Klipsch RP-600M Speaker Review

[maty]

Resistors with very low inductance -> speaker crossover and in others places
https://www.audiocircle.com/index.php?topic=166906.0

Note: Some images do not appear because the free FTP service that I have been using since 1998 was canceled weeks ago

 

[edechamps]

But up to app 15kHz it seems fine, right? Or am I missing something?

Uh, no. As early as 8 kHz, +30° horizontal is 5 dB below -30° horizontal. That's just weird for a speaker that, in the pictures, looks completely symmetrical in the horizontal axis. @amirm: is it conceivable that the measurement axis might have been off-center?

 

[edechamps]

It seems to me that practically every speaker that has been measured experiences a fall from 100Hz or so. Isn't that weird?

There is another discussion about this in the Revel F35 thread, where some suspect this might be a deliberate design decision to compensate for boundary effects.

I think in the case of this speaker it's fine because it agrees with third party measurements. I'm a bit more worried when it comes to speakers like the Neumann KH 80 or the JBL 705P where @amirm's measurements show a dip in the bass that does not appear in third party and manufacturer measurements. However it could very well be that Amir's measurement is the correct one because the NFS is supposed to be more accurate than other measurement systems in this frequency range.

 

[QMuse]

There's one more thing I noticed although I don't have time to check it via your Loudspeaker Explorer:

It seems to me that practically every speaker that has been measured experiences a fall from 100Hz or so. Isn't that weird?

View attachment 55120

Here is what I find strange. This is F35 measured response and here are the specs from manufaturers page:

Low Frequency Extension 55Hz, 46Hz, 35Hz (-3 dB, -6 dB, -10 dB)

Sensitivity 90dB (2.83V @ 1M)

Ok, so sensitivity is indeed there at 89dB, so very close to spec, but lets' now check LF extension:

35Hz - app 72dB (-17dB)
46Hz - app 77dB (-12dB)
55Hz - app 82dB (-7dB)

Not really according to spec, right?

 

[QMuse]

There is another discussion about this in the Revel F35 thread, where some suspect this might be a deliberate design decision to compensate for boundary effects.

I think in the case of this speaker it's fine because it agrees with third party measurements. I'm a bit more worried when it comes to speakers like the Neumann KH 80 or the JBL 705P where @amirm's measurements show a dip in the bass that does not appear in third party and manufacturer measurements. However it could very well be that Amir's measurement is the correct one because the NFS is supposed to be more accurate than other measurement systems in this frequency range.

You are right - I just checked JBL 705P measurement vs specs and it seems ok from the LF extension aspect. Mistery about F35 LF extension still remains..

 

[edechamps]

Here is what I find strange. This is F35 measured response and here are the specs from manufaturers page:

Low Frequency Extension 55Hz, 46Hz, 35Hz (-3 dB, -6 dB, -10 dB)

35Hz - app 72dB (-17dB)
46Hz - app 77dB (-12dB)
55Hz - app 82dB (-7dB)

Not really according to spec, right?

I wouldn't be surprised if the manufacturer numbers went through some "creative accounting" by the marketing department, for example maybe the numbers assume a reverberant environment (i.e. a normal room with bass gain and boundary effects, or half-space or something) as opposed to far-field anechoic.

 

[QMuse]

I wouldn't be surprised if the manufacturer numbers went through some "creative accounting" by the marketing department, for example maybe the numbers assume a reverberant environment (i.e. a normal room with bass gain and boundary effects, or half-space or something) as opposed to far-field anechoic.

Nooo.. You are not saying that even Harman would be doing something like that?

 

[Juhazi]

F35 is the only tower speaker so far, with three bass drivers and two ports on the backside. That might explain low bass measurement of NFS/amirm.

 

[Jim777]

Personally I never heard a Klipsch I didn't hate because the tone is completely off. Looks like the spinorama correlates with my preferences..

 

[ctrl]

I have verified that the Bennic cast resistors spoil the sound in my modded KEF Q100 a lot. The difference when changing them with Mills MRA was undoubted, and that I did the test with a single modifying loudspeaker compared to the other with the original resistance.

