Found Teapo Specs - For Replacement Reference.

[PCBONEZ]

Dunno if anyone's found this one. [Or if this is the right place to post it.]

http://www.cbs.it/teapo.htm

May be a handy reference when replacing Teapos.
Use it as a guide for the ESR/ripple the Mobo design engineer presumably intended to be there.

When I dunno the specs for what's coming off I'm always wondering if I'm wasting money with over-kill or falling short of the minimum requirements.

.

[Sparky]

Re: Found Teapo Specs - For Replacement Reference.

Thanks for that info

Thing I wonder about is why do teapos seem to not last too well on mobos but seem to be quite happy in a PSU?

[Galvanized]

Re: Found Teapo Specs - For Replacement Reference.

I suppose heat.
On mainboards the MOSFETs are right next to the caps on the pcb and connected by short copper traces.
In a PSU the heat generating FETs and diodes are bolted onto the sink and the caps are on the pcb.

[willawake]

Re: Found Teapo Specs - For Replacement Reference.

hmm everybody is always talking about heat
psus are pretty hot also. some are goddamn toasty

[Newbie2]

Re: Found Teapo Specs - For Replacement Reference.

hmm everybody is always talking about heat
psus are pretty hot also. some are goddamn toasty

An example of a toasty PSU is a L&C or DEER (same thing). Poor un-reliable fan with super-low CFM, tiny thin heatsinks. Those factors can kill the already poor-quality caps by heat in it, and it would be worse the PC runs hot, or in a hot climate, or both.

[gonzo0815]

Re: Found Teapo Specs - For Replacement Reference.

I think it is the ripple current and the heat in combination why caps in pSU`s are far less stressed than on vrm.
A typical vrm should deliver an excess of 50amps or more, thus the ripple current is a real concern here.
PSu`s don't have that high current density.

[gonzo0815]

Re: Found Teapo Specs - For Replacement Reference.

I think it is the ripple current and the heat in combination why caps in pSU`s are far less stressed than on vrm.
A typical vrm should deliver an excess of 50amps or more, thus the ripple current is a real concern here.
PSu`s don't have that high current density and caps are usually in the air stream of a fan, most vrms caps not.

[linuxguru]

Re: Found Teapo Specs - For Replacement Reference.

That site also has the datasheets for most Rubycon Aluminium Electrolytics (didn't notice MBZ there, but there's the interesting 125c series ZT):

http://www.cbs.it/rubyconalum.htm

Download them before they vanish due to obsolescence...

[PCBONEZ]

Re: Found Teapo Specs - For Replacement Reference.

The Teapos they use on Mobos may be different model Teapos than they use in PSU's.
That's something those data sheets might help you figure out.

I just found out the ripple rating on the Teapos I'm replacing is just under 1000 so I don't really need anything better than Nichicon PW's to meet the Mobo designers specs on these boards. (Of course if caps with a better rating are cheaper I'll go with those, but knowing what's actually needed gives me a wider range of caps to choose from.)

I wouldn't mind finding the data sheets for some of the other common crap caps for the same reason.
- Anyone know of any to save me some search time?

I noticed the Rubycon specs after I posted but figured everyone already has those.
I was aware of the high-temp ZT series and I have that data sheet in my 'collection'.
I didn't know that was something new.

There are also spec sheets for Jamicon fans in there.
I like to know the CFM before I buy a fan.
The physical dimensions don't really tell you anything about air flow.

BTW:
They don't have a huge selection but this is a often a good place for decent replacement fans at low prices. (And misc other parts.)
http://www.allelectronics.com/cgi-bi.../220/Fans.html

.

[Big Pope]

Re: Found Teapo Specs - For Replacement Reference.

That site also has the datasheets for most Rubycon Aluminium Electrolytics (didn't notice MBZ there, but there's the interesting 125c series ZT):

http://www.cbs.it/rubyconalum.htm

Download them before they vanish due to obsolescence...

It is a good choice if the size of installation is available.

[starfury1]

Re: Found Teapo Specs - For Replacement Reference.

