Dave brings out the big guns in this attempt.
Can a $12K Flir E60 IR Thermal Camera find the fault?
Once again, a "real-time" attempt to find the short on the Lecroy mainboards 3.3V rail.
Part 1: http://www.youtube.com/watch?v=F-VnbzXjH9I
Flir Cameras: http://www.triosmartcal.com.au/206-thermal-imaging-camera
ULIRVision: http://www.ulirvision.com/product.asp?info_kind=002
Forum Topic: http://www.eevblog.com/forum/blog/eevblog-401-lecroy-9384c-oscilloscope-repair-part-2/'>http://www.eevblog.com/forum/blog/eevblog-401-lecroy-9384c-oscilloscope-repair-part-2/
EEVblog Main Web Site:
http://www.eevblog.com
EEVblog Amazon Store:
http://astore.amazon.com/eevblogstore-20
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/
Can a $12K Flir E60 IR Thermal Camera find the fault?
Once again, a "real-time" attempt to find the short on the Lecroy mainboards 3.3V rail.
Part 1: http://www.youtube.com/watch?v=F-VnbzXjH9I
Flir Cameras: http://www.triosmartcal.com.au/206-thermal-imaging-camera
ULIRVision: http://www.ulirvision.com/product.asp?info_kind=002
Forum Topic: http://www.eevblog.com/forum/blog/eevblog-401-lecroy-9384c-oscilloscope-repair-part-2/'>http://www.eevblog.com/forum/blog/eevblog-401-lecroy-9384c-oscilloscope-repair-part-2/
EEVblog Main Web Site:
http://www.eevblog.com
EEVblog Amazon Store:
http://astore.amazon.com/eevblogstore-20
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/
Hi yes, we're back on the Lroy Oscilloscope. Uh PCB short slrep uh because I had so many people saying that uh, they wanted to see the follow-up to this because I couldn't uh, find the fault within the Um hour that I had last time. um you I'll link in the previous video If you haven't seen it. It is a must watch, You can't This probably won't make much sense to you unless you watch the previous video.
So I thought we'd um start again because we got to the point in the previous video where we found that uh the we we had a you know, a 0.1 ohm or thereabout short across the 3.3v rail on here and we tried to track that down and I explained how the usual culprits are the Um: either the bypass uh Caps or you know, one of the semiconductors or something like that. but of course I sucked off what? I believe were all all of the uh bypass capacitors on the 3.3 volt. Rail and the short was still there and then I um did some quick uh, thermal uh, temperature measurements of the only four devices which seem left on the 3.3 volt rail which are these As6 here because they don't have the memory modules plugged in and they were all at at the same temperature. uh pretty much so one of them didn't really stand out.
you know that was hotter than the others and I couldn't find any other uh place on the board. just you know, using my hand. just a quick look. um you know, a quick feel around on the board, to see if I could find any hotpots but there has to be one here because as I mentioned in the previous video, the 3.3 volt rail is only capable of 6 amps.
uh, the actual power supply and we applied external 3.3 volts and we'll get in 11 amps. so that extra power has to be going somewhere and has to be heating something up. You know I can't do a Scottish accent, but you know you can't defy the laws of physics Captain it's got to be there. So there's extra power now.
So there is definitely a short on this board. Contrary to what a lot of people said, a lot of people um, uh mentioned that oh no, it's probably normal, you know and that power supply is capable of more. and they did all sorts of, uh, thermal, uh, you know, back of the envelope calculations to sort of, uh, convince themselves that that was the case? No I can assure for you. um, there is a short on this board somewhere and that is obvious.
A typical system rail on this board, like a typical 3.3 volt rail will not measure 0.1 Ohms both directions on a multimeter with different multimeters showing no capacitive charge up at all. Um, you know that is just not normal. There is definitely a short on this board and um, thanks to Uh Sam at uh, the Lroy um uh users group. There's a Yahoo users group.
specifically to these Scopes He measured uh the rail and confirmed that it was above 1. ohm. So there is definitely a short there, so let's find it, shall we? It can't be that hard now. Um, well, it can't be that hard.
