Had this happen with a pocket FM/TV sound radio many years ago. Started with intermittent stereo which eventually failed totally. Turns out the the led was intermittent because it split in half due to the legs being bent when it was installed. This stress the bond wire until it broke. The intermittent stereo had me looking at a faulty chip and first but when I fitted a new led it worked 100%. The led was actually regulating the stereo detection voltage. When the led went open the voltage rose too high and muted the decoder. I actually read of a dead led causing no stereo shortly after that. Think it was in EA mag.
EEVblog, I remember it, too - the most annoying thing was the absurd brightness levels. I had to put black tape over a few of them so that I could sleep at night!
i had a router with blue leds that would make my bedroom bright enough to read a book... whoever thought that puting retina-burning blue LEDs in everything was a good idea has a special place reserved in electronics hell.
Blame the industrial designers. One of my favourite examples of this stupidity was an HP "entertainment" laptop which had one of those horrible glossy screens (protip, I don't care in the slightest what the thing looks like when it's turned off, but I'd like to be able to see my screen in any sort of well lit environment) and blue indicator LEDs across the top of the keyboard where they glared onto the screen making the bottom third of the display unwatchable.
My favorite LED failure was in a piece of machinery from probably 15-20 years ago that I had to deal with recently. Long story short, it has an RGB LED to indicate ready/warning/fault. I am not sure if it was something with the bond wires degrading/shorting or if there were some chemical breakdowns going on, but the LED lit up Yellow, Green, and Yellow. There went a couple of days of working on the entirely wrong part of the machine.
Actually that's a extremely common fault for LED's. I see indicator LED's going haywire like that all the time. On one second, off the next and then back on. In my experience it's LED's in "cheaper" electronic devices. High end products tend to use better quality LED's.
Many seem to. I think the bond wires break loose and make for a poor connection. Feed enough juice into them and they come on. Leave them on long enough and as they heat up thermal expansion seems to push the poor connection together and they will stay on. Have seen this in 7 segment channel displays on CB radios to.
Yes, I've seen similar defects in various modern LEDs. They can flicker at low current then stabilise at higher current. I suspect a semiconductor defect rather than bondwires.
I had the opposite happen in some GU10 LEDs a few years back, LED's tested ok at low currents but when 100mA + passes through them to light up the room something breaks and they turned into a dim flicker.
I've seen it a lot with cyan leds like those used in traffic lights. These were damaged by static discharge from not properly handling them. When looking at it under a loupe I see a black fissure-like damage on the die. When I power it it's like a short, but when I increase the current very briefly to 50~100mA I "burn through" the damage and the led lights up again at 10mA but not as bright as a new one. Sometimes varying the current brings back the short circuit or only partially and the led flickers. With the curve tracer I see a curve like resistor in parallel with a diode. A led that doesn't light up at all has a curve so steep that more than 50mA flows before the threshold voltage is reached. When I "burn through" the damage on a led, using the curve tracer, the slope immediately drops and becomes very unstable.
Yeah! I've seen something like that in the past. It's truly hilarious when that happens. Actually, it seems to be a quite common failure mode, if you pass too much current through the poor thing for an extended period.
At my company we used to have strange problems with blue LEDs going intermittent or off alltogether. In the end, the actual problem in production seems to have been ESD-induced. However, while we were investigating, we found that a moderate overcurrent through the LED for months would result in just the kind of behavior you show here (flickering which did not seem correlated with any mechanical shock or vibration). If I recall correctly, it was caused by driving maybe 30 or 40mA through a max 20mA-rated LED for about 6 to 10 months which did it.
Similar problem with white LEDs I used for lighting a portable astronomy dome. PVC pipe and nylon construction dome causes static charges. This was a basic wire-up before adding clamping diodes and capacitors to each LED. These white LEDs started to flicker on and off and turn green in between. The red set were unaffected. No problem since adding the clamping diodes and capacitors since. It is an odd fault with the blue and white LEDs due to the multi-layer construction of the dies.
