stillwater

I took (what I thought was -- see "Partial Retraction", below) the previously offending (and thus replaced) 2477S apart. Without the case on it is pretty quiet but makes a buzz that is audible, especially with the stethoscope. I was able to confirm that there was no noise or vibration from the piezo electric noisemaker -- it's something on the main circuit board (not the RF daughter board). Probably it's a surface-mounted ceramic capacitor singing -- though I didn't look closely. Here is a reference about such "singing" capacitors and what can be done about them in designing printed circuit boards. (Unfortunately not much help for us) https://sound-au.com/pdf/acoustic_noise_MLCCs_TI.pdf Possibly some goo like the black rubber sold in hardware stores as an alternative to electrical tape would reduce the vibration at the board, or taking steps to decouple it acoustically from the wall box and the wall would help. Anyway it certainly has nothing to do with the ISY. Of course whether this is the same as the seemingly louder noise from the currently installed one I won't know for certain until I replace it with the the newer one. The one mounted in the wall box makes a more rounded tone but I think that is because of various things being vibrated and not the source noise at the device. Edit: Partial Retraction. I see from previous notes that I likely replaced the 2477S a little over a year ago for a completely different reason -- related to problems with the 2477s communications that were ultimately resolved by factory resetting the 2477D in the same box. So I really won't know much for certain about what is causing the noise in the currently installed 2447S until I replace it and test it on the bench. However the similarity of the sounds via the sterhoscope makes me pretty confident that the problem will be a louder version of the sounds I have heard in the other switchlincs. I will report back.

Thanks!

I do not believe that it's the piezo "beeper" that's making the noise, though this is possible. To be sure it would be good to know where it is on the pc board) I could look into removing or cutting into the case and feeling around with an insulated probe to see find the source of the vibration. Also more careful listening shows the one installed in the powder room that I thought was completely quiet actually makes more-or-less the same noise as the new-in-box v.45 2477S.

Further information -- Not sure I am much closer to solving the mystery but here is additional info: I located the v.43 switchlinc (for toilet fan) that I removed from the system more or less a year ago. This was indeed made toward the end of 2014. I hooked it up to a test harness to give it power (no load) and added it to the ISY (after yet another factory reset). With a stethoscope on the case I can hear a very quiet buzz on all insteon traffic through the case. Now this is located in a different place from when it was installed -- it may be on a different phase electrically and also in a different position re RF (in the toilet room the switchlinc has a 2477D right next to it, so that may be blasting RF and powerline signals right at it. I also have a new-in-box v.45 switchlinc manufactured in 2018. I did the same test with it and it makes the same buzzing --though even fainter. Neither seem anywhere near as loud as when the v43 switchlinc was installed in the wallbox. (or the current one). This may be because of mechanical resonance to the wall box or wires jammed into the box. But it seems a greater difference than I would have expected based on that. Based on the noise the switchlinc makes when switched remotely by ISY, it has a relay to control the load -- so I doubt the fact that it was connected (like yours) to an exhaust fan has anything to do with it. (If it did it could only be by feeding RF back into the case? ) Also looking at the circuit board through the clear case there does not seem to be much that could be affecting the resonances... though conceivably it might help to put some foam tape or some other sound deadening material on the case, and not fully tightening the screws attaching it to the wall box, or the screws attaching the faceplate to the device. Of course the less insteon traffic there is on your system the less you will notice the noises.. So reducing retries by fixing Comms problems and disabllng the Query All program (or rescheduling it for when you are unlikely to be in earshot of the switchlinc) may help if you can't eliminate the noise.

PS the noise from this switch is one reason I disabled the 3 am factory query all program. Both of the toilet fan switches are v.43 based on the banner in the administrative console.

I have had what sounds like the same problem for maybe a year. I first noticed it some years after original installation. Factory reset (followed by restore device with testing in between) didn't make a difference. Replacing the 2477s with another one of the same vintage (probably 2014 or 2015) , followed by factory reset, also didn't help. However I have another 2477s (probably also bought from Smarthome in 2014 or 2015 and also installed in 2015 ) in almost the exact same configuration in a powder room that is completely quiet. It switches the same model panasonic fan and the same electrician wired both -- but conceivably it could be something about the fan. I do not think this is the actual device beep but I could be wrong. With a stethoscope it sounds more like a buzzing/fritzing sound than a beep. It does occur whenever there is insteon traffic, not just traffic specifically for or from that device. Maybe there is a coil in the 2477d that is in contact with the case, or something like that. The switch is in a toilet room that is very small and for a long time I thought the noise was coming from the fan and not the switch, but it based on the stethoscope it does seem to be coming from the switch. If I were you I would indeed try replacing the 2477s and see if that makes a difference. I'd be very interested in what you find out. I tried to query the insteon engine for each of the two currently installed 2477s from the Admin Console to see if they are the same type of device -- I believe that in previous years this would give some sort of result on the screen. The query runs (there is a brief blue rectangle and a progress bar that goes by too quickly to read) but I don't see any results on the screen. I have 5.3.4 firmware and 5.3.4 UI.

