Fairchild 8 pin IC part identification please - PG8AB - NDS - 99?

[tkamiya]

I am in process of repairing Agilent DSO 5034A.   Can anyone identify the damaged SMD 8 pin IC?  It is probably some kind of power regulator.  Google search and search around my favorite supplier yielded nothing. 

Maker is Fairchild
PG8AB
NDS
99 (something - 3rd letter is burned and illegible)

Board number is 54574-26402

[asis]

Hi,
Judging by the wiring, the complementary MOSFET. P-cannel destroyed.

https://datasheetspdf.com/datasheet/search.php?sWord=NDS99
https://datasheetspdf.com/pdf-file/474533/Fairchild/NDS9952A/1

[tkamiya]

Thank you!  FET is my weak spot....

So you are assuming the trace that goes under will go to pin 5 and 6, right?  I'll take it off the board and verify that. 

Supposedly, this will supply another regulator, so I'm thinking this is a safety cut-off of some sort.  I wonder why it burned crisp like this.  It's more like explosive deconstruction.

[asis]

This is typical of MOSFET's.
I am even sure that N-ch is serviceable.
You don't rush to replace after dismantling.
You need to look for the reason why this happened.
It is better to do this with a schematic diagram or sketch a fragment.
Look for the reason somewhere in the load.

Vladimir.

[tkamiya]

This device is between power source and a 3 terminal 3.3V regulator.  So it's somewhere there that is faulty causing over-current.  I saw some crude repair attempt, so human error is also a possible factor.  As this is more modern Agilent, circuit diagram is not obtainable.  I am thinking it's something like this:

https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/first11/part9/page5.html

I'm pretty sure it's a cut-off device in case of failure.  In fact, the defect I'm facing is, probe power not being available.  Error message shows it detected an over-current and it cut off the power.  Kind of fits the damage and possible cause.

Thank you for your expert knowledge and assistance!

[asis]

Happy to help.
However, if you showed a photo from the other side of the PCB, I would be more confident.
Good luck.

[tkamiya]

Hopefully, this will be helpful.

I went ahead and removed the device.  Upper half of the device basically fell apart!

Unfortunately, through one of the pad is damaged and the all important via seems to be ok, except for one.  That pad, pin #3, the missing part appears to have a connection to inner layers, and potentially a partial through-hole.

To help you look, I put one black dot above Q3400 to indicate the same via. 

[asis]

It is clear that this is a multilayer PCB.
You can give a photo of the whole PCB. It's easier for me to understand the circuit.
The difficulty is that the designations U3402 and U3401 (?) Have been erased.
It may turn out that the Q3400 is Dual P-ch or Dual N-ch.
{weekend}

[tkamiya]

They are not erased.  Just very faint and hard to reproduce.

U3401 reads 4373 FEUA
U3402 reads 311 E2642

It's obvious from looking at it, the whole circuit takes power source input and pass it to another layer of voltage regulation.  I think it might be a switch. 

I can send a whole board but you won't be able to read from it.  Let me see what I can do so you don't waste your time trying to figure out which via goes where.  Even having the board in hand, it's very difficult.  In fact, I can't make out some of them.

[tkamiya]

I think this might help. 

A large 6 hole pad right above "Q3400" silkscreen is directly connected positive power input.

Opposite of it, there are what appear to be filters that will ultimately connect to U3400 on positive input side.  This is a 3.3V low drop out 3 pin regulator.

Oddly enough, none of LT1086 nor Q3400, other than the 6 hole pad are directly connected to ground pins on power connector.  It seem to be floating.

[asis]

You are not mistaken in indexing U3401 = ?; U3402 =?

https://datasheetspdf.com/pdf-file/1208273/MaximIntegrated/MAX4373/1
https://datasheetspdf.com/pdf-file/570310/LinearTechnology/LM311/1

[tkamiya]

It's really hard to read even under 40x magnifications.  As far as I can tell, my readings are correct.

4373 makes sense because this is supposed to be a power source for powered probes.  The trouble I'm having is, it is reading overload.  That means there has to be something that's reading current or voltage.

[asis]

U3402 Is MAX4373?

[tkamiya]

Kind of make sense function wise.  I'm having difficulty reading patterns with both sides and through holes involved.  Marking wise, there is no assurance.  I wish I can see inner traces.

[tkamiya]

I wonder what would happen if I power this up with ammeter where mosfet switch is missing.  At least I know which one is the input and which one is the output.  There is a voltage regulator before it goes anywhere.  I'm fairly curtain this is a cut-off for over-current in the line. 

One big question mark is, what is going on with the ground.  Neither Q3400 nor U3400 has direct connection to ground. 

[asis]

You need to call everything back with an ohmmeter.

"It is better to do this with a schematic diagram or sketch a fragment.
Look for the reason somewhere in the load."

[tkamiya]

I took some detailed look and decided it can't be anything else but a complimentary pair.  You were right.  It IS a complimentary pair MOSFET.

It works now!  Thank you for all of your help.

One question though.  You were very quick to find out the part number with only partial information.  How did you do that?  Without your kind help, I'd be lost.

[asis]

Hi,
His Majesty! - experience.
I'm already 66 ...
 

[tkamiya]

I'm 54 but I guess that's not enough experience....

I'm totally new to SMD parts and their abbreviated naming convention.  Thank you very much to you and your EXPERIENCE.  The scope passed self-cal and it is now good to go.

[asis]

You cannot grasp the immensity, but there is no limit to perfection.
Good luck in everything!
It was nice to talk to.
Vladimir.