-533 and 667 available on junkbay. if u're lucky, it may even do ddr2-667 speeds at 3-3-3-9 or 3-4-4-10. the tighter timings also result in slightly higher mem bandwidth which results in better performance. hope this helps in squeezing out every inch of performance out of your system which is the point of any overclocker. hehehe

[Mattz_GT]

Interesting, the board doesn't support dual channel because of the dual RAM type design?

The chipset... VIA P4M800Pro is single channel (and could support Core2 but only in the V2.0 of this board). The VIA PT880Pro is dual channel and still hybrid DDR/DDR2
Also, the chipset officially supports 533, but the BIOS allows setting 667 (and my RAM is 667), so when it's all good, I'll benchmark 533 4-4-4 vs 667 5-5-5 - so the latencies are actually the same, but the 667 can burst a little higher, but I wonder if the 3/2 FSB to RAM ratio at 533 might be "cleaner" than the 6/5 ratio at 667.
Also, I think I'll bench using an older Cinebench rather than a pure RAM test, as real world access patterns may behave differently from flat out sequential and being back in step every 3 FSB clocks instead of every 6 might be important

[momaka]

No idea where I'd get replacement tips, they are screw in (Aldi Workzone / Lidl parkside).

Ah, I see.

I do have a curiosity in regards to the quality and performance of these Parkside tools. From what I gather, they are cheap, but aren't always terrible cheap garbage... though it really depends on what you get, I suppose. Some of their stuff is right down there with Harbor Freight Tools (in the US/Canada.) Still beats no-name Amazon and other China e-mall stuff for the most part.

Different versions of the board used different complements of caps.
The 1.0 used 9 short ones around the CPU and none in the other row. The 1.0A has the 6 and 3. The 2.0 looks like it has just the 6 short, but doubles up the mosfets.

Hehe, typical ECS. We'll use what we got when we've got it
The three tall ones on that TRW picture appear to Panasonic FL. But knowing ECS, some boards probably use OST RLX instead... which are pretty terrible (at least the 8 mm dia. caps always are.)

Strangely, it held together well enough to boot into DOS and flash the BIOS (needed to as I upgraded from a C1 stepping D915 to a D0 stepping D945 and the old BIOS wouldn't take the D0).

Oooff. glad it did. I wouldn't ever attempt to flash anything on hardware that has bad caps - never know when data corruption will screw something up and possibly brick the device.

Can't make it a dual boot if I put my best AGP card (Hd3650) in what is my best AGP host, as there is no Win9x driver... Looks like 9x support ended with the 9800 (my second best AGP card).

Well, no tears shed here on my part here. Winblows 98 is maybe truly only useful for a handful of games. Most everything from the late 90's will run OK on 2k and XP... with some patches, at worst. W98 is mostly just for re-living the 90's retro experience... and maybe some older DOS games, I suppose.
In terms of software, I have everything I need from the XP era, which often runs fine even on modern Windows versions (Office 97 and Photoshop 7, anyone?)

I still have an eye out for a potential better AGP platform, Core 2 would be nice but a few I've seen are a hellacious price compared to just building on an unloved PCIE platform instead.

Yeah, those prices make no sense really.
Funny enough, I do have two AsRock s775 AGP boards that I got in the last few years, one of which was free and the other dirt-cheap ($3)
So it's also down to luck. The bay is not such a great place for that anymore. I guess the word has spread about the retro hardware and now too many sellers are demanding insane prices.
FWIW, I did also sell some retro hardware last year, but I started it all with a low bid and let the Ebay zombie crowd determine the final price (surprisingly, the items still sold for a lot.)

My first time around XP experience was a P4 640 with a X740 (Medion), so this will be 200 more MHz, two actual cores instead of HT and way more GPU, though single channel DDR533 (667 can be set in BIOS - an OC for the chipset) gives away some RAM performance over the dual DDR400 I had before

Interesting, the board doesn't support dual channel because of the dual RAM type design?

[Mattz_GT]

No idea where I'd get replacement tips, they are screw in (Aldi Workzone / Lidl parkside). Different versions of the board used different complements of caps.
The 1.0 used 9 short ones around the CPU and none in the other row. The 1.0A has the 6 and 3. The 2.0 looks like it has just the 6 short, but doubles up the mosfets.
Strangely, it held together well enough to boot into DOS and flash the BIOS (needed to as I upgraded from a C1 stepping D915 to a D0 stepping D945 and the old BIOS wouldn't take the D0).

Wasn't that bothered about the hybrid RAM support... came with 2x 512MB DDR2 and upgraded to 2x 1MB after finding that it had to be low density and getting the right ones.