Well, we'll find out. right? as I said, these heat sinks were all the same temperature. So what I'm going to do is I'm going to use an IR thermometer and we'll go in there and we'll double check those and we'll look around the board with one of those cheap IR thermometers which uh, everyone probably should have in their kit and uh, see if we can find something because there's got to be something on this board on the 3.3 volt rail dissipating that extra power now, uh, quite. There's a few people who said, um, yeah, it might be, um, some sort of, uh, latch up to another rail cuz I'm not powering the other rails and stuff like that. Well, yeah, that could be the case, but that doesn't explain the 0.1 ohms. And when you find a a gotcha like that, you should track it down. That's why. um, in the previous video, that's entirely what I focused on.
because if you can't find that 0.1 ohm short, nothing else matters. so everything will be a red hering in terms of drawing the extra power. Now, it may be the case that there is some of that extra power dissipated in the other rails cuz it's going through protection, diodes and stuff like that cuz we haven't powered up the extra rails, etc etc. and that could be the case.
but there is definitely some extra power being dissipated there. So I expect there to be something else on this board that's heating up. Now let's just do a quick recap here: I got my external High current Supply at 3.3 Vols says 3.2 but it's near enough to 3.3 Um, and it's drawing about 11 amps. which yes, if you take the 0.11 ohms short, it should be drawing a lot more than that.
And people said uhuh, there's you know there's something wrong there but no, it is a short that is very common and expected. in something like this. you would expect the resistance to heat up and have a temp code like that. It is very, very common.
and then the current drop. Just because we measure uh 0.11 ohms at, you know, the milliamp or two that this multimeter is going to measure that current at doesn't mean that resistance is going to stay the same at 10 or 11 amps. it's not. So I would expect it not to match up with OHS law in that respect because the Uh short is going to be uh, nonlinear and in some respect, um and Andor it's going to have a Uh Tempco and it's going to heat up so that is perfectly normal, so don't worry about that.
now. A few people wanted me to measure the Uh voltage over here as well. I Don't think we'll get too much drop in those leads. maybe a few hundred molts.
Let's have a look there you go. 3.05 volts. So really, you know it's not a huge uh, it's not a huge drop, It's it's not. It's not a huge drop there at all now.
I've only had this thing on for a minute or two, but let's uh, have a look at the Uh temperatures of these heat sinks again. I Got my little fluke. uh 59 Mini IR Thermometer here and uh, that will tell us the Uh temperature on those heat sinks. Um now I think I was getting around about 70 or something before after they, uh, heated up for quite a few minutes and uh, yeah. I Have no idea if this is regular operating, uh temperature? not whether or not the clocks are working, whether or not you need the other power rails, uh, powered up, whether or not you need the processor board a It doesn't matter. Okay, we're looking for a temperature differential between these heat sinks, so let's measure that. Uh, you know it's going to peek at about 60 there. That one's very similar.
around 60. This one here is yeah, it's around 60 as well if I can get it. Yeah, 65. So there's not a huge There we go 63.
You know it's quite hard to get it with these things, but you know I would expect one of those to stand out. So really, there's no temperature differences between those. Forget it. Completely rule it out.
Now let's probe around other parts of the board and see if we can find it. Let's take the bench for example, as a reference: it's 20. you know, 24 25 degrees for example 27 I mean we're we're going to get some, uh, heat coming off the heat sinks and stuff like that. So as I said before, I don't think any part of all this stuff is powered up at all.
So really I wouldn't expect to find and these once I think just we're getting some residual heat from those heat sinks there. But these as6 of course aren't powered up. Now there's one thing I Wanted to check because we really have to go back to the the beginning. Here, we have to go back and consider that it could be a physical short now in the first video I Ruled that out because that is usually the least likely scenario for a product like this that used to work just fine.
it was out in the field, worked for years. you know they? they don't just magically get a short in the power plane or the or a connector or something like that. But I want to go back and start at the connector here and have a look Now these. um, there's no surprises for guessing that these alligator clips here get get reasonably warm because there's contact resistance in there.