Speach quality is never great when you have a two-driver crossover speaker system. The crossover point typically lies in the middle of the speech band. Try speech on a high quality 8" broadband driver like the Sica 1100, it blew me away how good it sounded.
I have the exact same speakers and they exhibit the exact same fault! I've had them for ~8 years and they've had this fault for at least 3-4 years.. I've been meaning to crack them open and see what's going on but you know, other priorities.. You've saved me quite some time with this vid. :)
Very common, even on significantly more expensive studio monitors that are bookshelf-sized, to tweak the LF response depending on whether they are against the wall or freestanding on stands.
No, a switch is more common. Check the Genelec 8010A, probably the all-time most popular active near field, two switches for 'bass tilt'. This is a speaker that costs about 6 times the price of these Alesis.
I'm saying having a switch to tailor the LF response is perfectly normal and not laughable. However it's not a very interesting topic so I see no reason to debate it further!
I have a cheap 12V led strip where one of the 3-led series strings behaves the same way. At low brightness, one segment is visibly dimmer than the others. It also sometimes flickers. If I happen to notice it, I just crank up the brightness until the pesky segment snaps up to the same brightness as all the others. Then it can usually be adjusted to any brightness without issue until it gets turned off for a while.
In my opinion one of the LEDs pooped itself and developped a significant leakage and started to behave like if it had the resistor parallel to it. That resistor formed a divider with the LED dropper resistor and voila! The voltage on the leds was unsufficient to light up the blue one. That leakage resistance is not that uncommon in LEDs, especially the ones that work quite hard in the lamps but I've came across such behaviour even in the HDD activity led in the PC case.
Weird coincidence: I just finished troubleshooting and fixing a musician's DI box with a similar problem. The DI box would randomly cut out, causing very loud "thumps" through the PA. The blue power LED would also come and go, even when the box was working OK. Long story short, all of the power for the unit was drawn through the blue LED (it only used about 1mA) and that LED was misbehaving. Replace the LED...all is well. I kept the faulty LED to experiment with and it really does behave weirdly. In experiments so far, it's forward voltage drop will vary between a perfectly normal 2.6V and 12V at 1-2mA, and it will "jump" between states quite abruptly. Strange behaviour.
I was trying to fix an intermittent LCD display fault on a Pure DAB stereo. I found that I could change the fault by tapping the board near the display, so I resoldered all the joints around it, but nothing changed. Eventually it turned out to be a reversed smd capacitor.
if you work much with White leds you see this blinking fading and random spiking all days… happens only with White and blue ones, because a White led is mostly an blue led with some phospor in it (yellow blob the crystal)
I don't like for the fact that they're using an XLR connector to connect the second monitor to the amp in the first one. XLR is almost always used as a line-level (or mic level) connector, and this makes it all too easy for someone to accidentally connect the output of a power amp into a preamp, be it the front-end of a power amp or one of the inputs to the console. SpeakON connectors are reasonably priced and are designed for this application.
So the transistor at 8:42 that bridges the LEDs is really driven in a reversed connection (C becomes E and vice versa) if the opamp output is negative and the system is switched off? This condition is not specified in the datasheet (I think one effect should be a significantly lower gain) and seems like quite an unusual design choice for a standard mass market design. Are you sure this is an npn and not a jfet and it is really connected like shown or might there have been a malfunction in the DaveCAD reversing feature?
temperature changes the contacts inside the LED so that's typical for cheap LED lamps all over... "auto" blinking appears a lot of time when overheated and stuff....
I have these same monitors, and had the same LED issue. however they have recently failed. all i get is a loud noise and the clipping light comes on. Poked around inside and looks all fine, no blown caps (that i can see) really upset and not sure what to do :(
I had a virtually identical issue with a red LED on an old MIDI interface - it was causing all sorts of random behaviour across the multiplexed front display. It took hours of troubleshooting and quite a few "repairs" before I finally found the little bastard!
