Hi. Just a quick heads up on this video. This is actually not the full version of the video. The full version is 50 minutes long and it's available over on my Eev blog 2 channel.

You can click up here or down below to see that because some viewers have already seen this said. hey, 50 minutes is a bit long. Can you cut it down a bit shorter? So this is a shorter one? just more focusing on the repair aspect to it. But there's a lot more in-depth discussion, especially on the soldering repair, the side of things, and other miscellaneous.

uh, like all the teardown, uh, commentary and all that sort of stuff. So this one's pretty much just the basic repair process. And uh, the full video has a bit more, uh, you know, dead ends in terms of, uh, troubleshooting as well. So if you want to play along at home and follow the full thing, check that out.

So do let me know if you think this is a good idea or not, like split these longer videos up and have the longer version over on my second channel. Anyway, here's the short one. Let's go. Hi.

it's product repair time. We've got more studio monitor speakers. I've done a few studio monitor speakers before. They're always fun and interesting.

So this time we've got our Jbl Series 3 Lsr 308s for those playing along at home. these are the new 3 series Jbl's and by most accounts they're actually really excellent. Uh, studio monitors At least you know for the price, but one of them's working. one of them's not.

Let me just switch it on. There's no hum. There's no like loud bang. I can't It doesn't smell bad.

always do smell test first. Um, maybe it will be when we open it. Who knows. But uh.

anyway. lead comes on and then it starts to flash. so I'm not sure what that means. I guess I should Rtfm.

Um, but the other one doesn't do that. it just stays on. now. This is really weird because I downloaded the manual for the 308 and I couldn't find any reference to blinking or flashing.

It was just the power Led, huh? Let's open up and the other speaker works fine in this one. I just get no audio out, so that is the symptom of this thing. All right? Well this? uh, oh yeah, yeah, she comes out easy enough. Special Whiz-bang tuned port on that baby.

so all you tuned port aficionados are all going wild. There you have it, we're in. Uh, he kind of stopped a spray painting and around there yeah, like it's neat and tidy. Earth points: One's there and the other one is down there.

There's two wires going off. one actually pops up over the top here. Anyway, they got a uh switching supply on this thing because well, its mains are straight in here the first thing of course. As I said, give it a smell test, there's nothing obvious and then of course you look for the caps.

and well, there's no bulges in the caps. There are none that polymer rubbish. They got good old liquid in them. no bursties on the vents or anything like that.

So that's all looking sweet. The primary side cap are usually, uh, they're fairly good. But yeah, the next thing, uh, to do, of course thou shall measure voltages and luckily, um, we have voltages over here. Check this out: 3.3 volts and 5 volts.

Why have we got 3.3 Because look, it's a crystal. and sure enough, that's got to be a micro. Why do we have a micro? And that obviously explains the flashing lead look. We've got another crystal over here, so this thing is.

just even though this is not a Usb like digital input thing, this thing has at least two intelligent doodads in it. And yeah, I think I'm right about that connector. Yep, yep, it popped off and there you go. And the bottom side of the board? up here, you can see all of the residue from the dip soldering the wave soldering process, so that's not going to cause an issue.

but you can see that we've got some primary side switching stuff over here. This handles that's the primary size switching controller. Anyway, the silk screen is reasonably informative, and I love the fact that they've actually put the silk screen of the parts on the top, actually on the bottom and underneath this heatsink. I can only presume that that's our main power amplifier.

I see it has to be looks like they got the crystal. I I thought it was going over here. Crystal is for under there. So I think we've got some intelligent class D type doo daddy? uh.

amplifier here, st. Micro down there. And what's that little jobby down in there doing is that that's a Cirrus? That's Cirrus Logic. Oh, does this thing actually sample the inputs and then uh, process them and then why you'd want to? I don't know.

Uh, because it. well, it's got filtering. but I would have assumed that was like analog filtering. but maybe they're doing that digitally.

So yeah, I reckon under there is some sort of Dsp Class D amplifier. Ah, it's a bit rude, isn't it? And yep, that's a Cirrus Logic 5341. that's an audio grade Adc, so that's clearly taking the inputs from here and just, um, basically straight in sampling those. So all that, uh, analog stuff.

