Help!! Irritating sound of my GainKlone

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I had my Gainklone mono blocks up and swinging. Refer to my last post.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=15425

Although, sound produced by it is ok; no hiss, no hum, quiet and with good bass, I do find it very irritating, not at ease at all, I just simply cannot immerse into the music.

After having the amps up and running for two weeks, I decided to change the value of resistor (in red) from 49.9kohm to 30.1kohm, Thus the gain increases from 13.6 to 22. After that, the sound is still very irritating, I would say no improvement at all.

I have also added two pairs of Rifa X2+WimaFKP1 cap before and after the rectifier, and it still sounded very inrriating.

What else can I do to improve this particular aspect of the amps? Can anyone share their experience with me? Would change to the inverting configuration improve the situation?

Thanks
SW Yeo
 

Attachments

  • gainklone.gif
    gainklone.gif
    9.9 KB · Views: 1,020
Hi!

You can try omitting the ERO MKP (the input cap) from the circuit (but measure DC offset prior to connecting speakers... probably anything up to 100 mV will be OK).

Or try substituting that input cap with a better one (I also sometimes use those ERO, but never as an input cap so far).

Also the cap from pin 1 to 4 is probably not necessary, try omitting it as well (but always carefully try out if the amp remains stable).

Also, what speakers do you use?

Bye,

Arndt
 
Re: With a PSU like that

ALW said:
You'll have ripple measuring volts I would think.

Try some adequate reservoir capacitance for starters

Andy.

That should not be necessary if you follow the GainClone discussions.

2 * 1000 uF directly at the pins of the IC should be sufficient, and some small caps directly at the supply should also suffice for stabilizing the voltage through the supplies cables...


Bye,

Arndt
 
overmind,

From the picture, the PSU cables look too long, and too thin...

The feedback resistor seems too high at 631K - try reducing it to 220K or even 120K (and changing the 49.9K so that the ratio of them is the same to maintain the gain)

Consider trying two seperate bridges, and locate them nearer to the amps. And try a pair of torroidal transformers - I found this made a big difference.

Try removing all the small PSU decoupling caps, leaving just the 1000uF's on the chip itself. You can alway put them back in.

Have you seen my WWW? Also check Nuuk's pages. Both of us have built inverting GainClones - have you considered trying an inverting version?

Cheers,

Mark ;)
 
Re: Re: With a PSU like that

I have tried adding two Nichicon KG 15000uF 63V and found that it made the sound more tense, the bass is also tighter. The difference was not big, but still do not like it. I had them removed.

Yeo

ALW said:
You'll have ripple measuring volts I would think.

Try some adequate reservoir capacitance for starters

Andy.


Cradle22 said:


That should not be necessary if you follow the GainClone discussions.

2 * 1000 uF directly at the pins of the IC should be sufficient, and some small caps directly at the supply should also suffice for stabilizing the voltage through the supplies cables...


Bye,

Arndt
 
Overmind,
What's the VA rating of that transformer?
It seams small to me.
Yes, 1000uf is good for a Gainclone, but you're supposed to have a good transformer.
Or else you'll have voltage variations that do no good for sound.
I didn't try the non-inverted configuration (and I'm not thinking in trying), so I can't say if that's the cause.
My Gainclone (inverted) sounds fantastic.:nod:
 
Cradle22 said:
Hi!

You can try omitting the ERO MKP (the input cap) from the circuit (but measure DC offset prior to connecting speakers... probably anything up to 100 mV will be OK).

Or try substituting that input cap with a better one (I also sometimes use those ERO, but never as an input cap so far).

Also the cap from pin 1 to 4 is probably not necessary, try omitting it as well (but always carefully try out if the amp remains stable).

Also, what speakers do you use?

Bye,

Arndt

I will try parallel the ERO input cap with another ERO KP1832, and if it still does not improve the sound, then I shall by pass it totally.

I am using the amp with a pair of Proac 2.5.

mhennessy said:
overmind,

From the picture, the PSU cables look too long, and too thin...

The feedback resistor seems too high at 631K - try reducing it to 220K or even 120K (and changing the 49.9K so that the ratio of them is the same to maintain the gain)

Consider trying two seperate bridges, and locate them nearer to the amps. And try a pair of torroidal transformers - I found this made a big difference.

Try removing all the small PSU decoupling caps, leaving just the 1000uF's on the chip itself. You can alway put them back in.

Have you seen my WWW? Also check Nuuk's pages. Both of us have built inverting GainClones - have you considered trying an inverting version?

Cheers,

Mark ;)


This just a make shift PS. I will use a better transformer + wiring once it is voiced to my liking.

I do not understand the effect on the value of feedback resistor. How would it affect the sound it this application?

I might change the amp to inverting configuration. I just wonder does anyone actually compare the configurations. And is the diff big?

I will remove the power supply bypass caps and see how.


GNAKED said:
It proposes to exchange power cables for the better. Gainclone is sensitive on used conductors.

What wire would you recommend? I have some multistrand copper wire laying around. They appears to be much higher quality and of course thicker. Will try it.



