Microphony considerations - test

Hi Dave,

Leaving CD players aside for a moment, what exists inside a Naim pre or power that is so different to the inside of the Q34, other than the power supply is external?
I don't see what would be gained or why they are a special case.

I anyone wants to loan a Naim pre I'm happy to run a test for you.
Then again they can do it themselves - this is hardly difficult!

 

Beats me why a 34 wouldn't exhibit microphonics but a 72 or 102 would. Then again, you're asking the least technically oriented individual on the planet as well.

 

No idea how this idea that the olive naim stuff was microphonic came about, certainly the 72/hicap/90 I had weren't.

 

...Through accidental discovery in our shop when I ran out of stands.

Placing the gear direct to carpet sounded far better than any stand we had in stock. Further experiments found subtle and not so subtle differences between many surfaces.

 

The reason gear like a 72 is microphonic is because of the plug in cards, try waggling one and hear the noise as your drivers fly across the room if you have a poor contact between card and pins. The Naim gear that isn't based on plug in cards shouldn't be microphonic, just like any other solid state amp that isn't a phonostage shouldn't be.

TT's and cd players are a whole different ball game.

 

Making and breaking an electrical connection by wiggling cards around isn't quite the same thing IMO. I'll borrow a Naim at some point and we can test those if there is interest. There is clearly some.

I've now re-done the Quad tests with some interesting results using drum tests at 107dB @ 20cm distance without cases. All posted within the hour.

 

Exactly, not the same thing at all.

 

Seconded...in my case none of the preamps had phono cards installed and in every instance, electrical connections were cleaned and reset after discovery for repeated tests. At the shop, sources were Naim, Linn, Meridian, Krell and Exposure CDPs. The same effect was heard with every trial. Sonic colorations were unique and repeatable with each surface. This was using various Naim preamps and power amps at the shop as well as different Naim preamps and power amps at home and at friends with both passive and active systems. What else could it be besides microphonics? (please, no jokes about "imagination", let's keep this on track.)

 

Yes, of course. However, surely it is not impossible that the rubber foot mounting actually contributes to the issue - isolating at some frequencies and coupling at others?

A moderate rap with ones knuckle is all that is required - no damage is done to anything. My point is to show that solid state circuitry is not quite as innert as some (you for example) might like to believe.

 

If you need a NAIM pre I can help.

 

Jolly good - I look forward to the results.

Note: those who do believe in microphonic effects often blame the case, so it might be worth trying it with the case fitted.

 

Part 2

In response to some of the points raised the first test is repeated here using a drum track as requested, and the casing removed as requested.

Before going any further I'd suggest downloading and familiarising yourselves with the test track:

http://www.mediafire.com/file/0jevlyubb7cfky8/02 Improvisation.mp3

The test was repeated with the same Q34 but this time without the outer steel casing. The component side of the pcb was placed sqaure onto the bass driver of the loudspeaker at a distance of 20cm. So a little closer than before and the transient nature of the test track meant that SPLs of 107dB could be produced at the test position. See below:










Set to line as per the initial test and full gain on both pre and recorder, these are the results obtained:



Sound recording:

http://www.mediafire.com/file/urr0tbwcuwmpp6r/drum micro LINE.wav

If you then amplify the WAV as per the first test what you get is this - still residual circuit noise (ignore the two spikes which are static ticks from the exposed circuits):




So now lets look at circuits with far higher gain.
I inserted a high gain MC card into the Q34 - 200uv of the type suited to low output MCs. This gives overall 60dB of gain for the stage IIRC.
No change to test conditions gives this:



This is the sound recording:

http://www.mediafire.com/file/yjxbnk4in46c7op/drum micro MC.wav

Gets interesting now. You can just about hear something down in the noise, more at HF a little further into the track. I'll leave you to determine any significance on this one.

Now it gets very interesting - tests with the WD Valve MC stage.

Same test method as before except that the output is taken directly into the Macbook and full gain applied. This is 3 tube (ECC83) stage similar to the EAR circuit. The active stage is MM and 12:1 ratio Sowter step-ups are in place to match low output MCs.
it is built to the highest standards I could conceive short of a fresh design, so high end fast diodes in the PSU, Balck Gate PSU caps, NOS Mullards, AudioNote paper in oil caps etc. In comes in two boxes with a steel perforated cover as seen in the pic.