You know, the high frequency, low frequency adjust, and all that. That's all done in software. So the micro which is an Stm uh, job down here. Nothing uh, special about that.

That's just like, uh, reading the input switches. Maybe it's like setting like, uh, the coefficients and I'd say the crystal is probably, uh, shared between both of those. So yeah, that's just. um, analog to digital digital out straight into there.

And then the micro here. Just yeah. um, tweaks some coefficients and does some stuff. and that's it.

So it's like a software-based speaker. It's quite frankly, it's rude. and there's a test connector down there. It's got like a reset line.

it's got boot. Um, can we actually maybe we could reset the process? No, But it seems to be going through like, like the process is almost certainly working. You're not going to get the power on delay and then the flashing lead when. Like, if the process is not working.

So that's fine. but obviously something wrong with the sampling or the apple. you know it could be just a blow-on output driver or something like that. Basically going to have to do some digital debugging in a bloody studio monitor speaker sacrilege.

What's the world comer do? And sure enough, there it is. That's an Sta 350b with a newfangled Dsp thing. It's some fancy 50 watt, uh, per channel digital audio system. Yeah, okay.

anyway. um, yeah. it has a heatsink on top. doesn't have a thermal pad on top though.

It's uh, I believe it's got a thermal pad on the bottom of the package which will no doubt be soldered down to the ground plane down there anyway, so we can safely power that up without the, uh, heatsink. It'll probably work up to 10 watts per channel or something without the heatsink, because you've got to remember, this is not a linear amplifier. This is a Class D amplifier. You don't need much of a heatsink.

it's going to the Class D. Audio amps are very efficient. So I'll start by measuring the digital power supplies. With any luck, that'll be easy peasy.

Three point, three, and five. All right. So the digital supplies are good. Let's go for the other stuff.

Uh, we've got a plus. eight volts there. And the analog supply rails. Um, these newfangled, uh chippies.

They don't have your traditional R plus minus uh supplies. It'll just have one rail. and there it is. Uh, plus v uh.

data sheet for this sucker says like a maximum of 30 volts so I don't know. it could be like 24 volts or something like that. 7.96 good enough for Australia and 26. Uh, yeah.

okay, that sounds fine. I think that's all she wrote for the rails on there, so I don't think it's a power supply related problem. Hmm. And just to prevent any brain farts where I accidentally touch the main section while I'm troubleshooting, uh, the low voltage side of this thing, I've just got an old food container takeaway food container wax over the top like that.

It's a Bobby Dazzler that'll just, uh, stop me. You know, accidentally brushing the live heatsinks. All right. So it's time to get the scope out, I think.

And uh, let's have a look at our lead line. there. There you go. Blinkity blink.

So our process is doing something now. As you saw before, there is a reset line there. I'm going to probe that reset line and that is normally high 3.3 So I wonder what happens if we just uh yeah. it's probably got like a power up reset thing, but I wonder what happens if we just reset that? So the way that you can do that, use your multimeter as a shorting probe working in the milliamps.

Gonna have to use my third hand here which is the scope just holding that probe in place there and let's reset this sucker, See what happens yet Low and then it should take that few seconds to reboot. And will it. Will it come? good? It's oh no. There you go.

constantly on. and then it started to blink again. Didn't expect it to fix it. but yeah, and that crystal down in there for the Tda output amp and also I believe the Cirrus logic chip.

There are 12.288 megahertz. Yep, no worries. Now there's a Tx and Rx on this little connector down here. Obviously like, um, probably some sort of like test connector or something like that.

Um, and now I can't get anything on there when I'm uh, booting up or anything like that. There's just no activity. Now I'm starting to realize. okay, so the processors obviously are working.

the clock up here is working. We've got to have some sort of like a preamp thing down on that bottom board, so I might just take it apart again and ta-da there it is. because someone could have blown the snot out of that input. So yeah, you never know.

That's just like that's just the switch uh, board. So nothing doing over there. But certainly we have an input amp down here and it could certainly just be that board. that'll be nice.

