Hi. This should be an interesting little video. It involves my new Eevblog BM 235 multimeter and this is something I Didn't realize, but people who have bought this a couple of reports started to flow in that it actually produces a high pitch whine when you turn the backlight on like that, there's a very faint if you put your ear like some people can like. hear it from, you know, like a standing distance away.

Other people have to like put it right up to their ear to hear the thing, but it's producing a high-pitched tone now. I Didn't actually notice this when I evaluated the original unit and ordered these things, but sure enough, it is on the original one as well. but I've got to hold this transmitted right up to my ear to sort of do that so they see and I've tested various Um I had like, you know it still had twenty or something here and I tested various ones and this seems to be quite a significant difference between units. Some are louder than others and there seems to be a slightly different frequency between units as well.

So I thought we'd just do an investigation, but it's already been done that by some forum members and people on the Eevblog IRC channel. Yes, there is such a thing. Apparently there's people who hang out on the Eevblog IRC channel and they discovered the problem in this thing and so we do know the solution. but hey, I thought we'd just take a look at it because it's interesting and it's something I've mentioned way back in episode number 33 Part 2: back when the YouTube had 10-minute time limits have to do two parts.

Hmm. now I'll do my best to try and get this sound on the microphone here. It's incredibly difficult I've tried different microphones and things like that, but I'll give it a bash here we go and that if I speak very softly in a whisper I've got my game turned all the way up and maybe you can hear it. Shh So that's the back line on and I'll turn it off.

that's now often, should be gone and back on. So there you have it. You clearly heard that tone there. It's around about five kilohertz or so and it's clearly coming on when you turn the backlight on it.

Otherwise, it's a completely silent, so it's something to do with the backlight inverter circuit that's actually driving the LEDs in here because this thing is only powered from two double-a batteries, right? So they're 1.5 volts maximum at 3 volts maximum. So the white LEDs in here obviously need at least that to work. So when the batteries drop in, haven't got enough voltage. so there needs to be some sort of boost converter in here to actually drive these.

LEDs and that's what the problem is. So I won't leave you hanging. Yes, it is the backlight that circuitry in here. but if we have a look at inside this thing, then when you hear something like this, there are three typical reasons why you would hear some sort of wine or tone coming from an instrument, but all of them involve some sort of electromechanical process that you know generate into vibe.

Something needs to vibrate to generate the sound. The audible sound that you're hearing, and of course, the prime culprit might be look, we have a little piezo ceramic buzzer inside this thing. It's designed to generate sound. Is that that? perhaps? So you'd naturally think of that, but that's fairly easy to rule out.

Now the second one is in inductors. Inductors can actually their coils of wire and at the right frequency, they can actually vibrate and do things like that. Well, the only inductor in here is this puppy on the bottom, but in and he looks like, oh, look, here's the backlight. the springs for the backlight down here.

Here's the backlight board with the LEDs on it down here and you might think oh, it's going over to the inductor, but if you actually follow the traces, it's not I Believe that inductor I Don't have the schematic for this, but I believe the inductor is actually part of the electric field detection circuit, so it's got nothing to do with the backlight at all. now. Another common one is that trance formers. Of course they've got laminations of material in there, and if they're not very tightly mechanically bound, they can actually vibrate.

You may have heard that classic. you know, 50 60 Hertz transformer hum coming from an instrument. Well, it's clearly not that we don't have any transformers in here. They're not whining, you know.

especially like a switch mode DC to DC transformer which operates at higher frequencies, which can be in the audible range. By the way, if there are, you know that lower frequency usually DC These converters are a higher, you know frequency in the you know, in the non audible range ia greater than 20 kilohertz. you know, maybe a couple of hundred kilohertz, even in the megahertz range. So generally you're not going to get them from DC to DC converters.

Although I have heard them happening. some that operate down in the audible range reductions you can see there's no traditional DC to DC converter in here. No I transform anything like that. so it's not that.

