First GC Amp .... something wrong?

[sl_1800]

I have built my first GC amp using Brian's boards. I measured for DC offset with not load resistor and measured a negative 110mv. How can I be getting a negative reading? And yes I'm sure I had the meater hooked up correctly and the outputs on the board are correct. Is there really a problem or should I check with a load resistor? All other voltages are correct.

I was pleased when I pluged it in and got no smoke and no flames but I really am confused on the negative reading on the DC offset.

 

[jaudio]

what are the measurement for supply rails. check solder joints

 

[kafka]

I really don't understand Your confusion... If You can have positive DC offset You can have negative as well, cos' the power supply is symetrical.

 

[jaudio]

kafka right but 110mv is kinda high

 

[peranders]

If you calculate what you _can_ get this is not extremely high.

Let's say that you have between offset voltage (according to datasheet) 1-10 mV and assume that the input bias currents are small in relation with chosen resistor values.

The gain is 30 => offset 30-300 mV!

 

[sl_1800]

Ok, it this negative DC offset is not out of the question then how do I go about reducing it from the 110mV to something a little closer to 0 ???

I have heard about a Ci cap which would be between Ri and ground, correct? If so then I'm assuming I would have to pull the ground end of Ri and insert a non polar cap inbetween ??

Also will a different value resistor for Ri have a effect ?

Thanks for the help.

 

[boholm]

It can have something to do with your (?) choice of values of the resistors. Can you please post a schematic with resistor values in it?

 

[sl_1800]

The values used are the ones that come with the BrianGT standard LM 3875 kit.

 

[boholm]

If I am looking at the right schematic - and I am not very familiar with BrianGTs products - there is a 220 ohm resistor between input and pin 7 of the LM3875.

This is the resistor to experiment with. In theory it should be 601 ohms ( 620 // 20k ). You may try with a trimmer (1 kohm) if you got one.

 

[sl_1800]

I can give that a try, but won't that change the gain of the amp or not?

 

[boholm]

I don't think you will notice it at all.

 

[transducer]

Did you have the inputs shorted when you measured the offset? When I first tested my boards, I didn't have them shorted, and the offset was about 50mV. With them shorted, the offset dropped to 0.8mV on one channel and 9mV on the other. Would be nice to get them both around 0V, since these will be driving tweeters directly in an active setup.

By the way, the amp worked perfectly the first time out, no surprises there 😉

 

[sl_1800]

The inputs were connected to nothing and the outputs were connected to nothing. I will try both and see what I get and report back shortly.

 

[boholm]

Did you have the inputs shorted when you measured the offset?

Damn, I forgot that - Yes, a very important step. And then maybe the 220 ohms produces a sufficient result? If not *then* you can experiment with the value.

 

[sl_1800]

Well the addtion of a output load made no difference in the DC offset but the addition of a input load, my Sony dvd player, the DC offset dropped to a negative 47mV. This is much better but I would still like to get it lower as this is going to be part of a active system.

 

[transducer]

Don't measure it with a source connected, you don't know if you're measuring the amp or the source DC offset. Short the input by using a male RCA jack, connecting the centre hot pin to the outside ground connector. I use the cheapest RCA connectors from Radio Shack for this.

You don't need a load connected, it won't make a difference.

 

[sl_1800]

Ok thanks, I'll give that a try in the morning.

 

[sl_1800]

With the inputs shorted I still get a negative 47mV on one channedl and a negative 49mV on the other. I'm going to try the Ci cap today and see what that brings.

 

[sl_1800]

I guess I won't be trying the cap today as RS doesn't have any that size so I will have to order some.

 

[boholm]

Remember: You still have the option of experimenting with the 220 ohm by changing it to a higher value (still measuring with the input and ground shorted).