Ok from what I understand I'll seconded Gonzo on that

When you look at the specs that are required from the CPU VRM circuit ...they are quite demanding very high current in short bust of time so the ESR of the cap becomes quite critical here

So if I understand the "Effects" of ESR correctly,
take a 1 Ohm resistor pull 200 amps through it you get a voltage drop of 200 volts across it
then get the power dissipated in watts 200A X 200V = 400 Watts
(how much and how deduce what is expended as true thermal heat I am not sure)

But basically thats a lot of heat

Now with the cap ESR the resistance is probably more like 0.01 Ohm
but you get the idea.

Again if I understand this correctly

Now the ripple is generated by the amount of ESR the caps has acting as a resistor causing a voltage drop to be developed across it by the demands of current being drawn from the capacitor
(these current demands are of course dynamic, changing with time)

So we end up with the capacitor acting as a resistor generating heat.
The higher the ESR the more heat (and ripple) you will get

I know this is not strictly speaking true and I haven't mentioned phase relationships of current and voltage or frequency, capacitor construction etc.

The Ideal cap would have zero ESR therefore no voltage could be developed across it resulting in no ripple and no heat.....the perfect cap in ESR terms at lest

I think this coveys the net result of what happens as an idea and thats all I am trying to do here.
Put it in more lay terms then true tech terms, as a somewhat DC model

feel free to correct me if I got wrong

PSU although they do have high current demands placed on them would not be required to dump as much so quickly therefore the stresses are less with regard to ESR
and the heat generated by the cap from this

So the caps in a PSu don't get the same brutal workout.
although you should use Good Quality Low ESR for both to maximize life expectancy.

Heat would of course factor into this.

I don't know for sure but Guess at around the 2000 period the Teapo brand were subject to the lousy half baked electrolytic formula which really badly hurt the reputation of teapo caps (not that from what I gather it was that great anyway.... along with others cap suppliers too)

Seems A lot of the pricey PSU's do Use them?
go figure.... if they are really bad caps why would you risk your reputation on them
(teapo is a bit like a dirty word round these forums and understandably so from the past issues)

One thing That I am wondering about (and haven't really looked into) is with all the new Graphics cards, as in the supersonic sub light speed ones that will do zillion frames a second at flee's @#$% resolution and that will kill a fat wallet in under a microsecond.

What current demands to they place on Power supplies? and whats best for them single dual etc rails.

Anyway that really going off topic here and probably best to start another thread on that (if haven't got one on it)

Just my thoughts on it

Cheers

[starfury1]

Re: Found Teapo Specs - For Replacement Reference.

I would think its more the bean counter's spec caps rather then the designers.....

on the VRM caps, only use good quality (ultra) low ESR (high current) types that are up to the task....if you want to maximize trouble free operation from this circuit, Ditto Output caps on PSU's

Thanks for the link thought, lots of pdf's to be Grabbed
(and I did )

Cheers

[PCBONEZ]

Re: Found Teapo Specs - For Replacement Reference.

In some cases more is better. - In other cases more is irrelevant.

....................

The Teapos I'm dealing with don't have anything to do with the VRM's.

In fact I haven't even SEEN Teapos used in VRM's on anything newer than an i810 or a BX board and probably much older than that. Maybe even 486?
- So, talking about Teapos and VRM's in the same sentence seems kind of invalid to me. If Teapos don't show up in VRM's then why do they need replaced with VRM quality caps?
- Granted, I don't do a huge variety of boards. Maybe YOU'VE seen Teapos in VRM's on recent boards. I haven't.

Teapos are usually scattered around the board and in those positions where it's doubtful that you NEED anything more than an a Pani-FC or NI-PW.
Those positions are usually just power to the PCI slots, sound, IDE, USB, and so forth and those frequencies haven't gotten to the point where you NEED high end caps.
- Crap caps there get replaced for reliability and life span, not performance. Exceeding what's needed there results in zero performance gain.

That's like watering you lawn with a 2 inch fire hose that's attached to the same old 1/2 inch spigot. - Boy! You have a nice hose there, too bad it doesn't make the water go faster. - The higher capacity hose is irrelevant...

Same-same with replacing a <1000 ripple cap with a 1250+ ripple cap.

Noted: sometimes 1250+ ripple caps are cheaper than 1000 ripple caps.
So, DUH, get the better ones.

