Setting GAIN on GAINclones

[Lucas_G]

Hi Folks,

I have build two 3886 monoblocks with the boards designed by BrianGT from www.chipamp.com

I used all the values as described on their site.

Now my gain is much too high.

I tried various preamps, but I can hardly move the volumecontrol or everything is becoming much too loud.

What is the best way to lower down the input sensitivity while still maintaining the good soundquality?

I considered adding a 10 kOhm resistor at the cinch-connectors...

In the manual it is stated that one schould increase R3 to lower gain. Is that a better method? How much higher should I go?
(National Semiconductor has a value of 1 kOhm in their circuit).

Please state the resistor numbers as shown on the picture included.

Why does the BrianGT version choose such an extremely high gain? It is much higher than all my other amps!


Thanks.

Lucas

 

[Steve Eddy]

In the manual it is stated that one schould increase R3 to lower gain. Is that a better method? How much higher should I go?

That's the better way to go than trying to attenuate the signal upstream.

Don't go any higher than 2k. The chip needs to maintain a minimum gain of 10 for stability reasons.

se

 

[SpittinLLama]

You can either lower the Rf resistor (22K) or increase the R3 resistor (680) or both. The gain is Av = 1+ Rf/R3. Right now it is at 33.3V/V or a tad over 30dB. Quite a bit of gain. If you are using the Ci cap then it would be better to increase the R3 resistor. National typically shows the circuit with a gain of 21 V/V (20K/1K). Or you can put a 1K pot in series with the R3 resistor and vary the gain from the current 33 down to 12. Figure out what you like best and then change the value once and for all.

-SL

 

[Lucas_G]

Hi Steve,

Thanks for your answer.

BrianGT states that the gain can be calculated: Gain = 1 + Rf / R3 and that with R3=660 Ohm one gets 33 dB of gain.
I do not know what Rf means here...

If I would up R3 to 1 kOhm, how much gain would that be?

Regards,

Lucas

 

[Lucas_G]

You can either lower the Rf resistor (22K) or increase the R3 resistor (680) or both. The gain is Av = 1+ Rf/R3. Right now it is at 33.3V/V or a tad over 30dB. Quite a bit of gain. If you are using the Ci cap then it would be better to increase the R3 resistor. National typically shows the circuit with a gain of 21 V/V (20K/1K). Or you can put a 1K pot in series with the R3 resistor and vary the gain from the current 33 down to 12. Figure out what you like best and then change the value once and for all.

-SL

Hi Spittinlama,

That answers my last question. Thanks!

Lucas

 

[schmalex]

I never adjusted my gain - it turned out fine - but I did ask a very similar question:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=59448

 

[Lucas_G]

I never adjusted my gain - it turned out fine - but I did ask a very similar question:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=59448

Hi Schmalex,

Thanks for that thread. It too gives some answers.

I have just tried the whole thing. I added a 1 kOhm resistor in series with the 680 Ohm resistor R3. This gives a R3 of 1.68 kOhm, and therefore a gain of 14.

I can now easily adjust the volume, but apparantly it kills the sound! The amp lacks openness and air...

Has anyone tried different R3's?
I will also try 1kOhm (as prescribed by NS), but if it is still less good sounding, I will try adding the input sensitivity by adding resistors at the input. I understand that it is not very elegant to attenuate before amplifying, but if the whole concept of the gainclone suffers from a lower gain, then there is no other option, is there?

Anyone same experiences?

Lucas

 

[CarlosT]

Heck...I'm going for a gain fo 60

 

[Lucas_G]

Heck...I'm going for a gain fo 60

Hi Carlos,

A gain of 60?????????

Then you will probably use a 500 kOhm pot and speakers with 80 dB (non-) efficiency....?

Speaking of gain. It seems that Peter Daniel, BrianGt and all the others use the Gainclone with a gain of 33 and then they choose a relatively strong pot of 25 kOhm to turn things down. Did they do comparisons versus the use of a gain of 20 and a 10 kOhm pot? COMMENTS FOLKS!

Lucas

 

[ianpengelly]

I use the chipamp.com boards and driven by an either an Audiolab pre-amp or my Cambridge Audio amp (with a unity gain buffer pre-out) there seems to be a reasonable amount of control over the the volume. However, the pot in the Cambridge audio is something like a 150k device.

 

[Lucas_G]

I use the chipamp.com boards and driven by an either an Audiolab pre-amp or my Cambridge Audio amp (with a unity gain buffer pre-out) there seems to be a reasonable amount of control over the the volume. However, the pot in the Cambridge audio is something like a 150k device.

Hi Ianpengelly,

That 150 kOhm makes quite a difference.
Meanwhile I also tried a gain of 23 and that sounded already much better than a gain of 14, but still not as good as a gain of 33, so I decided to go back to the R3=680 Ohm value and now I have put a 50kOhm pot into my passive pre. Now everything is working properly and the good sound is back.

So I warn everone not to go for a low gain on these strange Chips without listening and comparing. Otherwise these little b's may react with a meagre and muffled sound!

Regards,

Lucas

 

[CarlosT]

I think that I'm effectively running like 58 in the IGC I'm testing right this moment (R2 = 140K, R1 = 2.41K)...hard to tell since I'm using summing resistors to combine L-R channels.

Sounds dandy hooked to my PC with Rhapsody...listened to everything from Thievery Corporation to Van Halen and everything in between

I'm still thinking of turning down the gain since the real test will e with an iPod driving the little beast (see pics in related thread).