I will try the changes and report back.

Thanks

Yeo
 
On spec, the capacity of the transformer = 2 x 22V x 4A = 176VA. The quality of most the local make transformers are not good, we can that a discount on this, let it 150VA.

This might be the catch - GOOD transformer. :goodbad:
Have anyone try with cheapo transformer like me? Any improvement after the transformer is changed?



carlosfm said:
Overmind,
What's the VA rating of that transformer?
It seams small to me.
Yes, 1000uf is good for a Gainclone, but you're supposed to have a good transformer.
Or else you'll have voltage variations that do no good for sound.
I didn't try the non-inverted configuration (and I'm not thinking in trying), so I can't say if that's the cause.
My Gainclone (inverted) sounds fantastic.:nod:
 
overmind said:
This just a make shift PS. I will use a better transformer + wiring once it is voiced to my liking.
The point that you might be missing here is that the power supply contributes significantly to the sound you hear! I was surprised at just how much difference the power supply makes...

overmind said:
I do not understand the effect on the value of feedback resistor. How would it affect the sound it this application?
It affects the impedance that the non-inverting input sees. It's good practice to maintain low impedances where possible for lots of reasons. It keeps noise to a minimum, for example. It means that bias currents from the inputs have a minimal effect. Try it - it might not make any difference to the sound you hear, but I'd be very interested to find out what you find...

overmind said:
I might change the amp to inverting configuration. I just wonder does anyone actually compare the configurations. And is the diff big?
As you might have seen, my prototypes offer me the option of trying both configurations. But, I have to confess that I have only tried the inverting version so far. This is partly due to a lack of spare time, but also because I am so impressed with the way it sounds right now. As above, try it and let us know...

overmind said:
I will remove the power supply bypass caps and see how.

What wire would you recommend? I have some multistrand copper wire laying around. They appears to be much higher quality and of course thicker. Will try it.

I'm not really a big believer in magic cables - decent quality multistrand is what I'd choose. My GainClones have no wire in them - the wires from the transformers attach straight to the PCB...

Hope this helps,

Mark ;)
 
originally posted by Cradle22
You can try omitting the ERO MKP (the input cap) from the circuit (but measure DC offset prior to connecting speakers... probably anything up to 100 mV will be OK).
Or try substituting that input cap with a better one (I also sometimes use those ERO, but never as an input cap so far).
Sorry overmind that I can`t give You any good hint what could be wrong with Your amp. (don`t have any Gainclones experience.....not yet) unless the changes mhennessy suggested already.

But what I`m sure is that those ERO MKP`s coupling caps are not responsible for an "irritating" sounding amp. This are very decent caps.... if taken to extremes, well.... there might be "better" ones but this ROE are not so "odd" to produce the effects in sound You described (IMO).
 
Koinichiwa,

overmind said:
Although, sound produced by it is ok; no hiss, no hum, quiet and with good bass, I do find it very irritating, not at ease at all, I just simply cannot immerse into the music.

<SNIP>

What else can I do to improve this particular aspect of the amps? Can anyone share their experience with me? Would change to the inverting configuration improve the situation?

From my experience with other chipamp's (TDA2020/2030/2040/2050) the difference between the inverting and non-inverting operation is quite drastic.

HOWEVER, I would first recommend that you add the 0.22 Ohm resistor in series with the Amplifier Output. Use a good quality TO-220 cased "Powerfilm" from Vishay, Caddock or others, under NO CONDITIONS use cheap "white coffin" wirewounds or Metal Oxide units.

Secondly - remove the Wima Bypass Capacitors across the 1,000uF PSU Capacitors. These Capacitors will form a high Q resonance circuit with the parasitic inductance of the 1,000uF Capacitor and their fairly large value will move this resonance quite low in frequency.

Add to this potentially instability caused by having neither output LR Blocker or the simple resistor (which usually sounds better) and you have the potential for a very strained and agressive sound.

Also, if you use series feedback (non-inverted operation) reduce the feedback resistor to 22k and the gainset resistor to ground to 820R. Your values are very high and may cause further problems as a result.

Finally, the two parallel capacitors before and after the bridge rectifier are SERIOUSLY bad news in my view. remove the 0.47uF Cap and place it after the Bridge to not loose it's use. Then connect a 0.1....0.22R Resistor in series with each wire to the Bridge from the transformer (white coffin wirewounds will do fine). Connect one each of your 10nF Capacitors before and after the resistors across the wires.

I would suspect that doing this will calm down the sound of the Gainclone a lot, but of course I cannot be 100% certain, so it's "suck'n'see" I suspect. Let us (me) know how you het on and how accurate my diagnosis turned out (from the effects of implementing the remedy).

Sayoanra
 
Overmind,
The advice Mark gave you should clean up your sound. I would suggest trying the amp without the PSU bypass film caps across the 1000uF electrolytics. These have either hurt or helped in some of my projects so it may be worth a try. Are you using silver plated wire or tinned in the signal path? That can add an edginess to your sound as well.

edit: Kuei got in there before me. Good advice all around.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.