So this thing's working at 11 amps so they get quite warm. but um, maybe you know if there's a, uh, short near it or under the connector or something like that. As unlikely as it seems. Um, we have to consider that because this unit has, actually, um, had some physical shock or something to it.
If you've seen the tear down of this, you'll know that the case has got a big crack in it and, uh, all sorts of of things. so it's most likely been dropped or something like that. So maybe we've got some stress transferred onto the connector. or possibly, as a couple of people mentioned, the, uh, mounting hole down in there.
So possibly, um, there could be a short in there. look at that. We might be getting some residual heat from the connectors there cuz th, those connectors are quite hot, right? they're like 40 plus de they're they're getting quite warm cuz there's a fair bit of contact resistance in there, so it's going to heat up so, but that board next to it is getting quite quite warm. Hang on 35 and climbing on that on that pad down there. Look at that 32. Maybe there's possibly some heat spreading through that ground plane, possibly even all the way over to here because I wouldn't expected that to be, uh, warm at all over there. But this one over here is definitely hot. There's something going on there.
If you design this board properly and these large, uh mounting holes here, you should have a power plane pullback. What's called pullback from around the holes so that when you know you put screws in there and you tighten them all up, then it doesn't. uh, crush the power Lanes inside there. But who knows.
Um, you know, unless you had the Cad files for this thing that can check, or you dissected it, or you x-rayed it, or you know something. uh, like that. Um, you don't know how close those are, so it's not out of the uh. You know the bounds of possibility that when this thing was dropped, there could have been pressure on a through hole.
uh Moun uh, pad like that and it could have, uh, crushed the internal power planes. This is a six layer board, so there's not going to be a huge physical, uh difference. Uh, you know, a physical gap between the ground plane and the power plane. So um, it depends on what layer the 3.3v uh ones on.
Usually it's going to be uh, right next to the ground plane. So um, you know, with a very small amount of um, Fr4 prepreg between the layers. So it's not out of the bounds a possibility that you could crush that due to something like that. So what can we do to test that theory? That, uh, it could be something to do some some sort of short on the power plane in there.
Well, we can freeze it, We can cool it down and uh, see if the current changes. Heck, it might even go away if you freeze it down enough. Um, you know it, the the short could actually vanish. So if we're right, we might actually see.
um, that current change now. I I Don't have any freezer spray, but I do have air duster. So what you do with the air duster is simply turn it upside down and uh, all the um, uh, the coal gas actually settles at the bottom and that comes out instant freezer spray. Cool down that pad and see if this current changes you ready.
No, no, nothing. It's not changing at all. That's strange. Would have expected something.
Maybe our theory is not correct. You can actually see how it's freezing down in there. It really is. That is pretty darn cold.
and if I get the thermometer on that again, you know 16 138 you know it is very, very cold. That pad is now very cold, but it's still showing 10 and 1/2 amps. H Now here's that pad up close and I don't see too much. uh well. really any uh, major compression, uh, trauma or anything like that on the on the pad. Well, and the same goes for the bottom there. I Mean it's not like, you know, um, somebody's like really screwed this thing up and and compressed those? You know that looks pretty darn, uh, normal, uh, wear and tear on those mounting pads to me. So, um, possibly in the connector? I don't want to have to desolder the connector unless I Absolutely, uh, have to.
Now they look okay. but of course they could. Um, you know it's it's a remote possibility, but we're getting down to that point where we have to look at remote possibilities now. I Just thought I'd check something.
uh, these mounting holes. uh, that we're uh, curious about to see if they're actually, uh, grounded. So yes, they are, are actually grounded. There you go, right? So if there's not adequate, uh, plane pullback inside those things, um, they could, uh, certainly.
um, you know, uh, compress or short out or something like that with, uh, physical damage. Now I Think we've gotten to the point where we probably have to, uh, power up the other rails as well and just, uh, see what our current draw is on that 3.3 volt. Rail And yes, I've measured it. and this, um, 3.3 volt rail does seem to be dead on this power supply.
So um, the power supply seems to have, uh, died as I said in the uh first video. it uh, and the tear down, of course it was actually working and then it died. So I suspect that the short eventually uh, killed the thing. Now this is my theory on this.
