• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Noisy Purifi amplifiers

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,026
Likes
4,959
Location
Europe
Whose wrong/recommended practice is that? When you look around there seems to be significant difference of opinion.

I also suspect that which works better will depend significantly on the specifics of equipment grounding design.
AES 48 is not considered as an opinion :).
 

Chester

Senior Member
Joined
May 22, 2021
Messages
457
Likes
1,131
Since investigating further, the left channel of my Purifi has more hiss than the right. Neither is terrible but am interested in seeing if there’s anything I can do to remove it. Source is an RME ADI-2 Pro and speakers are Revel F208’s. Here’s a pic inside the amp, it’s a Nord. Can anyone spot anything inside that looks like a potential cause of this?

Again, appreciate this is just hair splitting differences, I’m a recovering audiophile though so the urge to tinker is still strong :D

Thanks
 

Attachments

  • 8BDF680F-C433-45E3-8ABE-EEC724B6998B.jpeg
    8BDF680F-C433-45E3-8ABE-EEC724B6998B.jpeg
    303.8 KB · Views: 124
  • 1C5110D0-204D-4B34-B280-9C92A8A355FD.jpeg
    1C5110D0-204D-4B34-B280-9C92A8A355FD.jpeg
    361.1 KB · Views: 118

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
8,170
Likes
13,888
Location
UK/Cheshire
AES 48 is not considered as an opinion :).
I admit to being unfamiliar with the standard - is it mandatory? Will compliant cabling always give the best result even if equipment grounding is not compliant?
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
8,170
Likes
13,888
Location
UK/Cheshire
Since investigating further, the left channel of my Purifi has more hiss than the right. Neither is terrible but am interested in seeing if there’s anything I can do to remove it. Source is an RME ADI-2 Pro and speakers are Revel F208’s. Here’s a pic inside the amp, it’s a Nord. Can anyone spot anything inside that looks like a potential cause of this?

Again, appreciate this is just hair splitting differences, I’m a recovering audiophile though so the urge to tinker is still strong :D

Thanks
Id say the most likely cause is coupling from the power supply to the left hand module (is that the one with higher noise?).

Possibly from that ribbon cable routed pretty much in contact with the module. If that is the noisier module I'd start with trying to space the ribbon cable as far from the module as possible. Then the thick black power lead.

After that, I might have to start gettting inelligent and working out what the cables carry - and what circuit elements they are in proximity to. (depending on this routing the ribbon cable closer to the power lead might make things worse)
 

JSmith

Master Contributor
Joined
Feb 8, 2021
Messages
5,418
Likes
14,153
Location
Algol Perseus
Can anyone spot anything inside that looks like a potential cause of this?
Unsure if this is the angle of the picture, but is that speaker output line in contact or very close, circled in red?

1643796271445.png


JSmith
 

Chester

Senior Member
Joined
May 22, 2021
Messages
457
Likes
1,131
Id say the most likely cause is coupling from the power supply to the left hand module (is that the one with higher noise?).

Possibly from that ribbon cable routed pretty much in contact with the module. If that is the noisier module I'd start with trying to space the ribbon cable as far from the module as possible. Then the thick black power lead.

Thank you, yes that’s right, it is the left channel that is noisier.
 

Chester

Senior Member
Joined
May 22, 2021
Messages
457
Likes
1,131
@JSmith Thanks, I don’t have it open at the moment but from other pictures I took previously, it certainly looks very close.
 

Attachments

  • BCE04F93-AC2F-4706-993F-47F98B6C3619.jpeg
    BCE04F93-AC2F-4706-993F-47F98B6C3619.jpeg
    337 KB · Views: 95
  • A0F76919-C45F-4EB7-BACD-B973B766E780.jpeg
    A0F76919-C45F-4EB7-BACD-B973B766E780.jpeg
    393.2 KB · Views: 95
  • 80E45FE0-0352-46EE-974B-437185576FB4.jpeg
    80E45FE0-0352-46EE-974B-437185576FB4.jpeg
    307.3 KB · Views: 93

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,452
Likes
7,992
Location
Brussels, Belgium
Purifi specify 11.5uV A-WTD.

No power amplifier is 'dead silent', especially when in close proximity to tweeters.
Truth be told most rooms' noise floor would drown any tweeter hiss when you're using a 120 db S/N chain. Unless you mean in anechoic condition but that's a hyperbole on its own.

If i remember correctly the 11.5uV specification is only possible with academic grade power supplies according to @boXem | audio
 
Last edited:

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,026
Likes
4,959
Location
Europe
I admit to being unfamiliar with the standard - is it mandatory? Will compliant cabling always give the best result even if equipment grounding is not compliant?
It is not mandatory in terms of "required by law'. It specifies how balanced connections should be done to avoid problems. Pro world follows it. On the non-pro side, there are more artists for whom following a standard is a limitation to their art ;).
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
13,010
Likes
39,903
Location
Gold Coast, Queensland, Australia
Truth be told most rooms' noise floor would drown any tweeter hiss when you're using a 120 db S/N chain. Unless you mean in anechoic condition but that's a hyperbole on its own.

If i remember correctly the 11.5uV specification is only possible with academic grade power supplies according to @boXem | audio

Don't confuse S/N with audibility of residual noise. The more 'signal' you have, the greater you can lift that signal above the noise. Wind up the voltage swing (power) and even an average to poor residual noise gets an elevated S/N number for 'free'.