We can isolate that, uh, but hang on for a second. Uh, that doesn't explain the flashing lead, Does it? If we had a blown input preamp, it just simply wouldn't work. The lead wouldn't be flashing. I'm going to rule that out anyway.

So I'm just going to go in there and probe the Adc here just to see that it's like outputting stuff. That's not too hard. You don't even need to know the pin out because there's a whole bunch of resistors over there. And I'm guessing that there are series resistors for the various uh, I2s, data and clock lines.

So get one side of the resistor. Yeah, look at that that looks like a clock 3.072 next to it. Yeah, that looks like data. There you go.

That's just random that'll change every time. Don't don't worry about the overshoots there, because I'm using a, uh, hideously long ground lead. all we're looking for. We're not looking for signal integrity here.

we're just looking for data. And uh, yep, there it is. Again, the Adc is doing its business and outputting stuff, so there's nothing wrong with that. Um, unless, like as I said, like, the signal is not coming in, but that even if the signal wasn't coming in, that doesn't explain why the lead is flashing and the lead flashing comes from the Micro here and it shouldn't be doing that.

And the flashing Led is only on. Uh is for the power down mode on the Mark Ii model. This is not the Mark Ii model. Can't find any information on the flash inlet at all.

I I presume it's some sort of fault code. So Micros working. Adc's working. It's outputting stuff.

But unless you want to decode the I2s data and stuff like that, you know I don't want to go down that rabbit hole yet. So I don't know what's left the output power amp. Okay, what I'm going to do now to see if uh, anything's getting into it. Like the input signal is getting into the Adc and the Adc is changing.

I've got a signal generator that's going to generate a one kilohertz tone. So this is the data that we get. Uh, at the moment, I just want to see if that changes that's fairly consistent. So let me play this and see if that changes.

Then we know that our input, um, you know it's getting into the Adc and the Adc is outputting different data. Yep, look at that. Yeah, there's more solid data there. Yeah, look, so that mean says looks like you know there's nothing inherently wrong with the input amps.

Something's getting through, and the Adc is certainly changing the data on the output. so it looks like the data is going to the Power Amp chip. So I would say the problem is with the Power Amp chip. But why is that lead still flashing? Why, Oh, there goes the phone, right? So now we have to start suspecting the Sta 350 chip here because our Adc seems to be working.

It's outputting data so we're getting no audio Out is our symptom now. Two lines that I'd immediately be interested in are like is this chip being shut down Now there's a reset pin at 31 Here This is where the data sheet kind of sucks. It doesn't really tell you you have to assume that it is positive polarity because it's It's not a not bar, so there's no, it's not. You know, slash reset or reset with a bar.

So you have to assume that it's a positive going. reset lines up. It's high, It's reset. If it's low, it's fine.

and then, uh, this Int line here is actually an interrupt line. This is an output and it's a fault interrupt. and the data sheet doesn't go into detail of this. And once again, it doesn't tell you the polarity of this thing or what it does, but it's basically a fault.

So if there's something wrong inside the chip, this should give you once again, presumably active high. It should give a high. out. And as it turns out, if you look closely on the chip, down in here, you notice pin 32.

It's a little test pad right there. so I'm going to probe that, see if there's a fault. And hello. Bingo.

3.3 volts active high. That chip is giving us a fault output. Um, maybe I can repair it. So let's leave the scope on there.

All right. There we go. It's drained and turn it on. It's low and then high.

Low high. It's giving a fault output. There's something wrong with that chip. and just for good measure, I'll measure that reset line.

It goes to our resistor here, and it's also high. Like, I assume that the fault line goes fault, interrupt goes back to the processor. it's reading it, and then it goes okay. there's a fault.

Well, I'll just reset the cpu. So here's the reset line. It's obviously under active control. Low High Low.

It seems to correspond to that, uh, fault interrupt line. So yeah, so yeah. I can only really come to the conclusion that the chip is faulty and it's got, you know, extensive Dsp and diagnostics and stuff in there and it's reporting default back to the processor. The Uh designers have implemented uh this properly and they're actually reading the fault interrupt uh line and the process is going well.