So the last remaining one. you have to actually go back to our episode number 33, part 2 as I mentioned that before way back in the old lab when I did a tutorial on capacitors and I mentioned ceramic capacitors and in particular, multi-layer ceramic capacitors. As you've no doubt you know, little Oh six, oh, three ceramic capacitors and stuff like that that you typically see in all modern electronics. Well, these can actually be microphonic.

They can actually pick up sounds, So if you're why I'm talking now. and trust me, these little multi-layer ceramic capacitors in here will actually be picking up that sound. And due to micro phonics, they will actually be generating. And my new, it might be micro volts, but they'll be generating minut voltages across the capacitor And that comes about because of the name of the things.

It's a dead giveaway multi-layer ceramic capacitors. And they're constructed using as a name implies, multiple layers in there. and there are ceramics. So they actually become piezo ceramic transducers just like this little puppy up here.

just like those transducers you used to in. Also, you seen all sorts of products. those little flat piezo ceramic transducers. Exactly the same things going on, except you've got multiple layers inside these multi-layer ceramic capacitors.

So not only are they microphonic and they can pick up sounds, it works in the reverse as well. They actually exhibit the piezoelectric effect just like a piezo ceramic transducer up here. If you apply a voltage to them an AC voltage in the audible range, they may actually vibrate and emit a tone. and the PCB can actually help amplify that and other stuff.

But yeah, these things can actually be little tiny capacitors in here can generate sound. Amazing. So given that this our buzzer is under a software control comes directly from one of the pins. and I don't think it's that.

given that we don't have any inductors in there in the circuit. I Don't think it's that we haven't got any transformers in there now. it's not that what's left these multi-layer ceramic capacitors. Bingo! Let's go to the datasheet, as you should have done first and you could have would have been probably obvious what the culprit was.

So here we are at the datasheet. The Hy 26 1 3 C is the one we have and Bingo right here it tells you. Built-in Charge Pump A LED Backlight built-in Charge pump. So if we go down, we should be able to find this I Hope you can read your Chinese but yeah, internal, we can't see it.

Is there a charge pump in there? somewhere? There's got to be. It'll show you somewhere anyway. If we keep going all the way with LBJ down here, then wait for it. Wait for it.

Wait for it all. Timing diagrams: Fantastic. Bingo Here it is. There's our lead driver, charge pup, engines, your typical capacitor charge pump, and you can see that it's actually got a current sense resistor down here, so it can actually control the current going through the LED So it's going to be a constant current generator operating at some frequency.

That frequency We don't know. it's not in the datasheet at all. at least. Well, if you can't read Chinese there it is.

Is it 15 milliamps? Is that the maximum? Obviously, it's going to probably the maximum. It's going to depend on the resistor value down here, but we can see that we have this 10 micro farad Well, assuming that they've followed The application. Note here: there's going to be a 10 micro farad output smoothing cap and there's going to be a 1 micro farad charge pump cap in here. So there it is.

There's the culprit right there. see 44 10 micro farad multi-layer ceramic capacitor and it's probably one of those dodgy you know why 5u dielectric material or something like that? You know, the ones that are really horrible have really horrible thermal and electrical characteristics, but they're great for bypassing applications and you know stuff like that. they're just fine for it. but they're a horrible dielectric and they can in theory be more susceptible to this sort of problem.

And by the way, this problem only happens to what Class 2 ceramics I Don't believe there's a single case of it ever happening with class 1 NPO type ceramics, those zero temperature coefficient ones that are much lower value. So what we want to do is actually get in here and prove this thing and see what happens. See if we can measure some stuff on here. But as you can see, the battery compartment has these two Springs here and they contact these little pads down here and it's really annoying.

