Toshiba M35X clicking noise from mb

[paul123]

The motherboard has a clicking noise coming from one of two transformers which are situated next to a maxim1902eai chip a Power-Supply
Controller.

Any ideas would be most welcome - thanks.

[paul123]

Re: Toshiba M35X clicking noise from mb

Photo of where the noise is coming from

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

Get 2 or 3 foot piece of windshield washer tube. The soft rubber kind.

Place one end in ear.
Place other end on components until you find the noisy one.
.

[paul123]

Re: Toshiba M35X clicking noise from mb

Thats a good idea - I have another laptop with the same clicking but have never been able to isolate the sucker.

Forgot to mention the laptop in question does not power on at all, is it likely the clicking transformer to be the cause or is it the effect of something else thats gone bad?

Whats the best way of removing these sm transfomers

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

The clicking noise is a typical symptom of a power supply in protection mode. One of the mosfets forming the synchronous buck rectifier is likely shorted.

Those are inductors not transformers. The best way to remove them is with a hot air gun. At the way they're packed, it's not likely that you'll have success via any other method anyway.

[paul123]

Re: Toshiba M35X clicking noise from mb

Yeah thanks TH3 I felt it in my water it was something else causeing the faint clicks.

On the reverse side of the board are 4x 8 pin dil chips two are branded

4810
BO and a delta sign
W48A

the other two

4800B
CD and a delta sign
W49B

Seven legs of each of the 4 devices are grounded.

Will they be the mosfets if so does anyone have the data sheets for them I can't find them on google

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

Seven legs of each of the 4 devices are grounded.

That's impossible. Sure sounds like a short to me. Remove CPU to be sure (CPU can cause false readings on vregs), but i believe they're toast. They're most likely Alpha & Omega devices (AO4810, AO4800), check this link and this link.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

The clicking noise is a typical symptom of a power supply in protection mode. One of the mosfets forming the synchronous buck rectifier is likely shorted.

Those are inductors not transformers. The best way to remove them is with a hot air gun. At the way they're packed, it's not likely that you'll have success via any other method anyway.

It's a laptop.
The PSU is a power brick.

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

It's a laptop.
The PSU is a power brick.

And the VRM of the CPU, isn't that a power supply too?

[paul123]

Re: Toshiba M35X clicking noise from mb

That's impossible. Sure sounds like a short to me. Remove CPU to be sure (CPU can cause false readings on vregs), but i believe they're toast. They're most likely Alpha & Omega devices (AO4810, AO4800), check this link and this link.

I agree its impossible, after taking another look. Its not the cpu removed that and it still clicks.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

And the VRM of the CPU, isn't that a power supply too?

Yes but there is no -rectifier- involved.
It's all PWM
.
Might wanna look here.

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

Yes but there is no -rectifier- involved.

Oh yes there is. The diodes are simply replaced with mosfets. They still act as a rectifier - you need a rectifier to create DC.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

Um,,, no...
You already have DC.
The MOSFETs are not acting as rectifiers.
Rectifiers work by blocking the reverse voltage in AC. Don't have any here.
-
If your point can be proven then show me the 'rectifier' in a VRM.
-- And prove it's a rectifier.
[It's not.]
.

.

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

You already have DC but you need to step down that DC to a level that the CPU takes. You do so by PWM'ing that DC. PWM is AC.

A regular buck converter uses a mosfet, which is a switching element, an inductor, which is the storage element, and a diode, which "catches" the reverse-going spike of the inductor when the pulse is switched off, and converts it back to DC. A synchronous buck converter such as a CPU VRM replaces that diode with another mosfet, for better efficiency.

A mosfet in "off" state has a very high impedance. A mosfet in "on" state has a very low impedance. For all switching intents and purposes, a mosfet is a low voltage drop, high current diode. But as opposed to a diode it needs to be driven externally.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

PWM is AC.

No, it's not.
It's Pulsed DC.
To be AC you would have to have -current- in the reverse direction.
In PWM current starts and stops but it does not reverse.
It is NOT AC, and therefore you aren't talking about a rectifier.
.
The MOSFETs in VRMs is act as a -switches- to a DC source.
- Same as the switch on a flashlight.
[Which also isn't a rectifier outside of this theory of yours..]
.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

A regular buck converter uses a mosfet, which is a switching element, an inductor, which is the storage element, and a diode, which "catches" the reverse-going spike of the inductor when the pulse is switched off, and converts it back to DC.