If there's if this thing's saying there's a fault, then I might as well reset it just so it doesn't blow up or something. Um, yeah, it doesn't like it stops driving, it disables the outputs so I presume that's what's happening. And for a second there I thought, well, maybe just maybe one of the speakers might be at fault and I don't know. shorted uh, coil something like that and it might be detecting a shorted load or some other thing because if you look at the data sheet it says about this fault, uh, interrupt line that it does.

You know various like over voltage and over tight protection and stuff like that. But no, just repelled it and the light down in there, you might be able to see it go on fleshy flash. so no, wasn't that right? So what's actually causing this problem? Well, we don't exactly know because there's just not enough detail in the data sheet to, you know, know exactly what's wrong. Apparently you can actually read out some registers and stuff like that, but obviously we're not gonna go to that effort.

Um, so basically there it comes down to two things. Whether there's something internal, uh, to the chip like you know there's some blow-on output mosfet, or you know, something like that. Um, and it's actually detecting that or whether or not it's something external. Now look, we've got some ceramic caps up here and of course, I've done many videos on ceramic, uh, capacitor failure and things like that.

Very fascinating stuff. By the way, linkedin at the end and down below if you haven't seen it. Um, and there's some like on the speaker outputs here, so like maybe something could be loaded on the output. You know, I might have a short in a cap.

So I'm going to get this board out again, do some visuals, of course, and then probably start Just you know, probing around just to make sure impedances are okay. with uh, various caps and things. There's not much else that can fail. There's the electrolytic there.

I have actually measured the ripple on that main, uh, supply rail and there is none. so it's just fine. So yeah, I'm hoping it's external because I don't want to have to suck off this thing with the pain in the ass pin pitch and the thermal pad down the bottom and actually replace that. that would be.

that'd be a bit of a kick to the guts. So yeah, yeah, I'm hoping that it's external, but uh, you know, Murphy. Yeah, you just want to go around, make sure none of those are shorted. like opens aren't going to cause a problem.

So like, I'm going to presume like it's some sort of, uh, that's measuring. uh, about nine meg. by the way. It's fine, so opens aren't going to be a problem, but if something's you know, shorted then that could cause a problem.

All right. Well, I've measured, uh, every cap in there and nothing's shorted. So with the power rail, uh, fine, I assume like 26 volts is okay. No external loads it.

I get the um, the same fault uh with or without the speakers loading it down. Gotta say, it's gotta be internal to that chip because all the external supplies are fine and yet it's reporting some sort of internal fault. Uh, So I haven't checked the timing of this, but I believe that this thing, uh, reports back. It gives that uh interrupt fault back to the Cpu.

and then the Cpu goes. oh, interrupt fault? Okay, reset and shut down. So that's why we're getting no audio. Well, here's a plot twist.

That test point we've got there, which is the uh fault interrupt output pin. It actually doesn't seem to go anywhere. There's no corresponding via under the bottom of that. so that interrupt pin is not actually going back to the micro.

So the micro is not actually detecting it via the interrupt. But as I said, there are registers inside that it can actually read. So regardless, the interrupt is just to alert the micro. So the micro is most likely uh, reading those fault registers and going.

yep, there's a fault there and it's just going blinkity blink on the Led and that's it pointing towards that chip. I? I've got nothing else. Okay, we have determined that the Sta 350 uh power chip here and Dsp thingy is most likely dodgy. so we need a replacement for that and Ta-da I've magically got some replacements.

This is the 350b W and it has the thermal pad on the bottom of the chip. They also do a version with a thermal pad on the top depending on what heatsink, uh, solution you're using here at least, So you're going to make sure you've got the right version. Uh, it cost me a whole four dollars. Um, each for one of these things.