There is definitely a short on this board. Okay, 0.1 Ohms is not correct, right? It's just not. There is definitely a short there and the thing. The power supply was good enough to continue to keep the thing basically working, but we were getting corruption on that 3.3 volt rail.
I.E all the capture memory which is kind of what we were seeing um in the Uh data. So all of the um, all of the capture memory which runs from the 3.3 volt rail. there's probably excess Ripple or whatever was going on there. it was hiccuping or doing whatever I don't know, but um, it was.
um, so it was able to still power through that short just like our high current supp here. can it can still Supply 3.3 Vols but it was drawing too much and it eventually got to the point where eh, no, I can't take this anymore. Kaput, it's dead. Okay, so what we're going to do is um, but all of the I have measured all the other rails and they all still work.
So I'm going to cut into this uh cable here the 3.3 volt one. we'll power that from the external Supply and we'll power up all the other ones as well and see what we get, right? So what I've got now is I've got it powered up and and uh, we check the voltages on here. 5 Vol There we go and we're getting minus 5. Vol So all the other rails Are all, uh, all up.
not exactly sure what they are, plus 15. so the other rails are up the plus 5 and all that. and but look, it's still drawing 9.6 amp. So there you go. That's busted. that theory that, um, it was. you know it was, uh, latching. You know the power was latching into other unpowered rails through protection diodes and you know, all that sort of jazz.
So my hunch was correct that there definitely is a short there. And it's not surprising because 0.1 Ohms is not normal now. um, Sam has also, um, confirmed that yeah, the 3.3 volt rail should be drawing about 6 amps. So it is.
And he also said as I said before that it's you know, it's over an OHM on, you know, when you actually measure it with a multimeter. So there is definitely a short on this board. No doubt about it. don't want to hear anymore And just to doubly sure.
Yes, I've plugged in the processor board here. and yes, we're still drawing 10 amps on the 3.3 volt rail. I Just want to try this uh, temperature again. But I leave the connector in now so that we've got the uh, well, the two pins going through to the power plane there, the third one seems to go off to somewhere else and um, and also, I'm using the ground point over here.
So um, we should get less heating on that uh on that pad due to the contact resistance we had before. So we're still 33 on that pad down in there. So whether or not that's actually normal, um, due to the um, the contact resistance of the connector at Um 11, 10 or 11 uh amps and then the heat is spreading through the power plane see, cuz even this pad right down here. you know, well away from the ground point over here and the contact resistance over here is still.
You know that's still 33. So clearly that's just the residual heat spreading through the power plane. so it is not obvious yet where this damn short is now. A few people ask why don't I use my aim TTI uh I Proba 520 It's because it's it's not really going to work.
You're going to get positional issues as I was shown in the video for this rotational positional issues on this sort of thing. So if you put it here, let's okay. I've got it into the multimeter instead of the uh uh scope cuz we're dealing with the DC here. It's just fine.
and if I rotate that, look at just the positional difference like I'm not I'm just rotating that. Okay, so just the ID just the you know rotating that is is going to be all over the shop there. You know you're pushing. uh, the brown stuff up the hill with a pointy stick trying to use this thing I'm afraid it's not going to work.
You could do it like using an AC uh method on the rail or something. but oh no. ugly. So it's just simply not going to work when you have a huge power plane like this thing.
So no, sorry. As cool as this thing is, it's not going to do the job in this particular instance. Now, as I mentioned before, you can do the thing with like feeding constant current through the thing and then using this, um, to measure, uh, voltage drops. But the problem with that is when you've got active devices on here, um, drawing huge amounts of current, you're naturally going to get the drops across here. anyway. So um, you know you're going to get drops across the plane. So here's our input connector and all these four Asic devices here are all drawing. You know, large amounts of current.
so unless you could, uh, isolate those, um, it's going to be incredibly difficult to find the issue. Like let's go from this cap here, for example, to this one: I think I got the right one, no the other pad here. Okay, so just from one As6 to the other, we've got a 13 Molt drop there. Okay, from there, from there to there across the Uh plane there.
okay, 13 mols. And then if we go a bit more, we'll find that we've got more drop Oh wrong pad. There we go. We got 40 MTS drop across there like that.
and then if we go over here, we're going to get a larger drop. If we can get in there, There we go. We got 81 m volts drop. So you know, how can you really essentially narrow that down? When you've got such large currents drawn across a plane like that, it's very difficult to do so.