If that were how it works then removing the inductor all together would improve the circuits stability.
- So that's obviously NOT how it works.

The inductor's field doesn't have -time- to collapse before the next pulse comes in.
There is no 'reverse going spike'. [Until you turn the whole thing off.]
.

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

The inductor's field doesn't have -time- to collapse before the next pulse comes in.
There is no 'reverse going spike'. [Until you turn the whole thing off.]
.

You are wrong. The inductor is the primary source of storage in any switchmode power application. There are two discrete modes of operation for a buck, boost or flyback converter - discontinuous mode where the current in the inductor DOES go to zero before the next cycle, and continuous mode where it does not. High density converters favor continuous mode because it needs smaller inductors and provides less ripple (there is a minimum inductance required for a converter to run in discontinuous mode).

But this doesn't change the operation of the inductor, it always, always produces a voltage spike when a current is applied and then removed suddenly. The diode/mosfet is there to make that spike do something useful. You may want to read some papers on inductors again...

I think what you are talking about is simply a technicality of language. Why does everybody else call it a synchronous RECTIFIER then?

Wikipedia
One PDF
Another PDF

So, these guys must be wrong too, ain't it so?

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

Wikipedia

Specifically referring to VRs with an AC input [Which a VRM and this circuit this circuit are NOT.]
>>> "Using active rectification to implement AC/DC conversion ...... ...... ...... which forces the current waveform of the AC source to follow the voltage waveform, eliminating reactive currents and allowing the total system to achieve greater efficiency."

One PDF

Notice how every example circuit has a transformer and VRMs and this motherboard DON'T...
This example does not fit the application.

Another PDF

This one clearly says what I just did. - The coil's field never crashes.
-
If the field did collapse the voltage between the high and low MOSFETS would go negative.
It is therefor your contention [by your stated theory] that Vcore goes to a negative voltage.
-
[Has anyone besides Th3_uN1Qu3 seen Vcore swing negative on their O'scope?]

So, these guys must be wrong too, ain't it so?

They aren't wrong about how that works.
They are wrong about what they call it.
Either way you are misapplying their information.
A VRM is a DC-DC Converter and there is no Rectification involved at all.
It indeed can be called an SMPS but there are many kinds and VRMs are a type with no AC and no rectifier.

Since you like wiki so much lets look at the definition of RECTIFIER.
http://en.wikipedia.org/wiki/Rectifier
>>> "A rectifier is an electrical device that converts alternating current (AC), which periodically reverses direction, to direct current (DC), which is in only one direction, a process known as rectification."
Clearly a 'synchronous RECTIFIER' with a DC input is not actually a RECTIFIER at all because it does not RECTIFY AC to DC.
It's a DC-DC converter that has been misnamed.
Smoothing Pulsed DC is NOT rectification.
.

[PCBONEZ]

Re: Toshiba M35X clicking noise from mb

Another wiki for you about "Active rectification"
http://en.wikipedia.org/wiki/Active_rectification
>>> "The control circuitry for active rectification usually uses comparators to sense the voltage of the input AC ..."

Lacking an AC input the device is a Synchronous Converter.
Not a Rectifier.
.

[Th3_uN1Qu3]

Re: Toshiba M35X clicking noise from mb

VCore cannot go negative. The mosfet acting as a diode catches that spike. But anyway, in the buck, that spike occurs at the inductor side that is on the switching node, not on the output. The same inductor is used both as energy storage AND DC filter. You have a diode any way you look at it, in the same way a switched inductor does not, and most importantly, cannot output DC by itself. It outputs an AC waveform, which, when measured across that inductor, does go negative. The voltage at the switch also goes negative without the diode, or second switch in a synchronous rectifier. That's why it's called a rectifier. A rectifier makes DC at whatever point it is connected at - and in this case the point is at the switching node. The voltage at the output has nothing to do with the diode (or low-side mosfet) itself because it is not directly connected to that diode.

In a buck converter the switched inductor is in series with the load AND the input, and ground is common to both input and output, so the load voltage cannot be negative in respect to ground (and also cannot be positive in respect to the input), but this changes in the buck-boost for example.

You have once again proven that you do not understand what an inductor does. Last time it was about the common-cored toroidal inductor in group regulated PC power supplies. At least that time you were more polite in asking us what its function is. I remember that kaboom also provided valuable information in that topic. I'll try to look for it. In the meantime, let's not clutter this guy's topic anymore, okay?