My power amplifiers have a 120dB S/N and I can hear residual in the tweeters from a foot away at zero volume in a dead silent room. The residual is less than 35uV (A).
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,452
Likes
7,992
Location
Brussels, Belgium
My power amplifiers have a 120dB S/N and I can hear residual in the tweeters from a foot away at zero volume in a dead silent room. The residual is less than 35uV (A).
And the gain at which the amplifier is set results in how much maximum output? because if it's set to 100 dBSPL at maximum input then the noise floor should be around -20dBSPL which should theoretically be inaudible. in normal room conditions where the noise floor could be anywhere from 10 to 40 dBSPL even at high frequencies.
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
96
Likes
321
Purifi specify 11.5uV A-WTD.

No power amplifier is 'dead silent', especially when in close proximity to tweeters.
With "dead silent" I mean that both 11,5uV and 20uV will be impossible to hear even with the ear very close to a standard dome tweeter.
Our purifibased amplifiers Vera P150/600 has a noise floor of 18uV measured on the speaker terminals. On my 92dB/W speakers it is below my hearing threshold even with my ear 10mm from the tweeter. That is how silent these modules are. I had to borrow a monster 7 kilo compression driver with a sensitivity of 108dB/W to be able to hear a faint hiss with my ear very close.

When people have noise problems it is in almost all cases caused by the source. Note also that simply removing the input will increase the amplifier noise. The only way to proper evaluate amplifier noise, is to short or terminate the inputs with 100 Ohm
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
8,170
Likes
13,888
Location
UK/Cheshire
It is not mandatory in terms of "required by law'. It specifies how balanced connections should be done to avoid problems. Pro world follows it. On the non-pro side, there are more artists for whom following a standard is a limitation to their art ;).
Thanks.
 

Chester

Senior Member
Joined
May 22, 2021
Messages
457
Likes
1,131
With "dead silent" I mean that both 11,5uV and 20uV will be impossible to hear even with the ear very close to a standard dome tweeter.
Our purifibased amplifiers Vera P150/600 has a noise floor of 18uV measured on the speaker terminals. On my 92dB/W speakers it is below my hearing threshold even with my ear 10mm from the tweeter. That is how silent these modules are. I had to borrow a monster 7 kilo compression driver with a sensitivity of 108dB/W to be able to hear a faint hiss with my ear very close.

When people have noise problems it is in almost all cases caused by the source. Note also that simply removing the input will increase the amplifier noise. The only way to proper evaluate amplifier noise, is to short or terminate the inputs with 100 Ohm

I quite fancy trying your XLR shorting method you mentioned earlier in this thread, to see what noise is then audible, but as a total novice in this area, sticking wire into two holes and turning it on just feels like it will end in tears.
 

tmtomh

Major Contributor
Forum Donor
Joined
Aug 14, 2018
Messages
2,844
Likes
8,421
My speakers are medium efficiency (87 db/2.83V/1m), and I cannot hear any sound whatever from them with the Purifi hooked up, even with my ear close up to the tweeter.

Same - my speakers are rated at 89dB efficiency and I can't hear anything at all with my Audiophonics Purifi-based amp. ear right at the tweeter.

I did hear a very low-level hiss when I had a ground loop in my system, though, and the removal of all hiss from the amp with the ground loop solved did surprise me, since I always thought the ground loop would only produce 50/60Hz hum (and perhaps low-order harmonics thereof).

I suspect some, perhaps a lot, of the variability in hearing or not hearing hiss could be the presence or absence of ground loops in the system, particularly since the other components do not actually have to be turned on (just plugged in to AC) in order for the effect to happen.
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
96
Likes
321
I quite fancy trying your XLR shorting method you mentioned earlier in this thread, to see what noise is then audible, but as a total novice in this area, sticking wire into two holes and turning it on just feels like it will end in tears.
It's safe
 

JaapDeventer

Member
Joined
Oct 29, 2018
Messages
31
Likes
34
I had a hiss with my first DIY AES EBU cables. A different cable with more shielding did much better. Strangely a pair of rca converters and rca cable was dead silent.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,366
Likes
1,135
Since investigating further, the left channel of my Purifi has more hiss than the right. Neither is terrible but am interested in seeing if there’s anything I can do to remove it. Source is an RME ADI-2 Pro and speakers are Revel F208’s. Here’s a pic inside the amp, it’s a Nord. Can anyone spot anything inside that looks like a potential cause of this?

Again, appreciate this is just hair splitting differences, I’m a recovering audiophile though so the urge to tinker is still strong :D

Thanks

Is it perhaps the channel closer to the power supply?
 

amarsicola

Active Member
Forum Donor
Joined
Dec 6, 2019
Messages
125
Likes
356
Location
Rome, Italy
...after opening the amp that feeling turned out to be right. It was set to "Configuration 2" for approx. 2.25V RMS. So I removed the gain jumper to get to the 4.2V RMS configuration. And gone was all the noise. Like magic. :)
...
Happy for you that noise is gone away, still I am now very skeptical about Audiophonics implementation.
Why should the 2.25V configuration be noisy?
It is not to be justified according to me
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
8,170
Likes
13,888
Location
UK/Cheshire
Because the gain is doubled, and hence any noise (internal from the amp, or coming from the source) is also doubled.
 
Top Bottom