I Don't think that's really going to work in this particular instance. It can work in other Uh instances, but this one not so much. So we can do the same thing again with the um uh Instead of the Uh drop on the rail uh itself, we can get the voltage drop of the rail instead of the voltage drop across the uh, the resistance of the tracks and the plane. So we can start out over here and we're getting at the connector.
we're getting 3.32 Vols input. Okay, 3.03 25 for example. And then let's go to our cap over here. Bang.
It's dropped a bit. You'd expect that just due to the plain resistance. then we'd expect it to drop a bit more. There we go.
2.96 As we go to this. ASC over here, 2.93 There you go. It is dropping as you progress across the plane like that. What does that tell? you? nothing? essentially.
Now, if we scan over the board with our IR thermometer again, we can have a look. You know, 61 63. It sort of Peaks that 64 63 similar sort of peak in 65 peak in there so you know there's not. There's nothing in that really.
Go to the other A6 63. You know that's got slightly more heat sinking on that one I think so you'd expect it to be a little bit lower, but uh yeah, you know, like there's there's not much. That one though is only I know there we go. We can get it going up to almost 60 something like that.
So you know very similar temperatures across the board and it drops as you go away. for example, pretty evenly. As I said that, um, uh, temperature is going to be, you know, fairly evenly spread across that plane. So trying and detect this is oh, almost needle in a hay stack. So this thing just doesn't seem to be doing the business. I mean it's only a single spot. What we need is more resolution. unfortunately.
I don't want to have one of those multi th000 um thermal. uh IR cameras. but I know someone who does. let's go.
So thanks to my mate Charles at Trio Smart Cal Pveers of Fine Test Equipment pornography I Have just that. Check out these puppies. I've got three um, infrared thermal imaging cameras. Fantastic! We have a Flur E60 Um, this puppy goes for about $112,000 It's $ 320x240.
It's the pretty much the Rolls-Royce in handheld thermal imaging night cameras I mean Flur are the best in the business and uh, the E60 is the topof the range. eer Oh beautiful beautiful bit of Kit And check out this folks. look at this made in Estonia Woohoo would you believe it? There you go. And then we have a uh UL Vision uh brand TI 160.
These are much cheaper, doesn't have the same resolution as um, you know the uh Flur of course, but um, still this one's much cheaper. I think Charles said this one's about $ 2,800 Australian dollars or uh, something like that. Don't uh, fully quote me on that and then we have the same brand again. UL Vision TI 395 This is um I think a couple of thousand more expensive than uh, this one and it's a gun style one and it's almost identical.
uh, look and feel to the SEO Xati camera I used to use for the blog. Check it out. it's like, you know it's um, these gun style ones. almost identical look and feel.
Oh man. I Love it. So that's a nice little, uh, funky form factor I don't really like this form factor on here. in fact.
um I'm probably after I finish uh doing some thermal measurements of this I think I'm going to do a separate video just uh, playing around with this thing. uh, these things. So I've only got them for the afternoon. So um I'll have a further play around with them later.
But what we want to do now? um, we've got the right tool for the job now folks. because um, this board if you remember just a few minutes ago, um, was, uh, you know, uh, we couldn't find that, um, any hot spot at all apart from the uh As6 we couldn't find it like around the hole down there and the wires and stuff like that. It seemed to be fairly evenly spread around the power plane in this thing. So let's uh Power it up.
And let's say start out with the rolls Warus, let's use the Flur E60 and see what this puppy is capable of doing. Welcome to the world of infrared Flur Oh, just want to hug this thing. Oh, it's pornographic. It really is.