Did you measure the resistance levels accurately beforehand? Were the values of the new resistors exactly identical to the old ones?

If not, the sound change is most likely due to slightly different resistor values (depending on the tolerance values of the resistors) and not due to the different "residual inductance" of the resistors.

Here the measurement of two selected resistors, both have almost exactly 6.8V.
Once an ordinary 20W "sand cast" resistor

and a low-inductance 4W MOX resistor

Attention, the measurement of resistance value against frequency follows
Pay attention to the scaling.
The measured difference comes from the slightly different resistance values despite selection.
(the slight resistance increase above 10kHz is caused by the meter)

This is the reason why also well-known manufacturers use the low-priced "sand cast" resistors.

UPDATE: Most resistors have a tolerance range of +- 5%. This means in our example with 6.8 Ohm that the values may lie between 6.46-7.14 Ohm at purchase.

 

[mhardy6647]

Personally I never heard a Klipsch I didn't hate because the tone is completely off. Looks like the spinorama correlates with my preferences..

I'd generally have to agree with this post (albeit with a couple of exceptions -- the K-horns and the Chorus are pretty acceptable to my ears).
Along these lines, though, I'd love to see the data for any generation of Corny Cornwall!

Lcornieoly by Mark Hardy, on Flickr

I guess they're not too Klippel-friendly, though -- maybe a Heresy? Any generation would be fine (although I guess the fairest test would be the current version -- a little pricey, though, for a small-ish sealed box monkey coffin with three drivers in it).

 

[maty]

Did you measure the resistance levels accurately beforehand? Were the values of the new resistors exactly identical to the old ones?

If not, the sound change is most likely due to slightly different resistor values (depending on the tolerance values of the resistors) and not due to the different "residual inductance" of the resistors.

Here the measurement of two selected resistors, both have almost exactly 6.8V.
Once an ordinary 20W "sand cast" resistor
View attachment 55129

and a low-inductance 4W MOX resistor
View attachment 55130

Attention, the measurement of resistance value against frequency follows
Pay attention to the scaling.
The measured difference comes from the slightly different resistance values despite selection.
(the slight resistance increase above 10kHz is caused by the meter)
View attachment 55140
This is the reason why also well-known manufacturers use the low-priced "sand cast" resistors.

UPDATE: Most resistors have a tolerance range of +- 5%. This means in our example with 6.8 Ohm that the values may lie between 6.46-7.14 Ohm at purchase.

No i do not. But trust me, the difference was too much. I expected a little improvement but not so much, hence the thread I opened because I was surprised. Perhaps it is the sum of two factors: a poor quality (value and more inductance) of the originals and a much better quality of the new ones, which would explain so much difference.

The next time I change the resistors I will measure them. I threw them in the trash, angry.

 

[ROOSKIE]

Hi. I don't have a link to any specific measurements when using the speaker connector tubes. Though, the sound can be heard in one of the reviews by New Record Day on YouTube. I was listening to the video, while commuting one day a few weeks back. In the review, he'd play a clip, then indicate what it was we had heard. To me, this was like a blind test. My guesses while listening to the sound were quite wrong when I'd hear the clip, then he'd reflect whether it was the stock version of the speaker or the one with the modified / replacement crossover. However, the sound tube clip had a noticeable change in the sound that could clearly be heard. The speaker tubes are about $40 or so a pair from GR Research. Might be worth considering when building a new pair of speakers or doing some extensive modifications on an existing pair of speakers.

I have listened to a couple of their comparison tests and they are not that well set up. In the one I remember the most ( a behringer speaker) Speakers A was always played first and speaker B second. They never mixed up the order. Of course speaker B was the modded speaker. Do you think that hearing a new track played for the second time you will notice more detail than the first? Also the tracks did not allow switching from speaker to speaker at the exact same point which I find is a must.

At first I heard a clear difference every-time. Then I mixed up the order on my own and timed it so I could switch between tracks at the same point in the song and that point I couldn't tell which was which. I will say there may have been a difference in the room or for other folks but I was not enthused after I mixed everything up.

This reminders me of several online flac vs mp3 tests, I have never met anyone who passed these. Some people with high end headphones pick the 160kbps vs the FLACC.

