Neumann KH120 II Monitor Review

[Pearljam5000]

The only reason I'm not getting the KH120 II is reliability concerns
A studio monitor should work for 15 years with 0 problems
Why isn't anyone officially from Neumann comments here about this?

 

[Karmacoma]

Why isn't anyone officially from Neumann comments here about this?

Because so far it's 5 dudes on an obscure forum.

 

[Pearljam5000]

Because so far it's 5 dudes on an obscure forum.

Actually that's not such a small number

 

[Karmacoma]

we've seen elsewhere that Neumann is using lower-grade power supply electrolytic capacitors that have a reputation for premature failure

I thought that had been debunked by a Neumann engineer on this very forum....

 

[Karmacoma]

Actually that's not such a small number

Is it?

 

[danila]

I see a lot of misunderstanding in that post.

It's usual for monitor speakers to get hot and to work like that. I use Neumanns for a few years now and before I had most of popular brands in this market - Dynaudio, Adam Audio, Genelec, Tannoy, Yamaha, Focal... and every single one of them got hot at some point if you just power them on. Not any one failed on me and with most cases I was a second or third owner. Maybe it's luck.

It's very popular for people to complain about something if it's wrong. I don't know how many Neumann monitors were sold, but I think that's quite A LOT, so question is - what is real failure rate. I worked for audio distribution of Focusrite and failure rate of Scarletts was REALLY low. But from time to time there's a guy that have that one unit that just died. That happens. And he's a YouTuber and gets a hundred thousand of views and of course people are gonna get a little bit angry. We like to. On the other side you have hundred thousands of users that are happy and won't post anything besides their recorded audio.

Volume should be faster to respond, but it's only for a calibration reason - it's set and forget type of volume.

I don't know what you expected about calibration, but it is done for specific position of speakers. If you change anything about it (left right higher lower), you change the acoustic response. If you got really messy room the calibration sweet spot can be very small and it is what it is. Mind you that calibration should be used as a helper, not a complete solver.
With calibration is quite the opposite as you wrote - in studios you could calibrate only the lows because that's the most difficult region to treat and to get right, so some massaging is always welcome. Mids and highs are very easy to absorb and/or diffuse and usually it's fine if calibration won't touch those bands. In living rooms and bedrooms you need to correct a lot more, because we don't usually have them symmetrical and response could get a little funky on one or the other side.

MA-1 calibration response gets the sum off all points you measured. Did you tried to measure the same points with REW and then made an average of them?

It's a known problem for MacOS users, but Apple have done nothing to solve it for years. Weirdly on Windows it's much easier because you can choose different amount of channels for audio driver. MacOS always use every channel, so sometimes I get DTS or even Atmos sound by default, because my interface got 16 output channels.

Thank you for a nice reply!

Yeah, I understand that I'm judging those monitors harshly. But that were my subjective expectations of them: I wanted an engineering masterpiece that I will put on my desk (on the correct height, on eight iso-pucks), automatically calibrate it for room/desk reflections, and enjoy the music. I was willing to pay 2000 bucks for the setup.

I only truly enjoyed how silent the hiss was.

To my ears, the music was very similar to that of my 100-bucks a piece LSR 305. REW could measure that Neumanns had less distortion in the 1–5 kHz range, but I doubt I could hear that.

The room calibration turned out to be finicky and very dependent on a lot of things going on in the room like the desk height and my position. These were my misguided expectations, I'm not blaming Neumann. Besides, maybe in a properly treated studio (where these monitors are supposed to work) MA-1 does wonders. On the other hand, the advertisement of the feature by Neumann exaggerates the usefulness of it, I think.

As for the software and how laggy it was, well, it was the same laggy behavior when turning the EQ (target curve). Maybe it was even more laggy, with you having to disable and reenable calibration for the EQ changes to take an effect, each operation taking seconds. While it's useable, it's also not a joy to use. As a comparison, look at what this guy did: https://github.com/LeanderBlume/kh_network_control_mac!