When the CPU, memory, and video frequencies went up (and voltages down) that necessitated higher quality caps in THEIR VRM circuits but that doesn't mean the rest of the board needs better caps too. The rest of the board is about the same as what it used to be.

I do go for higher ripple ratings if it's anywhere near the chipset, memory slots, or a video card slot. That just makes sense as a "just in case" because I don't trace the circuits out with a meter.

~~

I think in most cases the designers get the values right for the best performance.

Unfortunately when the production department gets it they load the design up with crap short life parts that happen to meet the specs the designer gave them.

- The crap parts either fail or drift out of THEIR designed operating range.
- The board performs like crap or fails.

But that doesn't mean you NEED to upgrade.
- All you need to do is return it to the original design specs to have a reliable board.

Since like no one hands out schematics for their circuits, to know the original design specs you have to look at the specs for what you are removing.
- I don't see many people doing that.

I see people janking the crap caps and guessing about what should go back in or failing by default to "get the best there is".
- That just doesn't make sense to me.

I don't need a 300 mph tire for my wheel barrel because I don't know what I took off.
- It's the same thought process.

Boy, I must be in a bad mood.....

.

[gonzo0815]

Re: Found Teapo Specs - For Replacement Reference.

Your right, crap caps are crap caps and i would like see them nowhere.
But there are many socket A boards where Teapos where used along with G-Luxon and others as VRM input caps.
Sure, manufacturer switched to other caps like Ost, and recently to solid polymers.
But if you crawl through this forum, you can`t neglect the fact, that Teapos do usually not fail in PSu`s like in boards.
For stable operation, on most boards you don`t need to use the most current low esr caps, but it would not hurt in most cases to do so.
And it depends on the particular circuit, if not cutting edge caps are necessary or not.
I sure can read data sheets and can determine what the circuit may need ornot, most others i think can`t.
On the other side, many people are on the mission to improve the boards for whatever reasons.
And i think the conclusion of the accumulated knowledge here is, that it is possibly to improve things by using better caps.

[Sparky]

Re: Found Teapo Specs - For Replacement Reference.

Side note on the G-luxon caps that keep coming up, I just recently sold an ECS K7S5A rev 1 that had those all over it and none of them was bad believe it or not. Well, one went poof but that was my fault

[PCBONEZ]

Re: Found Teapo Specs - For Replacement Reference.

But there are many socket A boards where Teapos where used along with G-Luxon and others as VRM input caps.

That may be why I haven't seen Teapos near VRM's.
I haven't had an AMD board since they were in slots.
(No particular reason, just how it played out. I woke up one morning and it was raining P3 boards.)

And to starfury1:

200A X 200V = 40,000 Watts , not 400

ESR = Equivalent Series Resistance.
It's accounting for Frequency in a DC circuit by trying to make the AC 'resistance' (properly 'reactance') equivalent to what the resistance would be if it was true pure DC.

[I'm using unrealistic numbers for clarity.)
Lets say you have a DC circuit that is 5 volts.
There is no perfect DC circuit so the actual voltage varies between 4.5 and 5.5 volts.
- That would be a 1 volt "Ripple Voltage"

Now, like V = I X R .. .. .. Ripple Voltage = Current X ESR

- Because there is a cap in the circuit and the Ripple component of the circuit's voltage is actually AC and not DC, Ripple Voltage and ESR are going to change with the Ripple *Frequency*.

To compare one cap to another they needed a standard frequency to measure all caps at. They chose 100kHz.

So, if you think about it. [ Ripple Voltage = Current X ESR ]
- Lower ESR => Lower Ripple Voltage.

NOW!
(Based on electron flow theory not the hole flow theory.)

In order for a cap to keep a voltage constant in a circuit it has to alternately supply and remove electrons (current) from the circuit. [When the voltage is too high it pulls some 'trons out. When the voltage is too low it pushes some 'trons in.]

When it comes to Ripple Voltage it has to do this very fast at relatively high frequencies.

That is what the "Ripple Current" rating of a capacitor is referring to.
- At a high frequency,, how much current can the cap supply and remove.

It might be more accurate if they called it "Ripple Current Capacity" but given that it has nothing to do with the capacitance rating that might confuse people.