I'd have to like solder some you know contacts on there because if I put the go, put the board back in where we can't probe anything can we? It's really annoying. Maybe like I could solder some wires on and then bring it back out for example, and then screw the board back in and bring it back out. That's one way to do it and another way to do it. which is what some of the guys on the forum have done is just physically remove the screen like that, whack the board back in and back in there we go.

and then you can access and probe stuff through the front here. But of course we're disconnecting the load from this puppy. We're disconnecting the Led. so I think I'll do it the other way.

I'll put the screen back in Oh solder some wires on to some various places and then we can not bring that back out. All right. What? I've done the solder for little wires on here we've got and and the our main output 10 microfarad filter cap which is the culprit here. and then we've got the a charge pump capacitor so both wires coming out there I've just been time around like that I could actually try and bring it out the optional Led hole here.

um, optional. It's not an option on this model, but there's actually curiously, there's a footprint down there. so maybe I should install an LED on there and see if it. uh, you know, see if it does anything.

Anyway, uh, maybe we can probe that. mm-hmm Anyway, I'm going to set that on there and it is a little bit tricky and we can actually get eventually get contact and pair it up. Beauty. And here we go: I'm probing directly across the 10 microfarad output capacitor.

that C-44 That's the output field of capacitor. It's a constant current drive boosted constant current drive and I've got it set to our DC coupling here. and my ground points right down there. 500 millivolts per division, so 1 volt, 2 volts, 3 volts.

There we go. So it's around about the 3 volt level, but we can AC couple that and bring that back up here. Bingo! Now if we have a look at the signal here, you'll notice that not only does it have this interesting looking little bump in there like that. Ok, but it's jumping around like a jackrabbit.

Okay, and if we stop it, you'll notice that there's look. there is another period inside there. It's sort of oscillating like that as well. and you also notice it's not a fixed frequency either.

You can actually see the difference between there that peak and that peak is shorter than the time period between that one and that one. and you can actually hear this. Um, sorry I can't mix it directly with the well actually I probably can mix it with the microphone. Hmm, but you can see it jumps between you saw.

Ah, there we go. Almost perfect, right? It jumps between periods of almost perfect and this is touch sensitive. Try I was getting it before. trust me.

Now it's not cooperating. the white Coat syndrome, but you can I can actually get it where it is almost a perfect tone. okay, and you can actually hear it as well. I'll try and mix in the audio in a second if I can get this damn thing to cooperate.

but it is sensitive to think all sorts of. Anakin What There it is there. it is. Bang.

You saw it right. It was almost perfect. and if you got your ear up to it, you can actually hear a perfect time. Okay, I'm whispering again.

hopefully you can hear this. I'm trying to mix in the audio I'm using my external wireless mic and I've really got to turn the gain right up. Here we go. That's perfect time and hopefully you can hear it glitching.

so as you can see, it's jumping around like a jackrabbit there. I had that smoothing mode on before by the way. I've just turned it to refresh update mode. So yeah, that charge pump um, seems to be jumping around like a jackrabbit like it's got some sort of maybe pulse skipping mode or something like that in it.

So yeah, it's you know, not a pure tone at all. and it just it varies like all the time. just like sitting like randomly and tense seems to also very like just sitting there. It'll vary and it also seems to vary when I physically touch and you know play around with the unit as well.

And if we have a look at the frequency of this thing, um, you know the frequency counter up here I don't know what it's detecting it. Anyway, it's still updating because it's a hardware out frequency counter in the background. But I've got my cursors here and you can see delta. T Here is a 3.86 kilohertz.

There we go. and if we turn the cursor over to that shorter period there, we're looking at five point One Two. so it's going to, you know, vary and jump between those two frequencies. That's why it often sounds very muddied.

It's not a pure tone at all, and if we have a look at the waveform on one side of the charge pump capacitor, you can see that. or you can see it's switching where it actually switches on here. So yeah, that's you know it's doing. Yeah, it's pretty consistent there actually.

But as I said, it varies and the chain in the tone changes when I actually disconnect the probe. so see if we can see a change up here when I disconnect this. Yep, see I changed it the capacitance of the probe and there's the other line there. Not a huge amount of difference, but you can certainly see if I disconnect that probe I will upset that.

