I wouldn't mind trying the modification if I could be sure or the source and the correction for the noise. I'm not entirely sure what to make of Nelson's post on Asylum. He suggests the noise does come from the auto-formers, but he also said he's adjusted the positioning of the adjacent transformer and mu metal shielded the auto-formers in the factory build.
Given this amplifier circuit is no feedback through MOSFETS and JFETS, I can't help but speculate that this noise I hear might be thermal noise from the gain stage, not PSU noise picked up by the auto-former. Wouldn't PSU noise tend to be more of a hum or LF noise? I am getting what I'd say is a higher frequency white noise as heard through my HF compression driver and 800 hz. cutoff horn.
I can understand why some typical noise hacks for amplifiers like relocating the PSU are often effective solutions, but as you say, I'd really like to know a modification is the right modification for this particular amp before I make changes to a factory build.
Thank you! That was my intuitive guess, but I have no experience to validate that. I think assuming the noise is from NFB MOSFET circuit is the right strategy and I can hide the noise another way in the speaker crossovers, etc. Right now I'm working with a very high efficiency compression driver and horn and they always require some attenuation to match midbass woofer below.
And thank you for the L-Pad confirmation. I have used crossovers from Werner Jagusch that have auto-formers to attenuate the response of the HF drivers in two way speakers and I have liked that solution for passive crossovers.
I have been trying to use active crossovers to bi-amp a woofer and horn two way speaker. The active crossovers from Xkitz provide easily changeable xover points and attenuation of the treble channel, so I thought they'd be easy to manipulate to confirm crossover design before building a passive xover, or even retain as the final build. But, without a passive crossover, there is no way to filter or attenuate the power amp output to try to lower noise floor.
I had wanted to use the M2 as my reference amp for the HF driver since some of my SET tube amps have considerable harmonic distortion and some of the compression drivers I'll be testing have very low harmonic distortion. I think the M2 will be better for my passive crossover speakers, or I'll use it for the woofer channel and see if the noise floor is low enough. I have had the M2 running moderately efficient full range cone drivers without too much noise being apparent. Or, as you've helped confirm, put in an L-Pad on the compression driver.
I am going to try a pair of Hypex NCore amps to test the active crossover bi-amp design. The measurements on the NCore technology are ridiculously good. Even if it's not the sound I want, they should be very good for measuring compression drivers. They would not be the bottleneck for harmonic distoration and I'm not sure they aren't approaching an ideal performance for ridiculously low cost.
Thanks again, Jamie
It'd be good to post some photos... remove the top cover... who knows, maybe the root cause would reveal itself.
Do you have other stereo gear placed close (left or right) / on top / underneath the M2 amplifier? Show us the whole setup.... magnetic field (from other equipment) doesn't care much about aluminium chassis... space is gold.
Swap M2 with any other amp.... it might be the crossover to blame in your case, NOT the M2.
Do you have other stereo gear placed close (left or right) / on top / underneath the M2 amplifier? Show us the whole setup.... magnetic field (from other equipment) doesn't care much about aluminium chassis... space is gold.
Swap M2 with any other amp.... it might be the crossover to blame in your case, NOT the M2.
A simple test as suggested by Nelson may be done to see if the Edcor transformer is affected by EMI:
M2 Edcor Transformer EMI Noise?
M2 Edcor Transformer EMI Noise?
Ben, it's not hum, it's noise - higher frequencies
at least as I understood
in any case - worth reminding of that old trick
at least as I understood
in any case - worth reminding of that old trick
Jamie, just to possibly rule out one more source of noise, what are the specs for your active crossover? Specifically, what is the noise spec for the active crossover's high pass channel?
Could you possibly share with us what the manufacturer/model number is for this crossover?
Could you possibly share with us what the manufacturer/model number is for this crossover?
Random thought...
I take it you've confirmed that the noise originates from within the amp and not further upstream like the frequency filter, preamp or source?
I take it you've confirmed that the noise originates from within the amp and not further upstream like the frequency filter, preamp or source?
Understood but EMI could be higher frequencies too, as harmonics of other frequencies, and from other sources.
I am not saying that it is the cause, only a potential cause.
