Will a slight reduction in gain impair sonics on AKSA 55?

Jens · « **on:** 9 Feb 2003, 02:38 pm »

Hi Hugh (or any other member with relevant input),

Since my speaker system is very efficient (96 db/m) I thought I might reduce the inherent background noise of my AKSA 55 Nirvana by reducing gain by 2-3 dB.

Would there be any detrimental effects on sonics of a reduction of this magnitude?

My AKSA is nearing its cosmetic completion (it's been fully operational and playing for about a month now); the 10 mm alu front is being anodised (black) and should be ready next week. It has a well in the middle where I will mount a couple of orange (valve glow

) LEDs. Will send photos when it's ready.

Cheers,

Jens

PSP · « **Reply #1 on:** 9 Feb 2003, 04:52 pm »

Hi Jens,
How much noise are you hearing from your 55w AKSAs? My speakers are 94db, and the buzz and hum I get are audible only within 20-30cm of the drivers; it is absolutely inaudible at 2-3 meters.

Can you put a shorting plug on the AKSA amp inputs? (shorting plug = RCA with a resistor soldered across hot and ground, resistance value somewhere around 50-200 ohms... I'm not sure why a total short would be a problem here though, but gurus usually recommend some small R>0)

so back to the amp... with a shorting plug in place (thus eliminating any noise from sources, preamps, interconnects, etc.) you are left with the possibility of noise from the amp itself, EMI from the surroundings, and/or noisy mains.

If you are certain that your amp is noisy, consider the following:
- lightly twisted signal leads, short and well away from AC. You could also try some good coax here, but I'd see that as a near last resort

- experiment with running a wire (or not) from Single Point Ground to chassis ground... one configuration may be quieter

- keep torroidal transformers seperated from the AKSA amp boards

- look for poor solder connections (I use an illuminated magnifying glass, mounted to my workbench, to be sure that all solder joints look good). Bad solder joints, however, usually sound distorted or harsh... but I suppose that a bad solder joint in the signal ground could lead to buzz or hum...

If these simple ideas don't solve the problem, post back and Hugh + the high gurus here will help with some more ideas. I've been reading "AKSA posts" since nearly the beginning, and I can't remember many (any at all?) "noisy amp" posts.

Good luck,
Peter

AKSA · « **Reply #2 on:** 9 Feb 2003, 08:44 pm »

Hi Jens,

Thank you for your post.

Peter's advice is the best; short the input on your AKSAs (there is no danger of hurting the amp since there is a DC blocking cap on the input) and test for noise with speaker connected. There should be none.

Once confirmed, move to step two, which is reconnecting the input signal wires, which stretch across the amp from the modules to the RCA sockets at the back panel. Remove the input short, so that from the RCA sockets at the back panel to the amp inputs the wires are now in the testing loop.

These wires are quite high impedance - around 43K - and are thus susceptible to hum intrusion from straying magnetic fields.

You may find that these wires are the culprits. If so, then the problem is likely magnetic inductance from the toroids, and the best way to eradicate it is to place a ferrous shield between the wires and the toroids.

Reducing the gain of the AKSA will certainly change the sonics, though not by much. Things become rather tighter at the expense of musicality.

Hope this is helpful

Cheers,

Hugh

Jens · « **Reply #3 on:** 9 Feb 2003, 08:48 pm »

Hi Peter,

Thanks for your thorough input.

I'm fairly sure there's nothing wrong with my AKSA. Hugh built the modules = absolute no bad soldering, and all the wiring tricks you mention are in place. Besides the signal leads are only between 15 and 20 mm long from board to socket. There is a very slight amount of hum, but that will be eliminated with an interior "wall" of magnetic iron between the toroids and the boards. I already have the wall, which will be installed when I have to disassemble a few things to install the new front. Anyway, the hum is so slight that it doesn't bother me. However, when it comes to hiss, the AKSA is just inherently a little more noisy than most of-the-shelf amps. Even with the preamp turned completely off or the inputs shorted there is still an appreciable amount of hiss.

So, when you have a preamp with a 12 dB gain (I'm going to change this as well - to 6 dB - or perhaps I'll get a GK-1) and very efficient speakers, the hiss is sometimes just discernible in the listening position.

Mind you, I'm not criticising the AKSA (you know that, Hugh!) - just trying to see if I can solve this very minor problem by tweaking things ever so slightly!

The AKSA is just one great amp (the best I've ever had) - and when I get the enclosure completely finished I'm going to take on a tour to visit a few friends to show them how much better an inexpensive DIY amp sounds than their ordinary amps that have cost at least twice as much!

Cheers,

Jens

Jens · « **Reply #4 on:** 9 Feb 2003, 08:57 pm »

Hi Hugh,

As you can see in my reply to Peter, the hum is very slight, and a magnetic shield is already waiting to be installed.

It really the hiss I'm after, but I will definitely not tamper with the gain if it affects sonics!

Perhaps the change in preamp gain will do the trick

Cheers,

Jens

PSP · « **Reply #5 on:** 9 Feb 2003, 09:23 pm »

Hi Jens,
Hiss... no ideas come to mind. Do you sit really close to your speakers? ....or perhaps the noise floor in your listening room is particularly low?

Some really random thoughts...
- are you using a line filter on the mains? If not, you could build the Jon Risch power line filter (or put in the first three caps, but not the inductors) across your power line after the power switch inside the AKSA. If you do this, be sure that the caps are rated for mains duty.

- are you using a power-on light (I do, blue LEDs, but I haven't been able to hear any additional noise when they're in the circuit)? If so, perhaps you can temporarily disconnect them and give a listen...

- dirty power sometimes gets quiet late at night when people have most of their stuff turned off... see if your hiss is present deep in the night or on weekends (of course the noise floor might be lower too).