There we go. Upset the applecart. ah doesn't like it, but it's actually more stable when I actually load it down with the probe. I'm using the X One probe here.

I'm not using X 10. If I switch at the X 10 then it's still loaded it down somewhat. There we go and check out the high frequency ringing on that. Look at that we and if we go over here, we'll see it here as well.

Love it! You that beautiful? Okay, everyone wants to see the money shot which is the microphones. I've got my wireless mic the output here I've got like max gain so you know it's going to be pretty horrible. Channel to the blue waveform here is obviously my voice. Check it out.

there we go. That's my voice and I haven't turned the backlight on so we'll turn the backlight on, but this is the background level. Okay, so let me show you that bingo correlation time. Now at first it may just look like crap, right? and it doesn't correlate at all.

I mean look, this peak here doesn't line up with this one, but uh-huh. Watch this. This is the acoustic delay. If you shift this over to here and you shift this one over by the same amount, they line up so have just got some delay there.

That's all tiny. a little bit of delay, but it basically this here correlates with this here. this one correlates with that and so on. This one correlates with that.

Bingo. So it's really difficult to actually pick this up. I'm sorry about that, but we can at least see the correlation there. And of course, if I switch the backlight off here, we go.

There we go. We've just got the background and you can see that there's no high-frequency stuff on there at all. Ah, this is better. I've increased the output gain on the microphone here and this is what I'm getting here we go.

This is much better and you can have a look at. See those Peaks look at that. those bottom Peaks precisely correlate with a fixed amount of delay. The delay is the acoustic delay.

So um, look I mean there's you Know it's not an exact match for the for the voltage waveform on the capacitor, but this has to do with although piezo ceramic nature of the thing and how it actually produces sound. But as you can see, the frequency correlates and that is what matters. That's what's producing the sound there. So that pain in the arse.

10 microfarad multi-layer ceramic capacitor gotcha. And if I start that again and then turn off the back light that's a normal background. It's just picking up the regular background hum and other crap here in the lab, but none of that 5 kilohertz frequency content. so that's all very interesting.

But what is the fix for this thing? Well, the fix is to change that capacitor. Now this could have happened between our various batches as well depending on what 10 microfarad cap if they didn't specifically specify in manufacturing in the Bill of Materials a specific part number and model of capacitor always from the same manufacturer. which is, you know, not all that untypical arm. When you're just talking about a Joe Bloggs 10 micro farad bypass cap like this, you may think, well, you know what's the big deal.

It's a 10. It's value doesn't exactly matter. You know it can be half that. It can be double it.

You know it's not a huge big deal. it's just for a backlight here. so there you know. Wouldn't surprise me if they haven't specified that so the purchasing people can just go and chew.

You know, I'm pretty much free to choose if it's not specified, free to choose any capacitor they can get from anywhere. As long as it's you know, exactly the same size, the same voltage rating. it's 10 microfarads, blah blah blah. But this is a problem that is most definitely going to vary between manufacture of capacitor, between model, between dielectric, between different sub physical sizes be at Oh 603, oh La, 5, 1206, etc.

What sort of our current that we're actually what? it's filtering things like that. it's kind of very a lot. So how can we fix this thing? Well, we can simply try and change the cap either to our another 10 micro farad, but it's a different brand. different dielectric.

whatever. I don't know. I'll just find one I've got lying around here. Replace it, see if it goes away.

And here's the first 10 micro farad cap I found lying around 25 volts. It's an X5, our dielectrics and on this Y5 you rubbish or anything like that. but I don't know what is actually inside this thing. and yes, the X5 ours.

You know your X7 ours. They can have exactly the same piezoelectric effect as well. So anyway, I like what are these in and see what happens. And here's a close-up shot of the cap I Took out there just D Soldered that puppy.