Can't make it a dual boot if I put my best AGP card (Hd3650) in what is my best AGP host, as there is no Win9x driver... Looks like 9x support ended with the 9800 (my second best AGP card).
I still have an eye out for a potential better AGP platform, Core 2 would be nice but a few I've seen are a hellacious price compared to just building on an unloved PCIE platform instead.

My first time around XP experience was a P4 640 with a X740 (Medion), so this will be 200 more MHz, two actual cores instead of HT and way more GPU, though single channel DDR533 (667 can be set in BIOS - an OC for the chipset) gives away some RAM performance over the dual DDR400 I had before

[momaka]

now waiting for a solder sucker I had to order as I can't find the one I had - just a plunger type, was tempted by a heated plunger type, but they seem to have a bad reputation for the mechanism failing (or an electromagnet one notorious for failing and tripping the breakers.

Well, I personally never liked using solder sucker / plunger for recapping motherboards anyways. I find the vias are too small and the copper layers are too thick (especially around the CPU), making it harder than it should be to remove caps. Same with braid. Instead, I just use the "rock back-n-forth" method - basically heat one cap lead and tilt the cap away from the board from that side. Then do the same for the other cap lead... then back to the first... and so on, until the cap is out. Usually it takes 15-30 seconds per cap, tops - that is, for more stubborn caps around the CPU. The ones elsewhere on the board are easy-peasy.

Got a temperature controlled iron/station (60W), standard settings are 200C idle, 300C (it suggests for soldering) and 400C (it suggests for desoldering), can adjust, choice of wedge or conical tip, TBH I've never used anything other than flat before.

I hope it's not a cheap clone that uses 900M -style of tips. For motherboards, I recommend nothing less than T12 -based irons, since the tips have much better temperature control. The fact that I prefer to use an old and bulky 40W soviet-era soldering iron over 900M -based irons should say what I think of them.
As for tips, conical is usually *last* on my list. Bevel (wedge) or chisel are the two I use the most. Knife / flat is so-so. It's all about heat transfer, and the tips that give me the best of it are the ones I use most often.

also from my experience, ecs boards tend not to be overclockable anyway...

Yes, more modern ones anyways.
And I think this one falls into that category too. If you look at the MOSFETs for the CPU VRM, you can see there are a few more spots where they could have added MOSFETs in parallel to the ones already on the board, but the didn't. So the CPU VRM probably got gimped a little on this board and may run a little hotter with very power-hungry CPUs (or if you try to run a more aggressive OC... if the board BIOS even allows it.)

u also have to be careful as some of the cap spots are in ambiguous positions. for example at the cpu vrm, there is an ambiguous cap spot at the top right of the cpu socket at the edge of the board next to the heatsink mounting hole for the cpu cooler. that could either be for the cpu vrm input which gets 12v or the output which gets 1.2v to 1.4v depending on the cpu.

It looks like all of them are connected to that lighter purple trace on top, so more than likely they are all for the CPU V_core.
That said, I just now notice how many free cap spots they left for the CPU V_core filter caps. I'm counting 6 free spots. Wow! And if you count the ones already installed, that's space for 17 caps total! I think that's a record (or close) for a CPU VRM. At worst, a few may be for CPU Vtt... , which is still less than 2V, so can still use 4V -rated polymer caps.
But dang! - That's a lot of filtering potential. With bigger inductors, I think ECS could have made quite an OC monster here.

anyway, this is a nice board well worth repairing as its one of the rare boards with hybrid ddr1 and 2 support. the agp slot and software support also allow a wide range of oses to be installed on it from win98 to winxp. so there is a wide degree of software flexibility on this board. a dual boot os setup would be the most common for this board.

Concur.
.

[ChaosLegionnaire]

Wasn't sure about altering values, or even overcapping the board, there are quite a few extra positions for the 6.3V ones around the CPU

that depends on whether u wanna overclock or not. if u do, increasing the capacitance and ripple rating and decreasing the esr values can help in better overclocking results. however, doing that also results in larger can sizes and the caps might not fit. also from my experience, ecs boards tend not to be overclockable anyway...

u also have to be careful as some of the cap spots are in ambiguous positions. for example at the cpu vrm, there is an ambiguous cap spot at the top right of the cpu socket at the edge of the board next to the heatsink mounting hole for the cpu cooler. that could either be for the cpu vrm input which gets 12v or the output which gets 1.2v to 1.4v depending on the cpu.