I assume the answer is no but thought I'd ask this forum anyway. Ideally via software but if reversible by opening devices and e.g. moving a jumper or cutting a wire or trace that could be re-soldered later. I guess it might be different for each device. I have KPLS, switchlincs, switchlinc dimmers, micro on/off, and micro up/down devices. The PLM I could replace with a 2448A7 (via MrWorf's raspberry Pi hack -- I have a PI model A that isn't being used.) THe motivation is that I have Soraa LED bulbs that weirdly interact with the dimmers and insteon signals by flashing bright from dim on Insteon signals. I have dealt with this by installing a choke and (powerline rated) capacitor on each load but was wondering if there was a different approach.

Also -- if the Pyle amplifiers were buzzing in the absence of a chromecast signal, then that might indicate that the HDMI converter is outputting something when it is not supposed to output anything. Using a multimeter to measure its output (with both AC and DC ranges) in the absence of a music being cast to the chromecast would tell you a lot without risking another amp. You could compare the output of the known good one and the possibly bad one. Of course it might take days for the problem to reoccur... Full volume is probably around 1 or 1.5 volts AC. Without any input it should be close to zero AC. (DC was discussed previously). But this is not dispositive because it might require some "exciting" for the unwanted oscillation to occur... Which could explain why it took a few days for the right input to appear. (It's possible some new program material or music type was played that hadn't occurred in the past 2 years and then again with the new amp... I'm clutching at straws here...

If @asbrill has a multimeter he can measure DC volts on the outputs of the HDMI converter --and if not zero he could put a film capacitor in series with the center pin of the rca connector -- preferably in a shielded housing to prevent hum. (value of the capacitor determined by the input impedance of the amplifier which is probably 15k to 25k ohms.. I suppose he could put 1.5 volts from an AAA cell on the input of the amp and see if there is DC on the output but that would only tell you if the the whole amp is DC coupled and not if the input is DC coupled, so probably not worth doing. But it's equally likely the HDMI converter is actually sending out a lot more power than the amplifier is expecting above the audio band (at least for us old people) and this is taking more power than expected and creating the buzzing by subharmonics or by overloading the supply. I actually think this is more likely than DC bias... because I very much doubt these relatively inexpensive amps are DC coupled.

Also, just to be sure, when you get back from your trip, before you do anything else, consider measuring the voltage between ground and each of the four speaker wires. If you have one you could first use a proximity probe to see if there is 120 vac present, and then use a multimeter on both AC and DC ranges. For ground you can use the 3rd ground terminal on an ac outlet or possibly the case of your power strip (if metal). It should be zero (up to a few 10s of millivolts might be noise). But of course this wouldn't tell you much if the problem only appears intermittently... But I am guessing it is indeed the HDMI converter or a combination of the HDMI converter and the wall wart that is the source of an input problem, rather than anything on the speaker side -- but if it were me I'd put safety first. (Again I am relieved that only the last amp died completely...)

Sure. Do you remember whether the buzz was more like a constant hum or more like a staccato bup-bup-bup? At about a minute into this video you can hear what 60 cycle hum sounds like (plus some distortion from the speaker) https://www.youtube.com/watch?v=Vf6Bq-vStLA Also worth looking at the various usb wall warts you are using and ensure that you have a relatively beefy one powering the replacement HDMI converter (and through it the chromecast).

Good luck! I feel better in knowing that the first two amps just buzzed rather than having their power supplies fried also. Was the buzz a constant hum (likely to be 60 cycle noise) or what audio engineers called "motorboating" (more a succession of impulse noises -- see https://en.wikipedia.org/wiki/Motorboating_(electronics) ) Motorboating could be caused by the wall wart for the HDMI converter providing insufficient current. Unless you the amps buzzed without being connected to the HDMI device it is plausible that you threw out 2 perfectly good Pyle amps.