Look for 10 microfarads that is very, very thin. It's not thick like this X5 are. So I think it probably most likely is a y5 u dielectric because that's the only way to you need that. A better dielectric better in terms of giving you greater capacitance for size anyway, but much poorer performance so it's most likely one of those you know.

Wi-Fi View dielectrics because look at the physical size of a 10 microfarads. but I don't know the voltage of it as well. This arm X5r I've got here is a 25 volt one. So yeah, it's a it's a betel.

technically a bit better dielectric and it's higher voltage, hence why it's bigger. So I'd be absolutely stunned if this thing produced the same frequency. Anyway, if it does produce a tone I Don't think it will I Reckon there's an 80% chance I'll probably fix this thing by working in this cap. Let's see.

So did it fix it? Let's find out Oh can Still, Here's something. It's not the same. it's not even close to being the same amplitude. Ah, it's really I've got to put it right up right up to my ear to hear that.

but it still something is still there. So ever it's the new 10 microfarad cap. or maybe it might be that charge pump cap doing a bit as well. So hmm, maybe change that one.

- All right. I tried replacing the charge pump cap. that one microfarad one is whacked in another crappy old the Y5v I don't even know where it came from. but anyway, um, let's have a look.

tongue at the right angle still there. but jeez, it's so low like you wouldn't even bother worrying about that. It's just like that is ridiculously low. I mean I don't even know if I'll be able to capture that on microphone.

That's what I'm getting now at the backlight on sorry I don't have any yacht sync signal anything to trigger from basically because I've disconnected the wires. but I turn the backlight off and now on. There's something there, but it's barely picking it up and if there is something there, well, we're talking 3.1 kilohertz now. So yeah, it's changed frequency which is exactly what you'd expect because you're putting a different cap in there.

so it is. Looks like it is the new 10 microfarad cap I put in there. but you can see like it. You basically cannot hear the thing anymore.

So essentially, problem fixed. It's just a matter of choosing the right cap. Some are much more susceptible than others. It's going to depend on all sorts of parameters.

which, yeah, it's almost. you know, suck it and see. really. So there you go I Hope you found that interesting.

Thanks to the people on the forum and people who've bought this meter and actually reported this thing because it's something that I Didn't notice it was so low. but once it was reported to me, you can kinda hear it. But now I've essentially fixed this particular unit I've reported a deployment. They're going to research it and get back to us.

No doubt they'll fix the issue. So that's an interesting little practical example of piezoelectric effect in multi-layer ceramic capacitors. Watch out for it. Can be real trap for young player.

You can really come a gutter I'm Ultimately though, it's it's an issue with the chipset. The frequency is down in the audible range and that's just dumb. If they switched it at 20 kilohertz or 30 kilohertz or something like that, you know, instead of down at 5 kilohertz, no one would have noticed the thing. You can go piezoelectric.

Your dog might hear it or something like that, but no your cat. but that's about it. So yeah, it's just a poor choice of frequency. The frequency might be dependent upon the capacitance though.

I haven't looked at it. There's no details in the datasheet for that sort of things. I'll link in the datasheet for the LCD controller for this thing down below which has the building lead charge bum. So there you go.

Hope you found that interesting. If you did, please give it a big thumbs up. Catch you next time! Ah, by the way, while I was playing around with this 8 and trying to get the microphone to pick this up, various different types of mics I actually thought I'd try my Stanford Research filter um what is it the SR something-or-other I don't know s FS 650 programmable filter you've seen in a previous video I thought oh, if I can filter out all the other crap, I'll get a nicer waveform and stuff like that that you know if I can like put in a bandpass filter it you know you know 3 to 6 kilohertz or something then I can filter out all the other crap I'll get a nicer signal. turns out, the fan in this thing as I noted in the previous video I did on this I think it's just so ridiculously loud it just ruined everything.

So yeah, time to upgrade the fan in this thing. Bloody. Teske with noisy fans Hate! um you.