Dave, while you're at it, check the brown~ glue used inside these Alesis units for leakage! It seems to be no better than that black tar stuff, it starts conducting on the part that have seen more heat than the rest.
I had the same issue with LEDs from some club RGB lights. Had no idea why they flickered randomly, no soldering issues, done some PCB tracking, no problems there. So I changed some of them for new ones. My theory is that the manufacturer, in this case not Alesis, but manufacturer of the light emiting thingy, used some off-spec LEDs thrown away. I once got UV leds from ebay and they had obvious faults like bubbles and some of them started to flicker after about 10h+ of usage.
Lot of drivers put (not road legal) LEDs in number plate lights on their car. I see very often those LEDs blinking on switching on and off while driving through holes. First VW Passat with round LED rear lights now have blinking segments. I think it is quite common except you normally don’t expect it on a single indicator LED :-)
I have a pair of Presonus Eirs studio monitor speakers that jsut did 1 year since I bought them,and this just happened to them!The Led lights up when I turn them ON but starts to fade away until lits off,the sound stills good though!!!
This happened to a led I cut the to off a little it went intermittent and did the same thing yours is doing , I did that to make them spread more light in a car door application
my Alesis 520 USB have the same led problem. But this week it started another annoying problem, all the sudden the volume goes to maximum (clippping red light) and cracks out loud for a couple of seconds both woofer/tweeter!! This is damaging those components when that happens, so where is my problem? is it the potenciometer? This crackle also happens when headphones are connected, also damaging it's drivers :( Thanks in advance!
I've had exactly the same fault occur in a DJ twin CD player where the blue back-light on one of the lcd's failed. I did exactly the same as yourself probing the pcb and finding that it worked fine. Being again part of a light diffuser it was too awkward to fix so had to live with no back-light.
Why could they not have used two really simple old school round LEDs? One green for power and a red for clipping. Super simple, ultra reliable, AND, it would have been cheaper.
I detest blue LEDs. They are overused. This sort of intermittent failure has shown up in repair work here. The on/off period can shrink seconds or less. The period tends to be chaotic rather than steadily periodic. Very rare for red, yellow, greeen or orange LEDs. But not so rare for blue.
One time I had one of the driver transistors fail for the stepper motor on a scanner. It used a transistor array IC which had some unused channels, so I just added a few jumpers on the IC. Some time after another channel blew, and I think there was still a free transistor to wire that one to as well.
Always been a fan of your work Dave, you really have an enthusiasm that cannot be matched. I enjoy seeing how excited you get when things happen (good/bad/weird) this video being a perfect example, Who would have thought the bloody LED would be the culprit and in such a dicky way!! Keep em coming, and ill keep watching and learning, thanks a lot from a friendly pom!
Had many LED's do this, they are a pain in the proverbial ! Just recently I had an LED Side light bulb fed with 12v DC which now strobes but if you up the voltage on the bench (Effectively increasing the current) it comes on !
im seeing that some jr tech assuming all LED's draw the same current and not realizing that they may not , or picked up the wrong LED entirely that do not match the spec/draw of the red LED calculations were initially done for that circuit. the circuit is simple, but component specs still have to be looked at.
Nice looking unit. If it were me, a bum LED wouldn't bother me as long as the speakers worked fine, but having that clipping light would make it worth it to fix. Good vid and subscribed!
Hey Dave, that fault is not as rare as you think. Actually, my first guess was a faulty LED. I have seen a lot of them failing exactly like yours. Seems to be a problem of cheap LED production and should be well known by the manufactures because of the frequency with which this error occurs.
Are these speakers older than 10 years or so? I ask because I've seen older blue leds fail in a very similar fashion. I understand that, due to the relative difficulty in manufacturing blue leds, earlier-genration blue led chips are very vulnerable to both electrical and physical stress.
@0:51 You're not using the "crappy soundcard inside the computer" when you use the USB interface, you're using the crappy soundcard inside the monitors. ;)
Put BigClive's Argon/Mercury tube driver in your speaker box. It will drive that led. BigClive Autopsy of faulty LED vid @14:53 WTF, youtube/firefox doesn't allow pasting to yt comment?