Oh man, what a great toy! Let's start out by taking a look at our heat sinks. Tada Here we go and uh, if I this is the focus on the front here. so if I adjust the uh, just a focus, we can uh, see the temperature and you can see the uh spot measurement there. We're talking.
you know, 75 something like that? I mean this one's only 79 but this second heat sink here seems to be much, well, by much hotter I mean you know three or four degre hotter than uh, what we were getting than what we get on the other ones. And considering that this one here is the same heat sink so it seems to be a few degrees hotter and uh, this flare camera has various, uh, different modes of course, which uh, we need to investigate and look at, but it you know, so that may I don't know if that, um, you know says anything about my original prediction in the first video that that second device there seemed to have, um, a slightly lower resistance than all the others. So maybe it's no surprise that it's getting hot. but um, there's check out the resolution on this thing. it really is very very and I can take a can take screen captures of that and all that sort of jazz. So right so we know that those heat sinks are all, um, you know, pretty much identical temperature except for the uh second one there. Now if we scan around the board, we can see our our connector down there. If you have a look, um, I've got the uh I'm I'm deliberately not um, putting it right on the pins there I'm deliberately getting it away so we can, uh, look at the current flowing through those wires with the thermal imaging camera and also, we should see, you know, hot spot around here where it's connected.
um, but let's see if we can find a hot spot on there and in particular, that um, mounting hole down in there. So if we take a look around there you go, you can see the connector and you can see the heat spreading out from from that connector there. I Mean that connector is getting up to 42 it is. It is actually pretty pretty darn hot.
Now if we have a look at the connector there, we can actually see you see how the two right hand wires there are like that one. You know it's like 40 on those wires and the one on the right hand side that I said had uh, no current flowing through it. a very little current. You can see that there is no current flowing through that other wire now.
It's tricky to sort of get this cuz we are getting heat spreading from this heat sink here. so um, it is. It is rather a bit tricky. it spreads across, but you can see that there's no hot spot on that whole mounting hole down there at all.
There is no hot spot, so there's nothing going on there. You can actually see inside the inside the actual connector as well. If you can see inside that, there you go, you can actually see the connections heating up inside the connector. but no hot spot on that mounting hole.
so that pretty much rules that out there you go. So it is not the mountain hole which a lot of people thought. In fact, the short doesn't seem to be anywhere else on this board. I Can't see it now I Can't see any other hot spots around the board here on uh, sorry.
it's hard to really get this thing on camera, but there no are no other hot spots apart from those four heat sinks. and of course that uh, Second one which we'll call the second one down there does look hotter than all the others, but there are no. There's like nothing I can't I can't see anything. There's no other components that actually, um, show any hotpots at all. So this is not looking good folks. Let me tell you it, uh, is looking like it is something to do with those As6. Now if we pan around the rest of the board. Now here's an interesting thing: You might see a hot spot on these on these B andc connectors here and you might think they're actually hotter than other you know than other parts of the board.
but that's not. That's just heat reflection. And you've got to be very careful with shiny objects like this. You will actually get, um, heat reflection off the these components.
So if you don't use it correctly, you might think aha, that B and C is getting hot. but it's not. It is, um, just reflection. If I put my hand near that, you'll see it change right there you go.
If I put my hand near that, you can see the heat reflect reflected off those shiny B and C's and it can also do the same thing on solder joints and things like that as well. So um, that's a real trap for young players. With using these things, you've got to know about the reflective properties you're actually viewing. Now let's have a look at the back of the board, see if we can get it.
and uh, you can see the hot spots on the As6 there of course. once. that's about 65 on the backer there. and is that yeah, that's the first one and the second one 67.
So once again, it's a bit hotter. third one, not 65, not quite as hot and and the fourth one down there. but there are no other hot spots. I Mean, there's the connector down there, right? it is.
It's showing, it's showing nothing around that mounting hole. There's let me get that. InFocus that's the mountain hole down there and I can't see anything. any issue there whatsoever.
If you go up here, you can see the connector up there. Got the ground clip? No, nothing. and of course no other part of the board is um, powered. Now here's another example of one of these traps for young players you might think.
Aha, Look, we found right in the center there a little hot spot. It's a couple of degrees hotter and it certainly shows up there. But it's not what is it. It is these shiny, uh, solder pads here.