another ambiguous cap spot is the one by the southbridge near a screw mounting hole for the board. it could be for the sb, agp or ram vdimm. if it is indeed for the ram vdimm, it is not advisable to use rubycon yxj on it as suggested by momaka as there is a buck regulator near the main atx connector most likely for the ram vdimm. the regulator is needed to supply the correct voltage to the ram depending on what type of ram is installed. 1.8v for ddr2 and 2.5v for ddr1. there seem to be 6 or 7 caps around the ram for filtering the vdimm so i strongly suggest u avoid using yxj for the 6-7 caps around the ram. i recommend panasonic fr 1000µf 6.3v Ø8x11.5mm caps for the ram vdimm caps instead.

anyway, this is a nice board well worth repairing as its one of the rare boards with hybrid ddr1 and 2 support. the agp slot and software support also allow a wide range of oses to be installed on it from win98 to winxp. so there is a wide degree of software flexibility on this board. a dual boot os setup would be the most common for this board.

[Mattz_GT]

Got the caps, now waiting for a solder sucker I had to order as I can't find the one I had - just a plunger type, was tempted by a heated plunger type, but they seem to have a bad reputation for the mechanism failing (or an electromagnet one notorious for failing and tripping the breakers.
Got a temperature controlled iron/station (60W), standard settings are 200C idle, 300C (it suggests for soldering) and 400C (it suggests for desoldering), can adjust, choice of wedge or conical tip, TBH I've never used anything other than flat before.
Found my side cutters, thought I was going to have to replace those too, and if I'm lucky, might find the braid I used last time.

Wasn't sure about altering values, or even overcapping the board, there are quite a few extra positions for the 6.3V ones around the CPU

[momaka]

Been a while and wondering how this went.

Cap choices....
870025174008 | Wurth Elektronik 680μF Through Hole Polymer Capacitor, 6.3V dc | RS (WCAP-PTG5) or A750KK687M0JAAE016 | KEMET 680μF Through Hole Polymer Capacitor, 6.3V dc | RS (A750)
2000H 105C 5100mA ripple 10mOhm / .. 4100mA ripple 16mOhm - so the Wurth are superior? Checked again, definitely read TMJ on those.
The 1800uF are 3.5mm pitch, that narrows my choice to one Panasonic 1800μF Aluminium Electrolytic Capacitor 6.3V dc, Radial, Through Hole - EEUFS0J182L | RS (FS)
The 1500uF 16V are 5mm pitch Wurth Elektronik 1500μF Aluminium Electrolytic Capacitor 16V dc, Radial, Through Hole - 860010375019 | RS (WCAP-ATG8) ...only 1100mA ripple ..Nichicon 1500μF Aluminium Electrolytic Capacitor 16V dc, Radial, Through Hole - UPW1C152MHD | RS (PW) better... only other option that are bagged 5
Struggling to pick the 1000uF 6.3V, searching for longer lifetime rating takes me to Rubycon 1000μF Aluminium Electrolytic Capacitor 6.3V dc, Radial, Through Hole - 6.3YXJ1000M8X11.5 | RS - Rubycon YXJ

For the polymers, you can drop down the voltage rating and use ones with higher capacity.... though both Wurth and the Kemet polymers you picked should work and will likely have lower ESR than the TMJ electros.
For the 1800 uF KZG, Panasonic FS should be OK to use in most cases. It's not quite up to the specs of KZG, but most motherboards won't have an issue.
For the 1500 uF 16V caps, see if there is Panasonic FM, FR, or FS in that size. If not, even going down to 1200 uF will be OK (or upsizing to 1800 uF)... whatever is available. Polymers might work too 16V 470 uF should suffice for the CPU VRM input.
Lastly, in regards to the OST RLS series... Rubycon YXJ series are good, but not quite on par with RLS in terms of impedance (almost half.) That said, I suspect most of the OST RLS caps on your board are used for linear circuits and general filtering on PSU rails... so in most cases, the Rubycon YXJ will work too. However, if there are any RLS caps that filter a buck rail and you put YXJ in there, the higher impedance of the YXJ might increase the ripple and possibly cause errors / instability.
If there really is nothing else that matches the RLS apart from the YXJ, just grab a few YXJ and keep them in a drawer or replace only the RLS that are near PCI slots, USB headers, and rear ports area.