Dave, a quick question since ur the handy man - I actually own the 320 ones - prolly for the last 7 years or so. couple of days ago i found out even when turning the knob to OFF the blue LED light stays on. No sound unless I turn it on and raise the volume. Any idea wot can cause that? hope I can DIY it. Cheers
didnt that transistor you took out test dead short one way? I think if they test dead short they are junk.. I have a Denon AVR that lost its side surround (7.1) and zone 2 channel becauise a couple transistors on the output dead shorted causing it to shut off and blink red. However with that board disconnected you can use it as a 5.1 just fine.
I've had a repair business for the last 11 years and I'm burnt out on all the rabbit holes there are. Only way to keep going is to hire technicians and burn them out on a continual basis.
Hi Dave, I quite like them, thanks for asking. They're the older ones, not the V2. I use them with the bass boost off and they're good as close up monitors as designed. I guess the bass-heads would want more bottom end though. They're plenty loud for monitor use with basic video editing and general TV/Video viewing from 10ft away. Getting on in years but they sound crisp with nothing to complain about. www.dropbox.com/s/p2j1et2u6ecvezh/Speakers.jpg?dl=0 I've tried the Rokit 5's but they're more than I need for in the house.
In use I'm only 2ft or so away; I can touch the drivers "when sitting up and using them." I do sit back and let them go a bit looser when just watching RU-vid, listening to podcasts or just hanging around in the room.
Might not be the blue LED at fault? Could the red LED just be unhappy about being reverse polarized essentially constantly for many years and giving sufficiently high leakage current that it's clamping the voltage to the point the blue led won't turn on anymore. 2v seems in the region of a red LED
Interesting! I had an almost identical problem, also with a powered studio monitor (M-Audio brand). It had two blue leds in series around the volume knob, which are supposed to get brighter when you turn up the volume (there's an extra ganged wiper on the volume pot in series with the leds). After I had it a while one of the leds started to flicker, but only when the volume knob was low. When it flickered out the forward voltage would go down, making the other led in series get brighter! I took it apart to see if I could find the problem and eventually came to the conclusion that the led itself was bad. Weird!
Why not try some spare working LED's and see if it works as expected Dave? Would of at least told you if its the LED"S themselves being awkward or something more silly going on with the supply. It is annoying at the same time amusing when things like that occur.
I have 8 smd LED's (just white)on my Belkin 4 channel KVM, 1 for Video and 1 audio. All 8 started out fine and after a year or so 1 started randomly flashing while it was active, the a couple more started and now all 8 do it. I checked the supply the current resistors the solder joints... it is the LED's, the KVM now sits behind the monitor where the random "disco" flashing lights can not be seen. :)
Perhaps it was becoming self aware as the fractured crystalline lattice of the chip was developing neural pathways like in that star trek episode? You may have destroyed a new species before it could attain its first identity crisis being a power indicator glued into a desktop speaker. Geez Dave.
i have an power amp i got from one some a long time ago that stopped turning on until I replaced the indicator led. The led appears to have been a part of a latching circuit.
Dave: The line about " i do't like dicking around with software.." when it comes to volume control. I've had an idea about starting to manufacture DVD driver size pre-amp/audio isolator that does two things: isolates the audiocircuits from PC to cut off all interference that may come from the PC (i've had lots of troubles previously with ground loops with PCs..) Second function is that user has hardware, old school volume control on the front panel. Something that NEVER fails. No matter what state your PC ends up in, you can always turn up/down and mute audio from the place where it should've been for decades. Does that sound like a feasible product? The electronics is basic stuff, very, very easy. Chassis is easy to manufacture, there is power supply, bing bang bosh: shouldn't be hard. The only question is that would it be any way profitable? There is some demand for it, specially if it allows to use external audio source that is mixed with the PC audio. That is one special need, people who want to listen their iPhones while gaming or browsing the net but want to hear what the PC is doing too. 2ch mixer with headphone amp would be the configuration, with SPDIF input/output (to make the isolation even better and for audiophiles... ;) ).
