Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

[bw1]

Got a Dimension 4700C in the other day (newer-ish slim tower) with all (8) Rubycon MCZ V-core (out) caps bulging (1600uf 6.3v) and leaking a LOT.

Curious as to what you'd recommend replacing them with? She said the computer ran fine up until the other day when it wouldn't turn on. LED on front lit amber instead of normal red.

According to my records we put a new power supply in here 2 years ago. Didn't see anything at that time, but wasn't necessarily looking as closely for this back then. Thought I should mention it at least.

[Uranium-235]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

mcz's or mbz's. check and make sure the cpu fan isn't clogged up and causing excess heat. those should not bloat for no reason, I can only think that heat, or a bad PSU might be the reason why. Also, make sure to try a different CPU, cause those caps might of over volted that one, but not until you replace them. get a PSU tester and make sure the vore isn't too high

afterthought-make sure you clean and re-apply thermal paste on that CPU too, if the heatsink isn't moving enough heat off it, it could cause the board to overheat

[yyonline]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

A Prescott P4 and a small form factor case are a recipe for overheating the capacitors, leading to premature failure.

MBZ and MCZ have been discontinued. Badcaps.net sells Nichicon HN:

https://www.badcaps.net/store/produc...roducts_id=152

Those should be perfect replacements for the blown MCZ.

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

I'll probably go with the HN's. I'm assuming Nichicon has their act together, from what I've researched in other threads here.

I should mention, I do have some 1800uf/6.3v Samxon GC's here, and the impedance is right on, but ripple is just slightly lower, at 2220 (vs. 2350). I'm never one to replace with lower-rated parts, but if you guys said that's close enough I'd consider it here. Otherwise, I'll wait and order more HN's.

Thoughts?

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Otherwise, would these polymers be compatible?

If the rubycon's didn't last, maybe it's a good idea to do poly's? I'm pretty indifferent between HN and poly, but for <$10 more, I'd consider it.

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Yes, those polys will work. Even better news: You're only gonna have to use 5 of them (based on ESR and ripple ratings). Actually it's a good idea to just match, not lower the original ESR, you could disturb the VRM otherwise. So use 5 polys and the board should have a long life.

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

There was one oddball, exact same cap as the v-core, but right next to the P4 12v plug that was also bulged out. I suppose you'll want a pic -- camera in truck. If necessary, I can. Wondering if I can do a single poly there? Otherwise, I don't have anything that would work in stock other than the Samxon GC mentioned above... and that's slightly under spec.

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Use a lower capacitance poly in that case. Try 820uF.

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Replace the existing (single, non-v-core) 1800uf electrolytic with an 820uf poly? Just making sure. Seems like a big jump.

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

The capacity used is high just to be able to get low enough ESR from an aluminum electrolytic. More capacity just gets more hold-up time, which is not at all required in a motherboard.

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

I was under the impression that some hold-up was used to further filter the rippled DC, no?

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Sure, but at the hundreds of kHz a VRM operates at, the capacitive reactance component has long become extremely low, with ESR being the limiting factor. In other words, if ESR would have been zero at that capacitance, there would have been zero ripple.

[kc8adu]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

post a pic.i bet the lone cap is on vtt.
check to see if it is connected to vcore.
even though 5 may do the trick i fill all spots with polys.this has not failed me yet.

[PCBONEZ]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Use a lower capacitance poly in that case. Try 820uF.

Randomly suggesting that outside of a VRM without analyzing the circuit can be very bad advice.
There are many places where uF matters.

Even in a VRM reducing cap count and/or uF ~just~ because the replacements are polymer is bad advice.
- There is minimum amount of uF needed even in a VRM.
- Not all polymers will have ESR lower than the original lytics.

If you are going to recommend a reduced cap count in a VRM Vcore then you want to make sure the asker knows to spread them around in the circuit rather that grouping them all in one spot. [Reason is trace length effects between the caps and the CPU. Want to the many CPU Vcore pins as close to exact as possible.]

The specific suggestion [for Vcore, not the odd cap] works in this case but you are giving the impression it works -just- because the proposed new caps are these 'magical' polymers.
- There is more to it than that and you should point that out so you don't give someone reading this later on the wrong idea.

~~
After looking at VRMs on many motherboards with polys as original I think it's safe to say 5500uF is a reasonable minimum value for the total in Vcore for CPU's with TDP in the ranges of P4[478/775] and C2D. Some really cheap boards push it down to around 5000uF [sometimes less] and better boards usually have more. Boards that only use low power CPUs [laptop processors or Atom and the likes] usually have less Vcore uF. Boards that support tron guzzling CPU's tend to have more Vcore uF.

8x 820uF [with appropriate ESR] will put it at 6560uF total uF and they run about $0.50 anymore.
.

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

- [B]Not all polymers will have ESR lower than the original lytics.
The specific suggestion [for Vcore, not the odd cap] works in this case but you are giving the impression it works -just- because the proposed new caps are these 'magical' polymers.

You're only gonna have to use 5 of them (based on ESR and ripple ratings).

http://search.digikey.com/scripts/Dk...105-ND&x=0&y=0
VS: https://cdn.badcaps-static.com/pdfs/...9fb8ee1c9c.pdf

[PCBONEZ]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Thanks.
I read that before I posted and you're being a moron.

The point was you are implying: "Because they are 'polymer' they have better ESR."
- And that's a load of bullshit.

You are destined to screw someone over saying crap like that.
.

[Th3_uN1Qu3]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Thanks.
I read that before I posted and you're being a moron.

Thanks.

The point was you are implying: "Because they are 'polymer' they have better ESR."
- And that's a load of bullshit.

You are destined to screw someone over saying crap like that.

If someone blindly uses polymers in a random power supply circuit without checking the specs, then he screwed himself over.

[bw1]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Thank you both for your help so far -- appreciated. I think it best if I back up a bit and post a picture to help explain both. Normally my pics are nice and clear - this one was a bit hasty, but hopefully will suffice.

As you can see, there is room for 10 v-core caps. This space was populated by 8 1800uf/6.3v 8x20mm MCZ's. To be honest, I really wouldn't care if these were replaced by lytics or polys. My reason for leaning poly was that they were already MCZ's and all went out, plus, Digikey doesn't appear to have any lytics in stock that would work.

I've also marked the one oddball below the v-core. Again, I would be happy to replace with lytic, but can't seem to find any and would really prefer to order from Digikey as they are close and I already have an order started with them.

[PCBONEZ]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

I would use 8x 820uF Poly mostly because of the heat there.
The ESR on the MCZ was .012 so go for something with less.
[And a lytic at 'stock' uF for the odd one.]

Digikey does not carry any Lytics better than Panny FM unless they recently added something.
And AFAIK all Mouser carries in lytics with better specs than FM is Nichicon HN and HM.
BCN should have what you need too.

Are those black/gold caps to the left HN and if so what is the date code?
Those dated '01 thru '04 were defective.
.

[lti]

Re: Dell Dimension 4700C - V-Core Rubycon MCZ's were ALL bad

Digi-Key started selling Nichicon HM and HZ a few months ago. They are a normally stocked item with no minimum quantity. They sell HN as well, but not very many values are stocked.