I’m in agreement with Ben. In post #5 the OP says “I'm hearing it on a HF compression driver that is on an active crossover high pass above 1200 hz. . .” It seem that the speaker isn’t capable of outputing the 60/120Hz acoustically, just the harmonics. This is assuming the LF is on a different amp and not just using the other channel (at least how I read it). You could put a scope on the amp’s output, but shorting the autoformer output seems like an easy enough and pretty definitive test.
olen
olen
Maybe OP should be guided where to do the grounding on original M2 PCB?
Don't know how experienced OP is regarding such a test. The linked schematic is for M2x and resistor / capacitor numbering is probably different than on original M2 boards?
Don't know how experienced OP is regarding such a test. The linked schematic is for M2x and resistor / capacitor numbering is probably different than on original M2 boards?
all what's needed and exactly enough, to test for noise with amp's RCAs with shorting plugs
Ok, and then M2 only connected to the mid horn speaker (with active crossover out of the chain).
For those of us DIY enthusiasts with the time and energy to build our own M2x, here is something that works really well. (Documented in the main M2x thread)
Dual-mono PSU in a 400mm deep chassis - CLC w/ synchronous rectification
Power transformers of the “right” size to keep best distance from the Edcor signal transformers
Added a high Mu wrap around the center part of the Edcors
Dual-mono PSU in a 400mm deep chassis - CLC w/ synchronous rectification
Power transformers of the “right” size to keep best distance from the Edcor signal transformers
Added a high Mu wrap around the center part of the Edcors
Attachments
These are all really great suggestions for regression testing and proving out the source of the noise, but the combined evidence of the published noise of the M2 and Nelson's comments in forums and some posts here by Zen Mod, have convinced me that the noise is coming from the output transistors and is intrinsic to this zero negative feedback design with these particular silicon. I have owned this amp for a long time now and used it in many different configurations. I can't say my listening has been well done critically and controlled well enough, but my gear and experiments tend to be with minimal electronics and very sensitive drivers. For example I have used this amp happily with vintage full range RCA SL-12's which are legendary and revealing. In that arrangement, some power supply hum or ground loop potentially from other devices, although very few in the source chain, only a phono pre-amp, would occasionally appear. The SL-12's are maybe 93 to 95 db efficient.
The active crossover by Xkitz has been reviewed by Joseph Crowe with well documented measurement techniques and I'm also convinced it is not injecting noise. I think I've probably already run the amp directly into the drivers while doing low spl frequency sweeps prior to working with crossover.
The noise I here on a very sensitive compression driver is very consistent with the noise specs and other technical characterizations provided by Nelson and other folks here. I was very prepared to consider doing mods to the power supply or other shielding, isolation or ground loop cures, but I just think the evidence is overwhelming that those are not the source of the noise, or that the time and effort of trying those fixes cannot be justified by good amp design theory. Certainly worth a try if easier to do or not requiring modification of the factory built amp which has performance characteristics which are special and unique and can be embraced and enjoyed easily just by avoiding direct drive of 100db+ sensitivity drivers without filter or other noise floor mitigation.
Thanks so much to everyone for such a great education about this amp.
Jamie
The active crossover by Xkitz has been reviewed by Joseph Crowe with well documented measurement techniques and I'm also convinced it is not injecting noise. I think I've probably already run the amp directly into the drivers while doing low spl frequency sweeps prior to working with crossover.
The noise I here on a very sensitive compression driver is very consistent with the noise specs and other technical characterizations provided by Nelson and other folks here. I was very prepared to consider doing mods to the power supply or other shielding, isolation or ground loop cures, but I just think the evidence is overwhelming that those are not the source of the noise, or that the time and effort of trying those fixes cannot be justified by good amp design theory. Certainly worth a try if easier to do or not requiring modification of the factory built amp which has performance characteristics which are special and unique and can be embraced and enjoyed easily just by avoiding direct drive of 100db+ sensitivity drivers without filter or other noise floor mitigation.
Thanks so much to everyone for such a great education about this amp.
Jamie
It's the Xkitz 2 way. I don't have the noise specs handy, but Joseph Crowe did a critical evaluation including measurement and did not find issue. I cannot 100% say I regression tested the crossover, but I am pretty sure I did my first frequency sweep testing of the CD and horn without the Xover to make sure I characterized the CD response without a crossover. I might have had a senior moment, but I don't think the swap in of Xover or other elements is the cause.https://josephcrowe.com/blogs/news/xkitz-2-way-active-crossover