As for reliability, yeah, the question is open, but there are people are saying that both speakers failed independently from other another, and that suggests a rather high rate of failures.

Are all active monitors like Dynaudio, Adam Audio, Genelec, Tannoy, Yamaha, Focal really heat up to 50 degrees C while being silent? Like, class D active monitors? But why? What are they spending the energy on? Mine LSR 305 are cold.

Anyway, I was packing them today to ship them back. I was having doubts, but still decided to return them. While I was disconnecting them, I noticed that one of the monitors was hot! Despite not being used for several days! I explicitly enabled the sleep on both. In fact, I enabled it twice because the first time I did that it didn't work (they forgot the setting). Apparently, the second time it either also didn't work (i.e. one of the monitors forgot the setting again) or it was the famous problem of these monitors with not going to sleep. I smiled and stopped having doubts.

 

[Kyle_PL]

Are all active monitors like Dynaudio, Adam Audio, Genelec, Tannoy, Yamaha, Focal really heat up to 50 degrees C while being silent?

Genelec's are cold. Even if mine Genelec 8330A playing music, they have 3x Celsious on DSP, and cold enclosure.

 

[solawind]

What do you folks think about KEF LSX as PC speakers

I quite liked 120 II sounds while they lasted. Unfortunately, one monitor has developed a crackling sound, just like many cases in this thread. I researched the market for passive speakers and bought KEF LS50 Meta, paired with a cheap Topping Mini 300, and after room correction, the KEFs are almost indistinguishable from the 120 II. The obvious difference is only on the low end, but since I use a sub anyway, there is almost no difference. I made an A/B test, sticking them on top of each other.. They are the same size and LS50s coax drivers works in nearfield quite well
Not a fan of the LS50 without room correction, tho — they sound noticeably worse than the stock 120 II..

 

[dfuller]

I thought that had been debunked by a Neumann engineer on this very forum....

No, it hasn't been. It was all very handwave-y.

 

[Karmacoma]

No, it hasn't been. It was all very handwave-y.

I guess he didn't know about the failure rate to act like that and defend the component choice.

 

[Torin Krell]

Genelec's are cold. Even if mine Genelec 8330A playing music, they have 3x Celsious on DSP, and cold enclosure.

My Genelec 8341a monitors usually run quite cool too (and have worked flawlessly for over 5 years). Like most newer powered pro Genelecs, they utilize their whole aluminum enclosure as a heat sink while the Neuman KH 120 II and KH 150 only use an average sized rear heatsink (the enclosures are not aluminum and not especially heat conductive). My Genelecs only get moderately warm when I use them like midfield monitors and push them hard for several hours.

 

[Steve Dallas]

I quite liked 120 II sounds while they lasted. Unfortunately, one monitor has developed a crackling sound, just like many cases in this thread. I researched the market for passive speakers and bought KEF LS50 Meta, paired with a cheap Topping Mini 300, and after room correction, the KEFs are almost indistinguishable from the 120 II. The obvious difference is only on the low end, but since I use a sub anyway, there is almost no difference. I made an A/B test, sticking them on top of each other.. They are the same size and LS50s coax drivers works in nearfield quite well
Not a fan of the LS50 without room correction, tho — they sound noticeably worse than the stock 120 II..

Fearing for my glitchy 120IIs, I recently trialed this exact setup, except with my OG LS50s. With room correction they sound just as good. My tracking desk up against a wall with 4" absorbers behind the speakers to fix SBIR, and boundary reinforcement is enough to provide satisfying bass. Since I don't listen loud enough to create appreciable distortion nor tax the drivers, I can extend the bass response to about 35Hz in-room, which is plenty for what I do. I will use the 120s until they die, but will likely switch to the LS50s and a Mini 300 afterward. The Mini 300 is such a great little amp...

 

[Blockader]

No, it hasn't been. It was all very handwave-y.

Was it, though? He was not saying "we tested it, so it's fine." He's addressing the actual mechanism.

