Ascend Sierra Tower + Emotiva XPA-11 = Burnt Voicecoil

[davef]

Hi Dave, thanks for the reply! I did test the amp with an old speaker and didn't hear anything. That's why I like the theory of failure at a certain temperature idea... perhaps under a heavier load.

I do have a multimeter...

In this case, I recommend reconnecting this same amp to the towers and connecting your multimeter across the speaker terminals (positive to positive, negative to negative). Be sure to switch your multimeter to the DC setting. Take a few different measurements.

DC at very low volume
DC at medium volume
DC at loud volume

Just in case there is a significant DC offset, do this quickly. A DC offset will not take out a woofer immediately, it does take some time for the voice coil to heat up.

Post your measurements and then we can take the next step.

another option if you like, is to send the amplifier to me and I can test this for you with proper equipment (oscilloscope + power resistor bank for a static load). However, you would be responsible for all shipping costs which could be quite expensive these days. I have done this in the past for other customers and I don't mind it at all. I miss working on amplifiers.

[jeeper]

In this case, I recommend reconnecting this same amp to the towers and connecting your multimeter across the speaker terminals (positive to positive, negative to negative). Be sure to switch your multimeter to the DC setting. Take a few different measurements.

DC at very low volume
DC at medium volume
DC at loud volume

Just in case there is a significant DC offset, do this quickly. A DC offset will not take out a woofer immediately, it does take some time for the voice coil to heat up.

Post your measurements and then we can take the next step.

another option if you like, is to send the amplifier to me and I can test this for you with proper equipment (oscilloscope + power resistor bank for a static load). However, you would be responsible for all shipping costs which could be quite expensive these days. I have done this in the past for other customers and I don't mind it at all. I miss working on amplifiers.

Thanks, Dave. I tested the amp without anything connected, and all channels were near zero. Will take some measurements with the towers connected and at different volumes...

BTW, I was over at the Emotiva forum asking about DC offset and people were very sensitive lol. They thought I drove the speakers too hard. Perhaps, I did... and if that's the case I've learned my lesson. If I were to use an SPL meter, what would you say would be the starting limit of upper threshold? (so I can stop the tequila rounds)

[jeeper]

In this case, I recommend reconnecting this same amp to the towers and connecting your multimeter across the speaker terminals (positive to positive, negative to negative). Be sure to switch your multimeter to the DC setting. Take a few different measurements.

DC at very low volume
DC at medium volume
DC at loud volume

Just in case there is a significant DC offset, do this quickly. A DC offset will not take out a woofer immediately, it does take some time for the voice coil to heat up.

Post your measurements and then we can take the next step.

another option if you like, is to send the amplifier to me and I can test this for you with proper equipment (oscilloscope + power resistor bank for a static load). However, you would be responsible for all shipping costs which could be quite expensive these days. I have done this in the past for other customers and I don't mind it at all. I miss working on amplifiers.

Hi Dave, so I just did this. Here are the measurements. This would be from a cold power on for the amp as well...

measurements.jpg

[davef]

Hi Dave, so I just did this. Here are the measurements. This would be from a cold power on for the amp as well...

measurements.jpg

90mv is a bit high, typically, any DC offset should remain below 50mv. However, 90mv wouldn't be enough to take out both midrange drivers.

Next step is to run the same test, but run it for a while under load with moderate to loud volume to let the amp heat up. With output transistors, sometimes the solder joints breakdown over time such that when they get hot, they expand and those joints fail.

[RandomName]

Typical Emotiva - never an issue with any of their products.

Such truth. I cannot assign blame to be clear as I have no clue when it actually failed, but I was using one of their 2 plug a/c filters specifically for the “dc offset protection” that the,6 plug one lacks. Anyway, long story short, one of my Funk 18.0c’s amps up and died at some point just sitting idle. Strangely enough, the CMX-2 is also seemingly completely fried. can’t even reset the breaker and it is stuck saying “open ground”.

I know very little about electronics, but it seems to me at least like it should be quite difficult for something downstream to blast the ac filter/breaker that was supposed to be protecting it. Seems way more likely that it took my amp with it when it died.

