Is There Anyone Who Knows Which Multimeter Can Test the Reballed Maiboard?

[Devian]

Hello, I am in electronic PCBA repair, now i want a Multimeter which can test reballed main boards, who can tell me which is the most suitable one: can meet my need but not that expensive.

[joelby]

What exactly do you want to test?

Normally to check BGA soldering, you would use an x-ray machine, not a multimeter.

[Devian]

I reball Xbox 360 mainboard, for example, the GPU,after reballed it, I need to test the electronic effect, that is why i look for a multimeter,any one with recomendations?

[joelby]

Can you describe what you mean by "electronic effect"?

Honestly, pretty much any multimeter will be equally suitable for this task.

[IanB]

If you tried to replace a GPU on an Xbox 360 with that apparent skill set, then you broke it. No need to test it, it's broken.

[DaveW]

I reball Xbox 360 mainboard, for example, the GPU,after reballed it, I need to test the electronic effect, that is why i look for a multimeter,any one with recomendations?

Do you mean reflowed; ie. got a hot air gun and heated up the GPU? In which case, testing with a multimeter is pointless, just reassemble it and try it. If it doesn't work, try it again but heating it for a bit longer.
There aren't methods for detecting a BGA connection that is high impedance without using high end optical inspection, TDR or x-ray, and none of these are possible at a typical hobbyist level.

[Devian]

Thank you guys, who can tell me how can I know the Xbox 360 GPU is well reballed?

[Devian]

There is any way to prove my reballed gpu work well? If does not work, it means my GPU already broke? And no ways to make up except only to buy a new one?

[Bored@Work]

Get it X-rayed. And get a clue.

[Neilm]

If you mean you want to test for shorts using continuity testing, remember the output voltage of the multimeter might be enough to damage the newly soldered GPU. For instance, to meet some of the wiring standards in the UK, they have to output at least 4V.

Yours

Neil

[IanB]

If you mean you want to test for shorts using continuity testing, remember the output voltage of the multimeter might be enough to damage the newly soldered GPU. For instance, to meet some of the wiring standards in the UK, they have to output at least 4V.

That sounds like a disaster. For electronics use I think I want my continuity and ohms measurements to use 0.5 V or less to reduce the risk of damage to other components during in-circuit tests. I really would not want my DMM outputting large voltages like that with the possible exception of the diode test function.

[codeboy2k]

If you mean you want to test for shorts using continuity testing, remember the output voltage of the multimeter might be enough to damage the newly soldered GPU. For instance, to meet some of the wiring standards in the UK, they have to output at least 4V.

That sounds like a disaster. For electronics use I think I want my continuity and ohms measurements to use 0.5 V or less to reduce the risk of damage to other components during in-circuit tests. I really would not want my DMM outputting large voltages like that with the possible exception of the diode test function.

The continuity tester of the DMM should definitely be well below 0.5V and I'd even prefer to see it below 0.2V. This is to ensure against forward biasing any PN junctions in circuit.

[EEVblog]

There is any way to prove my reballed gpu work well? If does not work, it means my GPU already broke? And no ways to make up except only to buy a new one?

No, there isn't a way to test it.
Apart from switching it on and testing it after you solder it. It'll either work or it won't. If it doesn't work, you won't know if that's because it's a bad chip, or bad soldering, or both.

Dave.

[Neilm]

If you mean you want to test for shorts using continuity testing, remember the output voltage of the multimeter might be enough to damage the newly soldered GPU. For instance, to meet some of the wiring standards in the UK, they have to output at least 4V.

That sounds like a disaster. For electronics use I think I want my continuity and ohms measurements to use 0.5 V or less to reduce the risk of damage to other components during in-circuit tests. I really would not want my DMM outputting large voltages like that with the possible exception of the diode test function.

Exactly my point - I don't know what most DMMs do but if you get one that does output 4V or more (BS7671 gives 4-24V) you will blow things up. I think I will measure what my ones at work give.

Neil

[alm]

Open circuit or for very high resistances, most meters produce a voltage of 5V or so. For lower resistances some meters are specified to be below 0.2V.

I don't really understand the 4V requirement, is this about puncturing oxides? How about just requiring people to use a megger if insulation voltage matters? Otherwise this should be a clear distinction between electronics (below diode drop) and electricians (>4V) meters. Probably a separate mode, like the Fluke Lo Ohms feature.

[hacklordsniper]

Just try it,  i have seen people heat those motherboards in mothers stoves just after she baked cookies and it worked. It would be good possibly to limit the power supply current from burning something if there is a short, possibly powering only the motherboard from lab power supply and see what happens

[Neilm]

I don't really understand the 4V requirement, is this about puncturing oxides? How about just requiring people to use a megger if insulation voltage matters? Otherwise this should be a clear distinction between electronics (below diode drop) and electricians (>4V) meters. Probably a separate mode, like the Fluke Lo Ohms feature.