They're the things that are showing up on there as a different temperature because they're shiny. It's not because they're actually warmer. So yeah, when you're using these things, just make sure you really know how to use them because there's quite a few traps in there. But no, sorry folks.
Um, this, uh, $112,000 piece of magic is telling me that there is nothing wrong with this board at all in terms of actual shorts on the board. and if we use the Ulir Vision camera as well, well, um, we get, well, the same result. I mean I can scan over the board again. but there's the uh, there's the connector down there if we're going to have a look at that and once again, there is no Hots spot on that hole at all. You can see the connector, not sure why it freezes there. it says rectifying. There you go, you can see the two wires heating up look that that works brilliantly actually. and you can see this one.
um has a uh track. Well, the Flur has it as well. It has an auto hotspot uh tracker which tells you the hottest spot on the screen there. but um, you can see see the two wires on the right hand side have all the current flowing through it.
The one on the left has bugger Roll. It's just getting some residual heat from the others. um, as we measured in the first video and there's nothing on that hole at all. There's nothing there.
and if I scan the rest of the board, there's nothing over the board either. Once again, hard to capture on camera. I've done it off camera and uh, thoroughly and there's nothing there. Zip! Now, if I switch the other power rails on here, you'll notice that it dropped from Uh 105 there to 8.6 so it looks like there might actually be, um, some current going through the other unpowered uh rails there either through protection DS or whatever mechanism um is in place on those uh rails to actually, uh, do that So, but still.
you know that doesn't explain anything. Um, that's a complete non secer. So H all that's left is to current limit this and see if we can get any temperature differentials. but I reckon all those are dead now.
I'll have a quick uh scan around the board again with just the 3.3v rail and as you can see apart from the Uh heat sinks and the connector, I've only just uh switched on. everything else is, uh, everything else is dark apart from those heat sinks. So let's um, switch on all the other supplies and we should start seeing a few other things heat up as well. So yeah, I can start to see some uh, other circuitry on the far.
Yeah look, you can see all those chips on the far side there start to heat up cuz they're all the 5V 5V rail stuff. you can see they they weren't lit before. This thing has a laser on it too. by the way.
you can actually switch the laser pointer on and it does, uh, show up. but there's uh, some parallx error there on this thing so it's not that great. But you can see some of the other chips in there heated up. Another one over there is getting a bit warm.
47 de and uh, but yeah, not that exciting. There we go. That one in there, there's another. That one in there is getting quite warm.
it's getting up to 73. that one behind there. it's really quite warm. Yow Ouchy.
Now this is interesting. Check this out. I've got a constant current uh of 1 amp over here, so there's my, uh, constant current and voltage has dropped right down. Of course it's no longer 3.3 volt. It can't provide all the uh, well, it's it's current limited. that's why. And if I switch on, switch on all the other rails, check out. This little chip gets hot really really quickly.
He gets up to 100. But I don't think it's foldy. it's just that the fact that that voltage on that 3.3 volt rail has dropped cuz that goes away if that 3.3 volt uh rail actually goes up to normal. So um, yeah.
well. I mean I've checked that Chip's actually connected to the 5V rail. but obviously there's uh, something on, maybe the inputs or something like that that's causing that sucker to uh to heat up. And if I turn on all the other rails without the 3.3 check out the hotspot inside those chips.
You can actually see the dye heating up through the plastic package of that that is really quite. That's really quite something. I like it. and once again, that uh chip over there is getting red hot as well.
That's if you don't power the 3.3 volt rail. So those chips really do not like that at all. With the power supply current limited on one amp, you can see the ADC heat sink heat sinks over there are still, uh, warm. and you can see once again that that second Asic we've been looking at is warmer than is still warmer than all the others.
So, but you know that really doesn't account for all of the difference in the Uh Power If these three chips were working and this one was faulty and somehow got Scr latch up or some other damage uh, internally which shorted it out across the power railers, then, well, I'd expect it to be massively, uh, a massive temperature difference compared to the others. But as it stands, it's only a couple of degrees. so really, there's nothing in that. Well, sorry folks.