Neumann got accused of using Capxon caps because of a known problem 23 years ago. But that issue wasn't Capxon-specific. It was a bad electrolyte formula from a shared sub-supplier that hit nearly every Taiwanese cap maker at once (the "capacitor plague," well documented). That got fixed industry wide.

Also, just testing a cap when it's new wouldn't catch this anyway. A cap can look fine on day one and still die years later from the electrolyte breaking down. What actually matters is how hard you push it. He says they run parts well under their rated temp and voltage, and that alone can take a "1000 hour" cap and stretch its life past 50000 hours. That's the real answer to "will it fail" and it has nothing to do with which brand you picked.

Neumann KH80DSP Teardown

Also interesting that they went with Cirrus Logic CS42526 chip to do both 2 channel ADC and 6 channel DAC functions with it. As they mention in the description of the speaker, they used both ADC channels to digitize the XLRs + and - signals separately, and sum them digitally. ADC THD+N is -100...

www.audiosciencereview.com

For those who want to read Markus Wolff's comment about their capacitor choice.

 

[respice finem]

IMHO it is also possible that the KH 120 II gets hot, because the cabinet is not aluminum any more (as it was in the predecessor IIRC).
But one way or the other - one of two speakers getting hot in idle can't be normal.

 

[dfuller]

Also, just testing a cap when it's new wouldn't catch this anyway. A cap can look fine on day one and still die years later from the electrolyte breaking down.

Yes, that's the thing that I'm talking about. Considerably shorter lifespan than the datasheet suggests, which could explain why issues are popping up much sooner than any electronics bathtub curve would suggest they should.

Neumann got accused of using Capxon caps because of a known problem 23 years ago. But that issue wasn't Capxon-specific. It was a bad electrolyte formula from a shared sub-supplier that hit nearly every Taiwanese cap maker at once (the "capacitor plague," well documented). That got fixed industry wide.

Even 20+ years after the capacitor plague, CapXon electrolytics still have a reputation among electronics repairers of premature failure, especially in high temperature, high ripple current environments (like that of a switching power supply without much in the way of cooling). And I do understand why they do this - it's almost certainly more economical to deal with the premature failures (which, to be clear, definitely still are a small percentage...) than to increase BOM costs with more expensive parts.

For the sake of comparison, though... Genelec is using Samwhas in their similarly priced speakers. ATC uses Jamicons in their actives, last time I checked. You definitely do not need ultra-high-end parts, but midrange brands like Samwha, Teapo, Jamicon, Yageo, etc etc aren't any significant amount more expensive in assembly line quantities, and are considerably more reliable. Hell, I took a KH310 apart at one point - all the major electrolytics were Yageo or a smattering of Japanese parts, it's not as though they are unable to source the better quality stuff.

But one way or the other - one of two speakers getting hot in idle can't be normal.

It would be if they used AB amps that idle with considerable current. Class D amps (should) use extremely little power at idle - e.g., Hypex NC252MP idling draws about 8.5W, a similar output power Class AB will probably draw more like 20-30w per channel if not more (depends on the biasing). Last I checked, my KH420s draw roughly 60w per speaker when idling, but that's ~550W per plate amp and includes PSU losses and the op-amps in the crossover.

 

[Blockader]

Yes, that's the thing that I'm talking about. Considerably shorter lifespan than the datasheet suggests, which could explain why issues are popping up much sooner than any electronics bathtub curve would suggest they should.

Even 20+ years after the capacitor plague, CapXon electrolytics still have a reputation among electronics repairers of premature failure, especially in high temperature, high ripple current environments (like that of a switching power supply without much in the way of cooling). And I do understand why they do this - it's almost certainly more economical to deal with the premature failures (which, to be clear, definitely still are a small percentage...) than to increase BOM costs with more expensive parts.