[Mattz_GT]

The 680 uF 4V UCC TMJ (is it really TMJ series??) definitely need to go. Many of their TM--- series are prone to failure, especially when used around the CPU VRM of a motherboard.
The 1800 uF 6.3V KZJ are also quite problematic (as are KZG with the same capacitance and voltage spec... and in general.)
The 4x OST (I'm gonna guess they are dark blue with gold writing and RLX series) 1500 uF 16V *probably* should be replaced, just in case... even though I find these to be a lot less likely to fail.
The small 6.3V 1000 uF OST (RLP series??) may be problematic too, or maybe not (see info further in my post below.) I'd replace at least a few of them around the RAM and chipset (Northbridge and Southbridge) areas.

1. TMJ... look like this 680uf-4v-8x8mm-tmv.jpg (1000×1000) says TMV, but pictures TMJ
3. Yes, that's what they look like
4. They are RLS.

[Mattz_GT]

Cap choices....
870025174008 | Wurth Elektronik 680μF Through Hole Polymer Capacitor, 6.3V dc | RS (WCAP-PTG5) or A750KK687M0JAAE016 | KEMET 680μF Through Hole Polymer Capacitor, 6.3V dc | RS (A750)
2000H 105C 5100mA ripple 10mOhm / .. 4100mA ripple 16mOhm - so the Wurth are superior? Checked again, definitely read TMJ on those.
The 1800uF are 3.5mm pitch, that narrows my choice to one Panasonic 1800μF Aluminium Electrolytic Capacitor 6.3V dc, Radial, Through Hole - EEUFS0J182L | RS (FS)
The 1500uF 16V are 5mm pitch Wurth Elektronik 1500μF Aluminium Electrolytic Capacitor 16V dc, Radial, Through Hole - 860010375019 | RS (WCAP-ATG8) ...only 1100mA ripple ..Nichicon 1500μF Aluminium Electrolytic Capacitor 16V dc, Radial, Through Hole - UPW1C152MHD | RS (PW) better... only other option that are bagged 5
Struggling to pick the 1000uF 6.3V, searching for longer lifetime rating takes me to Rubycon 1000μF Aluminium Electrolytic Capacitor 6.3V dc, Radial, Through Hole - 6.3YXJ1000M8X11.5 | RS - Rubycon YXJ

[andrewsawesomr]

Funny you say that. I was just playing around with some black OST RLP's on my bench that I pulled from a Biostar motherboard a few years back, due to seeing a few of the same kind already bulged. I was hoping I could re-use them in non-critical areas (3.3V rail near PCI slots and 5V filter for USB) on another old motherboard. What's funny is that they all read 30-50% over their stated capacitance - i.e. around 1300-1500 uF when they should be just 1000 uF. Clearly they've become electrically leaky. But outside, they look perfectly fine. And no, these were all far away from heat sources. The ones that weren't on that BIostar motherboard were already bulged/bust.

So I think with OST it really depends on the series and also on the specific batch. For example, I have older OST RLP with same capacitance and voltage rating (1000 uF 6.3V) from three different boards that are still completely fine after all these years. At the same time, I have another board that had all of its OST RLP caps go bad in a similar fashion. So it seems some batches of the RLP series are OK and some were not. The same seems to apply to their other series too - e.g. RLX. I find old RLX to be pretty reliable in 10 mm diameter and the higher voltages (10V rating or more), but the smaller 8 mm diameter and only 6.3V rated ones to turn "iffy" over time. I think their RLS series seem to be the most stable from the bunch... but even these I have seen failed.
So I always take OST with caution. If none have ever gone bad on a board and a few read well within their parameters when tested, then I may leave them alone for a little while longer or until failed. But if I see one or two failed from a certain batch, they all need to go from that batch.

Funny enough, the only OST cap I've seen fail was an RLG series - their general purpose 105C caps. It was in a Bestec power supply though (in a Dell Vostro 200), so it probably wasn't set up for a long life.

[momaka]

Examined board closely...
By CPU, cooking under heatsink, 6x 680uF 4v TMJ, looking closer, the 2 on one side look like they have blown their bungs, glad to see that as now I can be pretty certain I'm looking at the culprits
Spaced away, 3x 1800uF 6.3V KZJ - I'm guessing they are also CPU VRM side as other versions mix up the use of those and the short ones
12v side, 4x OST 1500uF 16V

Rest of the board, MANY OST 1000uF 6.3V, if OST are really bad, then I'm in big trouble, but their positions are probably not as critical

The 680 uF 4V UCC TMJ (is it really TMJ series??) definitely need to go. Many of their TM--- series are prone to failure, especially when used around the CPU VRM of a motherboard.
The 1800 uF 6.3V KZJ are also quite problematic (as are KZG with the same capacitance and voltage spec... and in general.)
The 4x OST (I'm gonna guess they are dark blue with gold writing and RLX series) 1500 uF 16V *probably* should be replaced, just in case... even though I find these to be a lot less likely to fail.
The small 6.3V 1000 uF OST (RLP series??) may be problematic too, or maybe not (see info further in my post below.) I'd replace at least a few of them around the RAM and chipset (Northbridge and Southbridge) areas.