I got two just in case. So how do we replace a thermal power So uh, package. In this case it's a it's a regular So package, but it's got the thermal pad on the bottom. So another way to do it is to actually, uh, separate the thermal mass of the chip from the pads so that you use different techniques to desolder the pads as you do the thermal pad on large thermal mass pad on the bottom of the chip.

and you do that by getting a sharp exacto knife in there and cutting the leads. Now you have to be careful like there's pros and cons of all these uh techniques of course. And then this chip, you can actually remove this without potentially damaging, uh, the more delicate pads on either side. Now this technique requires two things.

A It requires your tongue at the right angle so that you don't cut any traces. Next thing is it requires a shallow angle like this. So if you've got like large components around it, like if this chip was facing the knot, that direction was facing that orientation, I'd have a hard time getting in there with my exacto knife and actually cutting out those pins. Because if you have a high angle like that, you're going to put lateral force on those pins and there's not much adhesive on each one of those tiny little, uh, piss ant pads in there to, actually, um, stop them ripping off the Pcb and you've got to be careful not to cut the traces underneath of course.

But if you keep a shallow angle, uh, you're pretty much guaranteed not to do that unless you've got really fine traces and you come a guts up because you weren't holding your tongue at the right angle. Now what you want to do is not scrape along like this, but just push down and you want to get right in against the plastic package in there. You don't want to be doing it out here because then you could put leverage on the pins and easily lift the pads. As I said, not a high angle like that, but a low angle and use a sharp exacto knife.

This one actually seems really tough. I haven't Oh, there we go. There we go and just just had to get the initial bit. This is not.

This is tougher than other ones I've done. Wow, yeah, this one's really tough. More force on those pins than I'd like. This one is particularly tough and probably should get a new exacto blade as well.

but I've used this one and it hasn't been a problem with other chips. This one's just got real beefy pins on it. Yeah, normally it'd be much faster and more efficient than this. Trust me.

There we go. That's better. Just needed some slicing action, but you don't. As I said, like you don't want to really get in there and drag it across because that can put lateral force on the on the pins so you can see the exposed copper.

Yeah, this pad on the end lifted up because there's no trace on there. That's a problem with those pads, but it doesn't matter because there's no trace on it. They're not as strong as the ones that actually have the copper traces coming out of them. There you go.

You can see how we've sliced through those pins. I made a meal of the other side. But anyway, let's flip it over. So let's now try and heat this sucker up.

Come on. fall off. There we go. Just dropped off Ta-da There it is.

No worries. and there is our leftover pins. Now we can just easily, uh, get those pins off with our large tip. Just uh, scrape it across and Bob's your uncle.

Bingo There we go. Got a whole bunch the advantage of having a monster tip. You could almost do an entire chip at once. There you go.

Yeah, you don't want to scrape across the pads like that because as I said, that applies lateral force and the adhesive can come a gutsa because there you're heating it up. And depending upon the quality of your Pcb, you can lift pads. If that's not clean enough already, you could actually go in there with some solder wick. There we go clean those.

Now that sweet as so, let's just wick that. I think that's all pretty much gone. So there you go. Beautiful, isn't it? That's how you get a uh power pad chip off like that with putting minimal amount of heat on the pads.

although we did lift one. so I've come a gutser, but you know, can't be helped. But often this actually can be a better technique than. uh, just trying to like heat up the whole thing and you'll notice that we didn't heat up any of the other components around here practically at all.

Um, so nothing else is going to get, uh, collateral damage. There she is as the power pad. Well, the electrons are going to fall out. It's upside down Sta 350 Bw genuine and this one.

You can see that we don't have a solar mass between individual pads, so could get like a bridge short across there. But you know? yeah, you can clean that up easily. All right, solder paste time. I'll do it with solder paste because uh, you're you know? to get down onto that thermal pad.

you can't put solder on there and reflow it. but it's like lumpy. it's you've got to put it on first and then it's lumpy before it melts. It's kind of like a chicken and egg thing.

Oh, a bit too much there. Wow. Oh, that paste wasn't exactly lined up. That looks alright.

Now what I'm going to do now is this is tricky to get on camera. Now here's of course where we've got two options. again. I can try and get my iron on the pad on the bottom uh and try and reflow it that way.

whether or not the pins will actually, um, then melt as well don't know, depends on the thermals of the whole thing. Um, or we can hot air reflow uh from the top, but that's kind of like a desolder in it, so I don't know. I I think I might give it a go with the iron. There we go.