I Think that's all I'm going to do on this one today because I think um, I'm going to call this one and uh, declare it unrepairable I think something has died in this uh, power supply which is killed taking out everything on the 3.3 volt rail. everything being the only four devices on the 3.3 volt rail, which are these four A6 here. Unless somebody can come up with another very plausible um, explanation for what's wrong here. I mean I cannot find a spot where it's um, you know, dissipating all that extra power.
we know for a fact, the rail is not supposed to be uh, 0.1 OHS obviously. um, it's supposed to be an order of magnitude greater than that. We know it's not supposed to take 11 o We know it's only supposed to take uh, six tops. So all that extra power is going somewhere.
and we've used a $112,000 IR thermal camera G all over this thing and cannot find a hot spot at all. All the power is being dissipated in these 4 As6, so I can. and because they're almost essentially a uniform temperature. The only logical conclusion on this one, folks. as much as I hate to say it: I think it's unrepairable. four identically dead A6 that one slightly more dead than the rest I might have some more fun with it. Maybe I don't know. Maybe suck out this second device and see if it makes a huge amount of difference.
but I no think we're got a loser folks. Bummer. Anyway, if you want to discuss it, jump on over to the Evev blog. Forum Catch you next time.
Blimey the DAVE has to go to a mate to get a thermal cam ! how things have changed in 13yrs now his kids are prob playing wit one
Talk about grasping at straws. No one in the repair field would ever use thermal devices to trouble-shoot a short. Lifting the Vcc pins on each IC would have been the logical next step in this saga. It's also very rare for a bypass capacitor as being the fault. It's much more common that a semiconductor is the culprit. There is also the remote possibility that the short is between one of the layers in the PCB… 0/10 for this effort Dave.
I´m still wondering about the melted relay?!!!
i would think in this case hitting the UNDERSIDE of the board with the FLIR would tell you a LOT more
When you have a short, it won't become hot )
Usually, shorts are cold in terms of temperature )))
You'll see a lot of hot other elements, because when you putting a lot of current – you'll see overload on working lanes…
I was really surprised when Dave showed that the Flir was made in Estonia. I'm an Estonian and I didn't know that Flir has a factory in my home country. Estonia is a very small country (1.3 million people) and it is very cool to discover a shred of it from a video of world-renowned Australian Dave Jones:). Now I found out that Flir was a nominee for the 2016 Foreign Investor of the Year award (didn't win it though).
Well remove that second chip and see how many amps it drops it wouldnt hurt if it doesnt drop by alot then call it a loss if it does drop below 6amps them maybe thats ur answer
good job dave i had the same problem with kind of sets i dont like the lay out of it
Probably suggested before but use a thermal insulator such as a bit of plastic or foam to prevent spill from affecting your thermal readings.
bought one and I only check where red is and replace :-d
Was wondering why the Flir camera was not the first thing used (in this video and the previous)! Then I saw the date of the video, and heard the price at the time. Lol, how things have changed.
Difficult problem.
dam you murphy!! can u take that second one out and test it? crowd fund a replacement set!
I'm sort of a beginner in electronics but I'm in love with that thermal camera. I don't know how usefull it might be right now to me but it's totally cool. I feel like I need it xD
Do a Power up disconnect the 3.3 from the standard psu and then connect the 3.3 rail so something that can handle the current and see if it at all powers up. That's what I'd do.
I got it from a VERY RELIABLE source, to come here to see some sensational video…. WTF is their no EEVblog #401 – Lecroy 9384C Oscilloscope Repair – Part 1??
For fuxsake??
I'm pushing the brown stuff up what now?!
It's two wires on the right. Not two wires on the right-hand side. What does the 'hand' have to do with it., and why associate body parts with left and right. That's infantile kindergarten speak. Snap out of it. It's either on the left or on the right side. Maybe it's on the right ear side. Get the point. – Other than that .. Good video. So no more associating body parts with left and right from now on. We are not that dumb.
in all that board there is not a single zener or something for protection?, that psu 3.3 channel has a shortcut its rectifier?, like sending AC and its reverse polarity to the 3.3 raid?