 

[ROOSKIE]

Yes, I understand your point. But, if you listen to the clips, you can hear. However, the measurement component is essential too - maybe, the speaker tubes are increasing the efficiency by a few decibels, which may be the difference in sound that is heard. Truly, yes, measurements are key - not just impressions of the reviewer on a YouTube channel. We've realized with all the great reviews Amir has conducted here, that there is a big difference between all the hype, marketing and costs behind a product vs. how it really performs on the test bench.

There is no way they are increasing the efficiency by a few DB. NONE. The resistance of the wire is not changing much. It likely isn't affecting the sound at all however I am open to it. I have not seen conclusive proof they don't work so there is that.

 

[ROOSKIE]

Quite a few speaker manufacturers use sand cast resistors in their crossover networks. Is it done because of cost? Possibly. But, when building a crossover network, I personally now tend to avoid them. When I had my previous pair of of Elac B6.2 speakers, I replaced one of the capacitors and two of the sand cast resistors with another type of resistor from Madisound. I wish I had the proper measuring equipment to have tested the before and then the after results of the $10 modification effort (minus the cost of shipping). Even the "famed" Buchardt S400 speakers @ $1,900 a pair have sand cast resistors. The speaker has received many great reviews, with some testing data being provided - but none to the degree that our famed reviewer, Amir would have been able to provide. Admittedly, I should digress from posting - lots of work this week and the 4 hours of sleep per night is hampering with my thought and response process. Just 8 or 9 more hours to go, then freedom!

It is not clear from you comment. Are you saying that the modded B6.2 sounded better than it did before the mod? If so where you able to compare the before and after in some way?

 

[John Atkinson]

JA had this to say about the measurements:

Interesting measurements, Amir, and congrats on using the Klippel system. One point: the text you quoted me as writing was not complete: it actually read "The Klipsch's farfield response has a slight lack of energy in the crossover region. Despite the spatial averaging, the tweeter is still balanced a little too high in level. I wasn't surprised, therefore, that HR found that the RP-600Ms 'slightly emphasized the leading edges of notes.' HR also wrote that he had 'a slight preference' for the sound of the RP-600Ms with their grilles on. The grille reduces the tweeter's output by 1dB or so between 3 and 8kHz." - see https://www.stereophile.com/content/klipsch-reference-premiere-rp-600m-loudspeaker-measurements

John Atkinson
Technical Editor, Stereophile

 

[YogiN]

I have added the Klipsch RP-600M to Loudspeaker Explorer where it can be compared to other speakers.

The listening window is a bit eyebrow-raising, as the speaker is not symmetric on the horizontal axis. Highs roll off way faster at positive horizontal angles, compared to negative horizontal angles. This is perplexing. In any case this speaker fails to present a consistent response within the listening window in the treble.

View attachment 55116
View attachment 55117

Would this suggest that they would benefit from being toed out?

 

[maty]

John Atkinson
Technical Editor, Stereophile

Welcome to the forums!

 

[BurgerCheese]

There's one more thing I noticed although I don't have time to check it via your Loudspeaker Explorer:

It seems to me that practically every speaker that has been measured experiences a fall from 100Hz or so. Isn't that weird?

Most of the speakers so far has been small. Having good sensitivity below 100-150 Hz in a small speaker is simply not possible.

This speaker is probably worse than most others as Klipsch attempts to keep sensitivity high. The frequency response in the bass indicates a woofer with a very high Fs/Qes ratio (EBP) in a large enclosure relative to VAS. This typically translates into a gentle roll-off around 6 dB/octave between 100-200 Hz and the tuning frequency. This type of ported alignment is/was usually referred to as "extended bass shelf".

 

[Deleted member 2944]

Wouldn't the impedance interaction with a high output-impedance valve amplifier somewhat "compensate" for the on-axis dip at 2kHz, whilst simultaneously increasing the treble and creating other problems in the bass?

Yes indeed, and a good observation. I don't see the Z data in Amir's supplied data, but with that info you could do a pretty good simulation of relative response changes with an amplifier that had significant output impedance.

Dave.