Not the end of the world, I checked already and even if the amp is a brick, it should at least still just be in the hundreds to get like all the electronics repaired/replaced. It’s a sub after all, the veneered rounded Baltic birch cabs and the fancy custom drivers are all the work and $$$, and like all subs they just mount the appropriate sized class-d and are good to go (well and some sort of dsp usually, but even the “pro grade” 64-bit like I believe is in my sub, https://www.alldsp.com/product-1800B.html, aren’t that expensive). Switch mode noise is usually all high frequency after all. If I recall right my subs CAN play up to like 300Hz. And because of the c-curve or w/e (our reduced perception of bass) and the frequencies of the noise even on some garbage icepower amp, will usually be able to drive a sub just fine. The subs in this should be nicer than that, but still nothing compared to the guts of a proper speaker amp.

Either way, I’ve already started researching alternate options to get rid of my other CMX-2. It’s totally possible it was in fact the sub, but since I don’t plan to get rid of the working sub (and am getting the other one fixed), and the CMX-2 was pretty chep anyway. The real issue I’m having is just separating out the snake oil from the good stuff. There’s sooooo many different ways to clean and isolate power. Isolation transformers, crazy huge balanced power transformers, ps audio style dsd 120hz sin wave generator with a big amp and huge rails, and a million other options. I’ve been seriously contemplating just trying to get enough LTO batteries and a charger/inverter setup as a psuedo ups/HiFi session battery.

But I now am also doing the same thing op should, trying to investigate the current state of my building/apartment power. There are so many different ways to clean power, but depending on what the problem is, it may well be something that will keep causing issues, especially in big high density buildings like mine (~10-15 year old 13f multi-unit urban apt complex), and the increased building load just from all the bodies at home cause of Covid.

The main reason I’m tossing the emotiva is paranoia and if there is an electrical issue, a/c protection and repair and such is not really what they do. Gotta determine how much I gotta actually “fix” by finding some way to do like logging measurements. I’m sure something exists. Tho knowing fluke it won’t be chep. Hopefully china makes an economy version.

I’d probably try and check the A/C and perhaps even the amp out signal (to be clear high voltage dc is dangerous, only do that if you got the right equipment or know someone who knows how to do such testing safely) with with various sweeps and other patterns from a signal generator. Cause as best I’m aware, the only things that could cause both mid woofers to die like that would be, either it just blasted the speakers and the mids died because the tweeters protect themselves and the woofers both distribute load and just in general have less work to do (the spl to driver surface area growth rate is INSANE, it’s why I have twin 18” subs, they can output what a pair of 12s can doing massively (like exponentially or logarithmically or w/e it would be) less work. And at least with these drivers, they are suuuuuuper clean when they aren’t working hard. Oh, the 3rd possible reason, and this one I know very little about has to do with impedance, because no matter how good a crossover is, it can’t magically make all the drivers have the same impedance, and I’m pretty sure there is some type of failure that can have a bad reaction/interaction with various impedances.

It’s certainly not a matter of the main channels being too powerful or anything tho, I run AHB2 monoblocks on my raal towers which is 380w @ 8ohm, and while I don’t think I’ve ever turned them all the way up I’ve certainly cranked it when I’m listening to music through my Dac & pre rather than tv with my receiver. I feel like you would have to hurt your ears to really damage the speakers, since it would take EXTREME volume or excessive volume for an extended period (such that parts heat up and die). It’s always amazed me just how much transient power decent speakers can take… ‘as long as it’s clean’.



Anyway, TL;DR, I’d trust emotiva’s DC offset protection about as much as I would trust a ford pinto’s gas tank. It should be safe… it’s probably safe… but it just might explode.

I for one would much rather invest in ATI built amps either direct or through monoprice. Or maybe some Vidar monoblocks (my AHB2s are suuuuper overkill, I love them and gladly recommend them to other people who are like me and have money to waste, but there are so many good options these days, plus all the ncore and purifi module based stuff, heck, the ATI NC528 would have been a far better use of my money, but I wanted sexy monoblocks next to my towers lol… I am trash).