Too bad you don't have the amps that buzzed. For example It would have been good to know if they still buzzed with no input connected, or with their inputs shorted. Or if they worked ok connected to other speakers. In the meantime I assume the HDMI converter is cheaper than an amp. I'd treat the HDMI converter as guilty until proven innocent (and I certainly wouldn't risk the Denon amp to prove the HDMI converter innocent, which it wouldn't really do anyway)

It would seem that if the amplifiers have been replaced and the power supplies for the amplifiers have been replaced and the wall wart for the HDMI converter has been replaced with a known good one then the only remaining potential culprits are the HDMI converter and the chromecast dongle. (As a pure digital device I don't think the Chromecast could be responsible unless it were somehow modulating the 5 volt supply by shorting out or drawing too much current-- though this is is all so strange I can't rule that out completely.) I don't know enough about the internals of these amplifiers but the failure mode would be somehow the HDMI converter is injecting some sort (non audible) oscillation (or DC if the amps are DC coupled) that is causing the amps to draw too much power and blow either an internal fuse or some other power supply component. (Otherwise the red light would be on). It's a stretch. None of these things should kill an amp... let alone three. You would think an op amp would die rather than the power supply. In any case it's odd that it works for a while and then dies. This would seem to rule out some sort of wiring fault resulting for example from non-polarized plugs. The idea that you replaced a failed wall wart with another failed one is also a stretch. You might check that the Wall Wart is rated for enough current to power both the HDMI converter and the Chromecast. The HDMI converter on its own may draw as much power as its wall wart gives and then the additional draw from the Chromecast is too much. You can find power supplies (for example for raspberry pis or for Ipads that proved an amp instead of the 500ma that is nominal for a USB port) . But an underpowered USB supply on its own shouldn't be causing this problem, especially assuming the first one worked for a long time before the recent troubles. And I'm guessing you have the same configuration for other rooms... One more thing -- I don't understand the sequence -- the speakers play normally for some time (days?) and then without any change they buzz loudly (for how long?) and then the amp dies and the buzz stops at the same time? Is that right? I suppose there is another possibility which is that somehow power from a doorbell transformer or thermostat (24 vac) is being transferred to the speaker wires causing the loud buzz, and killing the amp -- but then how does the loud buzz stop (maybe ground path for the power is through the amp, and when the amp dies the buzz stops ? ) But then how does it work for days before the buzz appears? Any chance that it dies when someone rings a doorbell or when your heat does something? Was there any wiring change in the period before these failures started? (I certainly hope 120 volts isn't straying onto the speaker wires from lighting or exhaust fan wiring! This would be an electrocution and fire risk. ) Anyway I'd be very careful around any of this wiring until the mystery is solved. Sorry I don't have any better ideas....

Very possibly the problem is the power supply (wall wart) for the HDMI to RCA converter, or less likely with the HDMI converter or chromecast. Unless -- I assume you didn't reuse the same power brick for the 3 amplifiers... Maybe the power supply leaking high voltage into the HDMI converter. Otherwise I'd be surprised the amps are permanently damaged due to an input condition from a 5 volt powered source. (assuming the HDMI converter is powered from a USB wall wart -- if not maybe it's 12 volts... but still ...). Or maybe it's leaking a lot high frequency from the switching power supply and that is overheating the amps. How do you know the amps are toast? (Just re-read your original post. One or more of the amps could have turned themselves off with overheat protection.) If you haven't tested them separately from that input configuration it's possible they aren't actually destroyed. Even if they aren't destroyed the most likely culprit is still the wall wart for the HDMI converter (failed capacitor). (If the amps are not destroyed what could be happening is the following -- something in the HDMI wall wart or HDMI converter (less likely in chromecast) is failing. You unplug everything or turn off the power strip and these components cool down. You replace the amp. It works again until something warms up and fails and you think the amp has died... repeat.,

Yes, finding the short is necessary. As someone else suggested, putting a 4 ohm high wattage resistor in series with each channel seems sensible until you are sure you have found the short. You could probably get by with 2 8.2 ohm 5w resistors in parallel with each other or extremely conservatively 6 22 ohm 5 w resistors in parallel. You could also (or instead, if you think you've found the short) put a fuse in an in-line holder in series with each speaker channel output. It sounds like maybe you have similar speakers with an amplifier that works. You could measure the typical current with an ammeter and use a fuse with some margin for louder music or transients Or, based on 4 ohm speakers and maybe at most 40 watts, you could estimate a current of 3.2 amps. If the speakers are 8 ohms, then 2.2 amps. But I think 40 watts per channel is way high for a bathroom. I think once I got past seeing that the speakers make sound with the resistors I'd consider putting a 1.5 amp slow blow fuse in series with the speaker output instead of the resistors. If the fuse blows while normal music is playing I'd go to 2 and then 3 amps. This would likely protect the amplifier though I can't guarantee it. There is one other possibility, at least theoretically. I am not familiar with the Pyle amps. In the unlikely event that they are DC coupled, it's possible that whatever you are using as an input source has failed in a way that is putting a DC voltage into the amp and the amp is failing by trying to amplify that and drawing lots of power. But a DC -coupled amp would be very unusual. You could measure the output of your chromecast device (if that is the signal source with a multimeter set to DC volts -- there shouldn't be any or at most a few millivolts. But even if there is in a typical amplifier it is blocked by a capacitor at the input and between stages. (again unless it is DC-coupled, which would be unusual). You could also put a blocking capacitor (Ideally something like a 1 microfarad film capacitor) in series with the center terminal of the RCA connector between the output of each channel of the chromecast device (assuming that is your signal source) and the amplifier. But again I doubt this is the problem. Also if the chromecast device has a digital connection to your amplifier (coax or toslink) then it certainly isn't the problem. (Also, depending on the amplifier rail voltage and the input, if DC were appearing at the speaker output it might well blow your speakers.)