Interesting idea, but I can see a few problems there. 1. 5-1/4" bays seem to be going out of fashion these days. Some modern cases already don't have any, and I'm talking full-size, not Micro-ITX. Not exactly a trend I'm particularly fond of, mind you - I'm all for good ventilation but at some point it just gets silly. 2. All you can typically access inside the PC is front panel audio, where people would typically use headphones or a headset, no isolation needed. So you would have to include some contraption to get your audio through a slot bracket or something. 3. How do you intend to include SPDIF? That's almost a soundcard in itself? Or do you just want to have Toslink converters for use with the onboard SPDIF connectors that some boards have? I would also advise to a) implement some sort of power-on/off muting as you will be limited to single supply (which also means taking care you get a clean +Vs/2 reference and other single supply design caveats), unless you are willing to go to the trouble of including an inverter to generate roughly -11 V for a -8V regulated supply (many soundcards use 8 V supplies, I assume 9 V could go out of regulation under adverse conditions unless using LDOs) b) have balanced inputs for your output signals so you don't have to deal with internal ground loop issues and can easily use local ground (most outputs tend to range from anywhere between 10 and 200-something ohms, so I'd include 22 ohms between output ground and your "cold" input and call it a day - if the effect of impedance mismatch bothers you, go for a buffered circuit and increase common-mode input impedance by going for a T-type network for input impedance, e.g. 10k+10k + 100k to Vs/2) c) include your own microphone preamplifier with clean bias supply, 'cause the ones in sound cards very often _suck balls_ (technical term), and finding a good consumer (unbalanced) mic amp is not easy. With an NJM2068 at +30 dB (3k3/100R, maybe the latter as a 1k reverse log pot + 10R fixed if you want to get fancy) and 1k + 100µ to clean up the output of your main 78M08 (78L08) reg plus 3k3 bias R per channel and taking care to avoid any ground loops, you would be pretty much king of the hill, sadly. Just about good enough for a 600 ohm dynamic mic, as in 3 dB above thermal noise... (Possibly include a 4PDT toggle for "straight through" in case you get a digital / MEMS mic, though not sure how common PC mic inputs for them are.) d) do power / ground routing carefully. You are likely to have chassis ground available, and I would prefer to use that as a reference, but I'd still include Molex / SATA power ground, with ~10 ohms connecting them, just in case the connection from chassis ground to PSU ground is not good. (With many cases no longer being bare metal inside and on the back, I would not bet on it...) Speaking of mic inputs, one thing that would be handy to see would be a balanced XLR mic input (à la ESP P66 or using an INA) with phantom power, as basic "large diaphragm" condenser mics are getting really cheap these days. That would realistically require a boost reg though, so you can get a full +48V (some passive cleanup included), as there are a fair few mics that are not content with +9-ish V, and I do think a switch and indicator are mandatory. That said, this may have you competing with the likes of a Behringer UM2, and that's not easy...
Has anyone else noticed a bit of flash/flicker in the past few videos? I don't know if I missed a comment or response, but it is quite distracting. It is sort of a slow strobe, about once a second in the background, and reflects on whatever is in the bench. Am I going crazy or has anyone else noticed this?
Attila Asztalos The current wouldn't tell you much since if the other LED would be in reverse breakdown, it would clamp the voltage to less than required to light the blue LED, but the current would remain mostly the same since the voltage changed relatively little.
I expect the surface either of the anode or the cathode inside the LED to be corroded and increased resistance by whatever reason, so the higher current probably helps to pass it. We definitely need an LED Engineers opinion here.
Yeah, something like that. That's what I was alluding too when I mentioned a "point contact diode effect". It's not like there is enough current cause any heating/expansion effect of the contact, so more likely to be some sort of contact surface phenomenon.