BS7671 is the standard for building wiring in the UK so that is where the 4V is specified. Some meters pick that as a requirement.

I measured a couple of multi-meters at work - one of them is reasonably cheap and you had to select the range. In the top range (>200k) the output voltage was 5V until you had about 5k on in. More than enough to fry a chip or two.

Neil

[alm]

BS7671 is the standard for building wiring in the UK so that is where the 4V is specified. Some meters pick that as a requirement.

Why this 4V minimum? What trouble are they trying to solve?

[Neilm]

BS7671 is the standard for building wiring in the UK so that is where the 4V is specified. Some meters pick that as a requirement.

Why this 4V minimum? What trouble are they trying to solve?

I think it is due to other requirements in BS7671 - for instance limits on the resistances. The instruments have to output 200mA into a short circuit so you will probably find that there is a 20 ohm max resistance limit somewhere in the document.

Neil

[alm]

The minimum short circuit current doesn't make sense to me either, but I'm not familiar with the standard, so I digress. 4V 200mA sounds just fine for some diodes like ESD protection diodes. Better keep those meters far away from any electronics.

[Devian]

There is any way to prove my reballed gpu work well? If does not work, it means my GPU already broke? And no ways to make up except only to buy a new one?

No, there isn't a way to test it.
Apart from switching it on and testing it after you solder it. It'll either work or it won't. If it doesn't work, you won't know if that's because it's a bad chip, or bad soldering, or both.

Dave.

Hi. your opinion really help me much, but if there is an easy way to test the ship before reballed or before soldering? How to test a X360 gpu reballed chip to make sure it is working before soldering ? Many thanks

[Devian]

Open circuit or for very high resistances, most meters produce a voltage of 5V or so. For lower resistances some meters are specified to be below 0.2V.

I don't really understand the 4V requirement, is this about puncturing oxides? How about just requiring people to use a megger if insulation voltage matters? Otherwise this should be a clear distinction between electronics (below diode drop) and electricians (>4V) meters. Probably a separate mode, like the Fluke Lo Ohms feature.

Yes I have the same question in my mind.

[joelby]

It is basically impossible to test the chip without soldering it onto the board and seeing what happens when you turn it on.

Maybe you could do it if you had chip-specific testing rig with a BGA socket, but only the manufacturer would have something like that, and only they would know how to accurately test it. With a multimeter you could compare the impedance between every ball and every other ball (if you could guarantee that your test didn't destroy the chip), and see if this is different to another chip from the same batch, but on a large chip this represents hundreds of thousands of individual tests and you won't have any way to know if a difference is meaningful or not.

Have you considered taking it to an XBox repair place?

[Devian]

If you mean you want to test for shorts using continuity testing, remember the output voltage of the multimeter might be enough to damage the newly soldered GPU. For instance, to meet some of the wiring standards in the UK, they have to output at least 4V.

That sounds like a disaster. For electronics use I think I want my continuity and ohms measurements to use 0.5 V or less to reduce the risk of damage to other components during in-circuit tests. I really would not want my DMM outputting large voltages like that with the possible exception of the diode test function.

Exactly my point - I don't know what most DMMs do but if you get one that does output 4V or more (BS7671 gives 4-24V) you will blow things up. I think I will measure what my ones at work give.

Neil

Thank you, if so, is there anyway to check the gpu before solering? I just want to know why it does not working after I soldering it? it is me that damaged it during the solering process because of wrong operation ? or the GPU was already damadged when arrived to me?  Thanks,guys

[Devian]

Open circuit or for very high resistances, most meters produce a voltage of 5V or so. For lower resistances some meters are specified to be below 0.2V.

I don't really understand the 4V requirement, is this about puncturing oxides? How about just requiring people to use a megger if insulation voltage matters? Otherwise this should be a clear distinction between electronics (below diode drop) and electricians (>4V) meters. Probably a separate mode, like the Fluke Lo Ohms feature.

Yes I have the same question with you.

[tinhead]

Open circuit or for very high resistances, most meters produce a voltage of 5V or so. For lower resistances some meters are specified to be below 0.2V.

I don't really understand the 4V requirement, is this about puncturing oxides? How about just requiring people to use a megger if insulation voltage matters? Otherwise this should be a clear distinction between electronics (below diode drop) and electricians (>4V) meters. Probably a separate mode, like the Fluke Lo Ohms feature.

this is why i love such cheap meters like MS8910, CR2032 driven can't have any dangerous voltage nor current
(while high R only 1.25V, low R ~ 0.4V). Additionaly the ES51925 chip is really fast while in cont. mode.