- Ripple Current is the high frequency 'Bleed off' and 'make-up' current and not the main circuit's current flow which will be effectively a DC value.
- In the formula [Ripple Voltage = Current X ESR ] the 'current' is the main circuit's current flow. Not the Ripple Current.

Anyway, the higher the Ripple Current rating is the better the cap is at reducing the Ripple Voltage in the circuit.
- It could be thought of as: the higher the Ripple Current Rating, the faster the cap can respond to voltage changes. (Or, the higher the Ripple Current, the more 'trons the cap can 'move' in X amount of time.)

Ripple Current is also standardized at 100kHz.

.....

I learn so much when I explain things to other people....

.....

[starfury1]

Re: Found Teapo Specs - For Replacement Reference.

Yeah Humm dont think VRM and teopo was mention in the same breath?


I probably should have said "Generally speaking" and "in my opinion" in a few parts there.

As to you explanation of ESR which is much more in depth, and every little bit helps

I was really trying not to get too deep into it and was more getting at the causes of heating effects of ESR and way the caps dies quicker, (plus why you should use low esr caps)

Total truth it isn't and somewhat bent version of reality from a somewhat DC (moment in time) it is but if it gets the idea across thats what I am aiming at

Thanks for the Correction, was typing on the fly and didn't pick it up cheers
( I should have thought....)

Most people that find there way here usually have little or no electronics knowledge
Most problems that they have and cause them to find there way here are usually caused by popped outputs caps in PSU's or popped VRM caps.
Most of this is from the use of crap caps in these positions.

You are quit correct in saying that some parts of an MB or PSU you dont need super you beaut extra low ERS caps in. horses for courses as the saying goes.


I find the use of Teapo on more expensive(silverstone for example if you see some of jonny gurus reviews) supplies somewhat interesting and yeah they do tend to last longer in a better built supply....here is possibly were the right series has been used for the demands of the circuit, but I dont know this for sure and only guessing this is the case.

I do agree that if you don't know what was there before use the same type speced part if you can get it.
(preferably from a quality cap maker, I have even said this in other posts)

The other side is you will find posts around here on people asking if they can replace with GP type caps in VRM, sure it may well work but wont last for all that long...so thats why I try and stress use "quality low ESR" for maximum trouble free operation of the circuit (barring other issues of course)

Mostly you will find people will recommend a better spec part for the job, if whom asks the question posts what it was or a picture....
(sometimes its just down too how available a part might be in that part of the world)

But yes I agree, it does not mean that all caps on every part of a MB or PSU need to be of this caliber

Cheers

[PCBONEZ]

Re: Found Teapo Specs - For Replacement Reference.

I think we are in agreement with a slight bit of dialect confusion.

I was not saying use cheap parts as replacements.
- ALWAYS buy quality parts.

But you don't need to exceed the (advertized) ratings of the original parts to get a solid reliable board when you are done.
- Just need to match them.
(So keeping specs sheets for crap parts handy makes sense. - Hence my post to Teapo specs.)

It's not the specifications and ratings that make crap parts crap.
It's the fact that they don't STAY at their specifications for very long.

Quality parts will STAY at the advertized ratings for a very long time.
Crap parts will only stay at the advertized rating for a short time (or fail.)

.

[starfury1]

Re: Found Teapo Specs - For Replacement Reference.

Agree

Having reference to what a parts specs is always handy regardless

The site its self was a good find

it listed each series on one page with the pdf

pitty there wasn't a few more manufactures there thought
(ever got hunting for pdf's you'll see why the site has a good idea)

Err yeah folks a few typo's in my last post Doh!

Cheers

[davmax]

Re: Found Teapo Specs - For Replacement Reference.

PCBonez


I do not understand your theory statement.

Quote:ESR = Equivalent Series Resistance.
It's accounting for Frequency in a DC circuit by trying to make the AC 'resistance' (properly 'reactance') equivalent to what the resistance would be if it was true pure DC.

The fact of ERS is that it is seperate from capacitive reactance being part of the pure capactitance and also seperate from the inductive reactance (ESL). ESR represents the the resistive component in which loss/heat occurs. There being no loss in either of the two reactances.

You have me puzzled.