This is what we get:



This is the sound file:

http://www.mediafire.com/file/a0r7ppmpchg4sv7/valve micro.wav

In summary I would suggest:

- Valves can be microphonic as expected and that is hardly news. You do however need to consider the extreme test condition used, because if the unit is placed a couple of meters away from the sound source and a normal listening level is used, the trace and audio recording sink to zero. Performance would seem largely dependent on the tube used as other stages are clearly very microphonic. You can also see from this test that a slightly different result was obtained for the two channels - again not unexpected with valves.

- High gain SS phono stages are microphonic, but to micro degrees, if that makes sense.
The output, given the massive amount of gain applied here is remarkably tiny. Put the case back on and move the unit only a couple of feet away and again you have flat line traces and nothing on the recording.

- Lastly the line stage. You can make up your own minds but I for one will never be the slightest bit concerned about these things, or having such electronics inside the loudspeaker cabinet.

 

Exactly what I was expecting, but I'm sure there will be the exclamation 'but i can hear a difference when I...' from the usual suspects.

ps,
I think you deserve a well earned rest with a 'tasty beverage' of your choice!

 

Sounds broken to me!

Who makes this amplifier?

On the first point, yes quite true but you cannot then get the pcb face onto the loudspeaker, and I want this to be an extreme test.
Bit like testing a rope. If you stress test it pulling 20 tonnes then it will have no difficulty pulling 2.

 

LOL - thanks, I'm now putting my living room back into some sort of order and will be opening a bottle of wine, then listening to some music on my hopelessly dull Quad amplifier :beer:

 

Rob, you'd be doing us all a favour if you tried it with the case in place- a good test to see if vibrating cases effect the sound, I'm thinking eddy current myths here...

Great stuff so far.

Re the knuckle wrapping Ynwan mentions- yeh but thats with plug in cards in the amp, they'll always squeal

 

I agree with your conclusions here. On the line input if you crank headphones up to pretty insane levels it is just audible, and the MC stage is surprisingly close. The tube amp was amusing as you can now clearly hear what is going on, i.e. what was a very subtle rhythmic excitement of background noise turns into a very clearly identifiable kick drum, toms, snare, hi-hat and cymbals. I deliberately listened to them in the order posted, i.e. not giving myself the clue as to what track I was listening to on the Quad. By the time I got to the tube amp I easily recognised it as a track you'd played me in the past through the Kensais.

How many db down do you think the Quad noise floor is, i.e. how far below a typical line level input? Certainly based on this test the effects of microphony appear to exist around the same level of residual mains hum etc, i.e. way down on what I'd expect to hear in room.

Very interesting IMO, and thanks for taking the effort to record and upload this stuff.

Tony.

 

I very much doubt it would make any difference at all, if placing the circuit board inches from the speaker clearly has no effect.

Sorry to be blunt, but that's just nonsense, as I mentioned as few post ago, my 72/hicap/90 produced nothing from the speakers when the cases were clouted.

 

Well...one thing's for sure.

You can plop your gear down on the carpet and if you hear a difference -make adjustments in what you use for a stand if you like the difference. If you hear no difference, do nothing. If you prefer the sound with your stand, return it to your stand and you've still spent nothing. If you believe this is a load of shite, do nothing and you'll gain nothing. Sounds like a win-win to me regardless.

 

Tony, the noise spec on the old 34 (don't have figures for the later one) on line are -88dB, so I'd say better than -90dB for a later one with reduced gain such as this.
Incidentally, the 34 line stage applies 32.5dB of gain at maximum setting as used here, so that needs to be considered along with any gain being applied by max setting on the Macbook analogue input.

I'd like to look a eddy currents and I'll include it in the next round, perhaps with the Naim or any other unit.
The first test was made with the case on and although noise was used, it was bloomin loud and wide band - so I'd expect to see something on the traces/audio.

Interesting to note that the overall residual noise levels went up by about 5dB with the casing removed from the Quad, so the metalwork was also doing something beneficial in the screening department.