[tinhead]

Thank you, if so, is there anyway to check the gpu before solering? I just want to know why it does not working after I soldering it? it is me that damaged it during the solering process because of wrong operation ? or the GPU was already damadged when arrived to me?  Thanks,guys

ehm, sure, you can solder a very expensive BGA socket to a working unit, plug GPU and power on ... so no,
no way to DIY test without proper circuit around.

[alm]

Connecting it to a multimeter which supplies 24V might be a good test, you would be certain that the GPU is dead after that test, no doubt about the condition.

[nrgeek]

Dear God.. I know this is a pretty old thread .. but WTF ! ... wana test the GPU or CPU .. like other have said hook the board up and test ... Ive been a computer tech for over 20 yrs.. Ive never thought gee could I test CPU to make sure it works b4 I put in the motherboard LOL OMG much less turn it on to test RFLOL ..

wana test it lol .. just put the chip on one nipple and hook up a lead to your other nipple .. and plug that into your wall outlet .. you will find out one thing .. electricity hurts when you dont know what your doing .. and maybe one less dumb ass in the world lol

Im a a newbie at building electronics from scratch yes .. but with all the you tube videos on xbox repair .. Ive never once and Ive seen a shit load of them seen anyone mention or try to test a re-ball with a DMM .. hell the oven trick, hair dryer, hot air gun etc trying to fix an xbox is just a temp fix anyway .. if you dont kill it !  Most dont survive 3-6 months before the same thing is happening all over again !

Only way to truly fix, is to buy the proper re-ball tools, stencil and chip clamp, board heater, and hot air or infra red rework system, this too is all over the net .. or just pay some one to do it for you 40 -75 bucks! There are some people that are just to damn stupid to drool or lazy or both lol

Forgive me father for I may have just sinned lol

[joelby]

Turns out that you can just use a multimeter - as long as you're safety conscious.

From The Amp Hour:

[Bored@Work]

And make sure you have a proper manicure done - essential for doing proper measurements.

[RJSC]

Well, most meters in continuity test cannot light or measure a white or blue LED, so, the are well below those 4V.
But even so, aren't the multimeter outputs current limited?
Almost every IC has integrated diodes clamping the voltage to the power rails so, If the current is low (5 mA for instance) it shouldn't damage them.
Do the modern CPUs/GPUs lack the input clamping diodes?

The diodes I'm talking about are like this:


I thought it was standard/good IC design practice to place them on all I/O pins?

[joelby]

They may or may not have clamping diodes, but the real point is that the CPU/GPUs in question have over a thousand pins - what exactly are you going to test by checking them with a multimeter? There are no references that state what readings you should expect. If you have two identical chips, you could compare them against each other, but without an automated test jig, this would take days/weeks and be indescribably tedious.

You might as well just solder it in place and flick the switch, or just buy a new XBox.

[RJSC]

They may or may not have clamping diodes, but the real point is that the CPU/GPUs in question have over a thousand pins - what exactly are you going to test by checking them with a multimeter? There are no references that state what readings you should expect. If you have two identical chips, you could compare them against each other, but without an automated test jig, this would take days/weeks and be indescribably tedious.

You might as well just solder it in place and flick the switch, or just buy a new XBox.

Sure, on a situation like that I'd just power it up to see if it's fine.

It's just that I use to probe for shorted power rails on PC motherboards who fail, from family friends and neighbors to find which one is shorted and then check the capacitors on that bus, but you left me wondering that if I probe a non shorted 1.25V rail going into the north-bridge or a dedicated soldered GPU I might damage them.
So far, on 2 computers, it never happened.

[joelby]

If you're measuring voltage rails using a voltmeter with the CPU/GPU connected, there's almost no chance of damaging it, unless you were to short different voltage rails together, etc.

[RJSC]

If you're measuring voltage rails using a voltmeter with the CPU/GPU connected, there's almost no chance of damaging it, unless you were to short different voltage rails together, etc.

Maybe I was not clear.
When a power rail is shorted the power supply does not switch on (because of the shorted output drawing too much current), so I must probe the rail in continuity test mode with the power switched off. In this case, the multimeter will supply power.

[siliconmix]

i'd buy a new xbox swap all the serial numbers .take your duff one back for a  returns.meh 

[Uncle Vernon]

And make sure you have a proper manicure done - essential for doing proper measurements.

You Americans are so vane!

[Erickben]

multimeter is for dc current, resistance meansurement.

[Dave]

Nope. Also, don't revive old, pointless threads.