For the sake of comparison, though... Genelec is using Samwhas in their similarly priced speakers. ATC uses Jamicons in their actives, last time I checked. You definitely do not need ultra-high-end parts, but midrange brands like Samwha, Teapo, Jamicon, Yageo, etc etc aren't any significant amount more expensive in assembly line quantities, and are considerably more reliable. Hell, I took a KH310 apart at one point - all the major electrolytics were Yageo or a smattering of Japanese parts, it's not as though they are unable to source the better quality stuff.

It would be if they used AB amps that idle with considerable current. Class D amps (should) use extremely little power at idle - e.g., Hypex NC252MP idling draws about 8.5W, a similar output power Class AB will probably draw more like 20-30w per channel if not more (depends on the biasing). Last I checked, my KH420s draw roughly 60w per speaker when idling, but that's ~550W per plate amp and includes PSU losses and the op-amps in the crossover.

Neumann claims that they test those capacitors and they are "good enough" to work for next 10 years. Whether this is truly the case or not is beyond my expertise. I agree with you overall.

Part of this is on us too honestly. A lot of buyers only judge speakers by measurable performance, frequency response, distortion, whatever shows up in a graph and never factor in long term reliability. That kind of pressure pushes companies to cut corners where it's invisible on a spec sheet, and BOM cost is exactly the kind of thing that's invisible until years later.

related to other topic above: Genelec 8361 draws 20W while being idle. 1000W max power, 20W idle draw. 1/50 ratio.

 

[respice finem]

Yes, that's the thing that I'm talking about. Considerably shorter lifespan than the datasheet suggests, which could explain why issues are popping up much sooner than any electronics bathtub curve would suggest they should.

Even 20+ years after the capacitor plague, CapXon electrolytics still have a reputation among electronics repairers of premature failure, especially in high temperature, high ripple current environments (like that of a switching power supply without much in the way of cooling). And I do understand why they do this - it's almost certainly more economical to deal with the premature failures (which, to be clear, definitely still are a small percentage...) than to increase BOM costs with more expensive parts.

For the sake of comparison, though... Genelec is using Samwhas in their similarly priced speakers. ATC uses Jamicons in their actives, last time I checked. You definitely do not need ultra-high-end parts, but midrange brands like Samwha, Teapo, Jamicon, Yageo, etc etc aren't any significant amount more expensive in assembly line quantities, and are considerably more reliable. Hell, I took a KH310 apart at one point - all the major electrolytics were Yageo or a smattering of Japanese parts, it's not as though they are unable to source the better quality stuff.

It would be if they used AB amps that idle with considerable current. Class D amps (should) use extremely little power at idle - e.g., Hypex NC252MP idling draws about 8.5W, a similar output power Class AB will probably draw more like 20-30w per channel if not more (depends on the biasing). Last I checked, my KH420s draw roughly 60w per speaker when idling, but that's ~550W per plate amp and includes PSU losses and the op-amps in the crossover.

Thanks for the info on caps in the KH 310A, mine are now 5 years old. I'm using them for "serious listening" only, and turn them off when not in use, so maybe 3000-4000 power-on hours yet. All good so far. They have a large radiator, and when playing loud, the whole backplate becomes a radiator, so I'm not hugely worried. I think that in my usage scenario, they might last me 10-15 more years (hopefully).

 

[dfuller]

Part of this is on us too honestly. A lot of buyers only judge speakers by measurable performance, frequency response, distortion, whatever shows up in a graph and never factor in long term reliability. That kind of pressure pushes companies to cut corners where it's invisible on a spec sheet, and BOM cost is exactly the kind of thing that's invisible until years later.

Yep. Gotta hit a certain price point, optimize as you will. Reviews move units so optimize for that, and if it gets through the warranty period you're good to go.

 

[Kalla94]

What am I supposed to see? A dude blowing a fuse?

I'm not sure where the internal fuse is located, but the spark came from the woofer.

 

[Karmacoma]

I'm not sure where the internal fuse is located, but the spark came from the woofer.

I didn't see that. I can't check as the video seems to be offline now.
And everything's located on the back plate so I really don't get what could generate a spark "from the woofer".