- ferrite clamps between the AKSA power supply board and the AKSA amps?

- snubbers on the power supply diodes?

- is your AKSA in a metal enclosure? Is the hiss sensitive to amp position? I'm thinking EMI here....

Honestly, I'm just shooting in the dark here. Hugh???

Good luck, keep on chuggin'

Peter

AKSA · « **Reply #6 on:** 9 Feb 2003, 09:47 pm »

Hmmmmmmm.... no, that should be hissssssssssssss......

Jens, this is interesting. Haven't noted any hiss in my installations, and can only conclude the following possibilities:

1. Very sensitive tweeter.
2. RF finding its way into the amp.
3. A noisy semi in the modules.
4. Hash from the power supply
5. High expectations with such a sensitive speaker.

Questions:

6. Is it the same for both channels?
7. If you shorten the speaker wire to maybe 30cms and listen for it, is the amplitude the very same?
8. Does it change at all according to time of day?

The AKSA has been praised by many for very low noise; it is possible there is some intrusion from the power supply or RF. OTOH, it could be that the very sensitive tweeters exaggerate a problem normally subliminal.

Dunno. We'll talk more.

Cheers,

Hugh

blizzard · « **Reply #7 on:** 9 Feb 2003, 10:13 pm »

Hi Jens,
I gotta back Hugh up on this. I have a GK-1 and an AKSA 55-Nirvana. Even with the GK-1 at full gain, there isn't even the slightest hint of hum or hiss.
As compared with your 96db speakers, mine are only 90db efficient. So, I don't know if this is a fair comparison or not. But, I would guess that your hiss problem can be remedied.
Good luck,
Steve

Larry · « **Reply #8 on:** 10 Feb 2003, 12:59 am »

Quote from: AKSA

These wires are quite high impedance - around 43K - and are thus susceptible to hum intrusion from straying magnetic fields.

Hi, Hugh,

I am interested in this. Can you elaborate a bit on this? How is this related to the input impedance?

AKSA · « **Reply #9 on:** 10 Feb 2003, 01:47 am »

Larry,

Any high impedance wiring is susceptible to magnetic coupling from nearby magentic circuits, such as tranformers, which create weak currents in wires in close proximity. These magnetic flux lines are all the more influential in nearby wires if they are terminated at a high impedance circuit.

If the target circuit has an input impedance of say 100R, then the weak currents induced by nearby magnetic fields will be greatly reduced in magnitude because they are heavily loaded. Thus, low impedance = low hum susceptibility.

Going in the other direction, we have the very high input impedance of the grid circuit on the GK-1, of the order of several megohms. The wiper wire is extremely susceptible to hum intrusion, since there is little to load these weak currents and they are amplified without restraint.

Hope this helps!

Cheers,

Hugh

Larry · « **Reply #10 on:** 10 Feb 2003, 03:45 am »

Quote from: AKSA

Going in the other direction, we have the very high input impedance of the grid circuit on the GK-1, of the order of several megohms. The wiper wire is extremely susceptible to hum intrusion, since there is little to load these weak currents and they are amplified without restraint.

Hi, Hugh,

Thanks. It certainly helps.

I will revisit this when I properly wire my GK-1 up in its permenant case. (Everything for the case ready except awaiting your release of shaft extension

) I believe the slight intrusion in my GK-1 is due to messy wiring, as discussed earlier, in my temporate setup.

So, in this case, at sensitive position like that between wiper and grid input, do you recommend coaxial cable?

Is the 43K you quoted ASKA 55's input impedance?

AKSA · « **Reply #11 on:** 10 Feb 2003, 04:22 am »

Hi Larry,

Yes, you betcha, the 43K is the input impedance of both the 55W and 100W AKSAs. This is deliberately high, to accord with the common input impedance of tube amplifiers, which which these amplifiers compete, and I wanted the same input conditions to reduce the loading on the preamp.

The extension shaft is now being made and on schedule. I'm having 200 made to begin with. It's all mild steel, no aluminium, for robustness, and uses a bright, attractive coating. I will offer these to the general public as well, not just Stage 2 GK-1 kit builders, as I see here a need for a good product at a reasonable cost - these things are often required in the DIY community.

Cheers,

Hugh

Jens · « **Reply #12 on:** 10 Feb 2003, 11:01 am »

Hi Hugh,

As I said, the "problem" is very minor, but if something can be done, then ...

Quote from: AKSA

Hmmmmmmm.... no, that should be hissssssssssssss......

Jens, this is interesting. Haven't noted any hiss in my installations, and can only conclude the following possibilities:

1. Very sensitive tweeter.
- True, my Scan-Speak Ring Radiator is very sensitive and has a very extended frequency (beyond 40 kHz)

2. RF finding its way into the amp.
- Could be, but I generally don't have much of this. Could try RF-stopper on the speaker cables?

3. A noisy semi in the modules.
- Dunno, could you specify?

4. Hash from the power supply
- How?

5. High expectations with such a sensitive speaker.
- Definitely true!

Questions:

6. Is it the same for both channels?
- Yep

7. If you shorten the speaker wire to maybe 30cms and listen for it, is the amplitude the very same?
- Will try that one

8. Does it change at all according to time of day?
- Nope

The AKSA has been praised by many for very low noise; it is possible there is some intrusion from the power supply or RF. OTOH, it could be that the very sensitive tweeters exaggerate a problem normally subliminal.
- I believe that is true.

Dunno. We'll talk more.

- Perhaps the change in gain in my preamp will do the trick, but I'll investigate some of the above possibilities. As always you've been most helpful.

Thanks, Hugh.

Cheers,

Jens