how do i lower my dc offset 89mv

[joka22]

Hi i have build a gainclone as must of you know. I have measured the dc offset i get around 89mv. The setup is stereo gainclone with caps at rectifer bridge board which are 2 per rail one is elna 50v 3.3 and one is 50v 10uf panasonic. The ch is connected to the ground for both channels . pg- and pg + are connected to the sg on both amp boards to ground them. I do not use volume pot i connect the input wire straight to the output of my pc sound card.

I read on decibel degeon (http://myweb.tiscali.co.uk/nuukspot/...gainclone.html) that a carbon 56 k resister must be connected to the input wire to the ground chassis to lower dc offset.

After one hour my chip get quite hot i have on full volume most of the time. if it get too hot it starts crackle. anyone know why this happens??? stability problems??.

[lineup]

what circuit you use?


http://myweb.tiscali.co.uk/nuukspot/...gc/nixgccd.gif
If is like this one, inverted, in this image
you can experiment with

put one resistor from pin7
to Signal Ground Star
** put this resistor close to pin7 and other end to ground

Depending on the value of such a resistor
you may get another DC-offset at output.

Try with:
10kohm, 22kohm, 47kohm, 100kohm

If the offset not get any smaller with such a resistor,
then connect pin7 to signal ground again.

[joka22]

hi lineup i bought the peter daniel kit i cant rember if it was inverted or non inverted but it was a basic kit. i will try the resister method thanks for the input.

[danielwritesbac]

Originally posted by joka22
hi lineup i bought the peter daniel kit i cant rember if it was inverted or non inverted but it was a basic kit. i will try the resister method thanks for the input.

That's a non-inverted. Details are here: Commercial Gainclone kit- building instructions

[lineup]

1. 89 mV ( 0.089 Volt DC ) is not very big offset.
It is one you can live with. But we do not wish to have close to 0.1 Volt

2. See attachment. Typical non-inverted LM3875 circuit.
it is taken from The Datasheet
http://www.national.com/ds.cgi/LM/LM3875.pdf

3. The cap *Ci = 22 uF is to block DC feedback to GND.
When using Ci, we can often get close to 0.0 Volt offset.

4. When NOT using Ci, then we can get such DC-offset at Output.
To remove this offset is not so simple as only use 1 resistor.


There are several ways to fix the offset when not using capacitor Ci.
But simplest way is to ad a very good quality capacitor.
There are good electrolytic:
22, 47, 100 or even 220 uF may work.


regards

[Peter Daniel]

Originally posted by joka22
how do i lower my dc offset 89mv

The simplest way is to change the value of R2 from 22k to 10k.

[lineup]

this will lower the gain to about 50%
if you with R2 mean the feedback resistor

simplest way to refer, is the refer to the attached circuit in my previous post
or post your own attachment of your R2

[Peter Daniel]

By R2 I mean input shunt resistor: http://www.diyaudio.com/forums/attac...amp=1210693738

[danielwritesbac]

Kudos Peter! That works like a charm.

Originally posted by joka22
. . .After one hour my chip get quite hot i have on full volume most of the time. if it get too hot it starts crackle. anyone know why this happens??? stability problems??.

A stability problem is somewhat unlikely on such a nice circuit board. More information is needed:
How's your heatsink size in C/W?
Did you use thermal compound?
Can cool air get into the bottom of the enclosure and then somehow also let the now-hot air out of the amplifier enclosure?
What's the operating voltage?
Did you use flux for nice smooth soldering or are there issues?
Are your speakers 4 ohms or 8 ohms (what's the load)?
How long are your speaker cables?
Did you use the Audiosector recommended capacitor models at the amplifier board, or did you use something else?
And, can you post a photo of the amplifier view as from inside the enclosure?

[Dxvideo]

I had a LM4780 amp with Ci (47uF) and my offset was ~0,1mV on each channel but sound was boring!
Then I decided to bypass Ci caps, the sound has improved incredible! Especially the bass stage has upgraded to another league (this is an English idiom). May be it was a bit psychological. Anyway, after that operation the DC offset was increased to 90-100mV
However, in my gainclone the input shunt resistor (Rs) was 47K while the Rf was 20K. So.... As a friends recommendation, I've replaced Rs with a 20K (= Rf). Then my DC offset was reduced to 40-50mV on each channel! I think these values are more acceptable and I've never heard any kind of click or plop sound in on-off operations.
IMO: Dont use Ci capacitors, reduce the input shunt resistor values.. 50mV is not too bad for DC offset! Trust me it worths.