If I take it from RS, any reason not to go with the Wurth Elektronik at £0.206 ea (pk 5 - 2 packs)
For the second, Rubycon at £0.223 ea (pk 10) more than I need
For the 3rd, oh, back to Wurth at £0.165 ea (pk 5)
And some Panasonic 1000uF 6.3V to cover the last ones - I don't really want to do all the less important ones, well, depends how good at it I get, ah, I DO have another dead board I could practice on

What series of these caps are you planning to use? Like to the exact product would be nice to see, just to make sure you're not buying some general purpose 85C caps or the like.

In regard to the aliexpress comment in your previous post... YES, it's better to avoid AliEXP for electrolytic caps, because you may not get good quality ones (especially if the listing makes no mention of the brand or series specifically.) However, I have heard that when it comes to polymers, AliEXP is not that bad. Actually, most stored that are operating on Ebay probably also have an AliEXP store front too. Still, I myself prefer ebay. Here's for example some quality polymers that come from a seller that I'd consider trustworthy (they've been on there for a long time.)
https://www.ebay.com/itm/235599133044

the thing about OST is if they arent cooked they seem super reliable, i have AT psu's that had working ost's in them.

Funny you say that. I was just playing around with some black OST RLP's on my bench that I pulled from a Biostar motherboard a few years back, due to seeing a few of the same kind already bulged. I was hoping I could re-use them in non-critical areas (3.3V rail near PCI slots and 5V filter for USB) on another old motherboard. What's funny is that they all read 30-50% over their stated capacitance - i.e. around 1300-1500 uF when they should be just 1000 uF. Clearly they've become electrically leaky. But outside, they look perfectly fine. And no, these were all far away from heat sources. The ones that weren't on that BIostar motherboard were already bulged/bust.

So I think with OST it really depends on the series and also on the specific batch. For example, I have older OST RLP with same capacitance and voltage rating (1000 uF 6.3V) from three different boards that are still completely fine after all these years. At the same time, I have another board that had all of its OST RLP caps go bad in a similar fashion. So it seems some batches of the RLP series are OK and some were not. The same seems to apply to their other series too - e.g. RLX. I find old RLX to be pretty reliable in 10 mm diameter and the higher voltages (10V rating or more), but the smaller 8 mm diameter and only 6.3V rated ones to turn "iffy" over time. I think their RLS series seem to be the most stable from the bunch... but even these I have seen failed.
So I always take OST with caution. If none have ever gone bad on a board and a few read well within their parameters when tested, then I may leave them alone for a little while longer or until failed. But if I see one or two failed from a certain batch, they all need to go from that batch.

[stj]

lead spacing seems to be standard related to diameter

[Mattz_GT]

Both the Wurth and the Kemet come up as 8mm round by 8mm tall in the specs, also, I guess lead spacing is another parameter I haven't checked... both 3.5mm - hopefully that is standard for what they were

[stj]

use kemet polymer for the 680's atleast because height matters.

the thing about OST is if they arent cooked they seem super reliable, i have AT psu's that had working ost's in them.

[Mattz_GT]

Examined board closely...
By CPU, cooking under heatsink, 6x 680uF 4v TMJ, looking closer, the 2 on one side look like they have blown their bungs, glad to see that as now I can be pretty certain I'm looking at the culprits
Spaced away, 3x 1800uF 6.3V KZJ - I'm guessing they are also CPU VRM side as other versions mix up the use of those and the short ones
12v side, 4x OST 1500uF 16V

Rest of the board, MANY OST 1000uF 6.3V, if OST are really bad, then I'm in big trouble, but their positions are probably not as critical
Now to read some more of the badcaps list

If I take it from RS, any reason not to go with the Wurth Elektronik at £0.206 ea (pk 5 - 2 packs)
For the second, Rubycon at £0.223 ea (pk 10) more than I need
For the 3rd, oh, back to Wurth at £0.165 ea (pk 5)
And some Panasonic 1000uF 6.3V to cover the last ones - I don't really want to do all the less important ones, well, depends how good at it I get, ah, I DO have another dead board I could practice on

[stj]

RS has good price on poly's for the vrm output and panasonic FR for the 12v input.
gonna get a handling charge for orders under £50 though,
or CPC has a minimum free delivery limit of £20 but less of a selection of caps