As a replacement amp, consider the rather small Class D Aiyima A07. In the user manual it says it has overload protection, though it says this only indirectly in a FAQ. It is marketed as a 300 W amp but of course this depends on power supply voltage and current capability and I don't think the box has enough cooling for anything like that on a sustained basis. (You probably won't need more than 20 watts for a bathroom anyway) The amp is rated down to 4 ohms (which some amps today aren't). $80 on amazon (+ power brick) or cheaper on the Aiyima site if you can wait for delivery from China. From China you can also choose LM4562 op amps which have better specs than the standard NE5532 though if the circuit is properly designed I doubt anyone could hear the difference.

@gzaharabove said "2448A7 holds half as many links as 2413U/S, so if you have a lot links in your current setup, you might want to go for 2413x. " I've seen this said elsewhere too. Smartlabs at one point said the 2413s/u could store 2000 links, but this seems to have been false. So maybe they have the same link memory capacity after all?

I am confused about the max number of links on the various PLMs... Elsewhere in the forums the 2413s is reported to have a maximum of 992 links, which seems precise and based on knowledge (Search 992). So does the 2448a7 have order of magnitude 1000 link capacity, equal to the 2413s? If half of 992, then @waffles may have too many links to fit into the 2448a7? You see my confusion...

I should clarify that with Sonos you only need one device in your sonos system that is Airplay enabled. Then any other Sonos devices on the network will be able to play that material, even if the receiving device is muted. For a small Lake cabin you may not be looking for multi-room sound.

Yes Apple Airplay works with the Sonos Ports and some other Sonos devices. Airplay over wifi (no internet connection needed) is higher fidelity than bluetooth. But there are cheaper alternatives than Sonos to do this, especially if you have a receiver/amplifier with a optical (Toslink) input. Belkin makes a "SoundForm Connect Audio Adapter with AirPlay 2" that will allow you to stream from an iOS or MacOS device. This is $80 on Amazon and B&H Photo. People complain that the analog output is sort of low level. I can't attest to the quality of the DAC built in and thus the analog sound quality, but the digital out should be CD quality if the file on the phone is. There are also some Yamaha, Onkyo, and Denon receivers (maybe Sony also) with Airplay or Airplay 2 built in.

I had this problem with a particular KPL. Nothing I could do would fix it, but changing the communication mode in the scene dialog (I think from Insteon to Command or something like that) turned the connected load on and off most of the time, though with an annoying delay. I had another KPL right next to it so I moved the button to that one (This one's load was not a member of the scene.) I reported this on the forum at one point but I can't find it now. I have other KPLs that have analogous configurations that work fine. It's weird.

@lilyoyo1 Thanks. I will take the class at some point. My only motive in this last post was to document the official character of the statement that in the near future Lutron intends to enable using two RA3 processors to more or less double the devices served. I wasn't arguing with the detailed info @upstatemikeprovided.

The reference to two RA3 processors serving 400 devices is about half way down this page from the official Lutron site: https://radiora3.lutron.com/us/en/whats-new The footnote describing that 200 devices is actually up to 100 of each kind is also on this page.

@upstatemikeThanks for the clarification on the device serving capability of the RA3 processor. I haven't again found the the Lutron page that referred to a coming capability to link two RA3 processors to serve 400 devices but the language was very similar to that found on the following page, which seems to be essentially copied from a Lutron press release or webpage. https://www.prolighting.com/blog/2021/12/03/lutron-reveals-game-changing-radiora-3-for-2022/ "Each processor controls up to 200 Lutron devices** with Lutron’s ultra-reliable ClearConnect® RF technology. Coming soon — the ability to pair two RA 3 processors to create 1 system of up to 400 devices." "**Up to 100 new RadioRA 3 devices (dimmers, keypads, switches) + 100 compatible RadioRA 2 & RA2 Select devices." (consistent with your clarification)