Our pins are melting now. Pins are melting. Are they? Yep, there we go. Pads are melting.

So that that did get through to our pins, so maybe we could have just instead of cutting. You can see though, the problem here is that it hasn't properly self-aligned because the surface tension of the power pad is pulling that upwards. Yeah, that that didn't work. So maybe somebody didn't design the thermal pad properly on the layout? Perhaps I don't know.

There we go. Okay, way too much solder, so we'll fix. We'll wick that away. It's no problem.

So yeah, don't be afraid that you've shorted your pins. It's not a problem. Yeah, the alignment's not great, is it? It's really pulled that chip over. It's hard to get in here though.

I've got these vertical components are in the way, but unfortunately. um, those pins are askew, but they're not touching. They're still soldered, no worries. So that was.

There's still a short there, so I made a bit of a meal of this one. This is not something I do every day. You do this every day. Yeah, get pretty used to it.

There you go. Sucked him. It is off kilter, but I didn't want to. Um, go in there and and force the issue and try and make it come back.

Those ones aren't touching so it'll still be good. It just looks a bit how you're doing, but um, it'll work. A treat. All right.

Well, let's switch it back on. Uh, before you do that, of course, you measure all the rails. make sure that they're not shorted anything like that and done all that. So we're looking for the lead up in there.

Nothing yet. It's on. It's on. It's staying on.

No more blinky blink like I'm gonna presume, like it's some sort of, uh, that's measuring. uh, about Nine Meg. By the way. it's fine.

Win a winner chicken dinner. Um, that works in absolute no tree so it doesn't sound as good as my focals. I still like my vocals, but yeah, that's repaired. That is a very nice repair it.

that was an overly long video. you know, in real terms? Um, this was probably only like 30 minutes of repair work. Uh, you know it takes longer to shoot video and stuff like that. So in the end it was just the power amp chip.

But the interesting part is the journey actually getting there and actually troubleshooting and narrowing it down to pretty much guaranteeing that it was that. we measured it like all the signals going in. everything was fine. We measured that, there were no shorts on the outputs and things like external components causing any issues.

We measured all the uh, power supplies and everything's just hunky-dory Uh, we looked at the data sheet and it had this thing where it actually reported could actually read back fault registers so that made sense with the flashing Led as an undocumented uh fault code. I could not find that anywhere on the interwebs. I guess here it is now and you know it would have been nice if they like told you that in the manual. Um, I could have just went yep and just replaced the power chip.

No worries, it's simple repair in the end, but I liked the process actually. getting to this and I sort of documented more here than what I would normally do like. Some of these things take you know, five seconds, ten seconds to check these sort of things, but I turn them into five or ten minutes worth of dialogue. So anyway, I hope you enjoyed that and found it useful.

If you did, please give it a big a thumbs up. And as always, discuss down below, let me know if you like repair videos because when I got these speakers, I had no idea that they had this digital, uh, Dsp rubbish in there. I thought you know it'd just be like an analog thing, a power supply fault, something like that. So I you know, not always, we get this lucky with an interesting sort of thing where we have to deep dive down.

And by the way, our my Patreon supporters would have already seen this video like or at least up until replacing the chip like a week ago. Especially ones like this where I'm waiting for parts or something like that, I'll just upload a pre-edit uh, to Patreon supporters, subscribe Star Supporters Forum supporters. Yeah, if you want to see that sort of stuff, I don't like plug my Patron at all and my other subscriber stuff, but yeah, I occasionally post uh, stuff like that. So anyway, I hope you enjoyed that as much as I did.

That was, uh, fascinating. And yes, by the way, um, these speakers. They have the hiss in them. Probably the worst I've heard.

They're supposedly really good speakers. Anyway, I've done a video on our tweeter hiss. It's just spoiler alert. It's the inherent noise floor of the amplifiers and the massive dynamic range and efficiency of the speakers in these things.

So anyway, hope you enjoyed that. Catch you next time you.