Need help with my LM1876 amp-project

[Nenola]

Hi all!

I'm about to build an amp using two LM1876. First I was planning to bridge the other amp for sub use to get a 2.1 system, but later I decided to go with two stereo amps; one for speakers and one for subs, total of 4 channels (2.2). Now I'm having a hard time finding a good schematic, because I'm not so much in calculating values for components and the datasheet doesn't give enough information for ME

Today I found a amplifier kit with LM1875 from some web site and it had some kind of schematic included. I've understood that the 1876 is a dual version of 1875 so probably I could use the same shematic? I'll attach it with this post so if someone out there would be so kind to look at it and tell their opinion.

Thanks all!

[AndrewT]

some web site

Much be as much of a secret as your location.

[paulb]

What's wrong with the schematic in the LM1876 datasheet? It gives component values.
I bet the LM1875 kit uses its standard datasheet schematic.

[Nenola]

Hello!

Thanks for the hint. There was nothing wrong with the schematic, I just didn't get it at first.

Now I've come up with some kind of schematic, I'll add it here so you can see it.

[AndrewT]

Hi,
add an RF filter to the input. Ranging from 330pF to 1nF in parallel to R3 should do.

The NFB resistances are quite high.
I recommend reducing them to 1k5 and 27k, or better still, 1k2 and 22k but with C3 increased to 27uF//1uF

Is R6 correct? It seems very low!

[gfiandy]

Whilst this amplifer may be stable without a Zobel network I would recomend adding one especially if this is your first design. It is likely to help keep it from oscillating. I just had a look at the LM1876 data sheet and it appears they recomend 4.7R and 100N.

Depending on the cables you connect to the amplifer you may also need the output inductor 0.7uH bypassed with a 10R resistor. (As shown on page 6 of the data sheet). However if your cables are not capacitive (ie you are not using coaxial construction) then you can probably get away without this and amps normally sound better if you can leave this out.




Regards,
Andrew

[Nenola]

Hi!

Thanks for your quick answers, they were really helpful.

Some of the component values are different from the schematic, because I got them from the overture design guide -excel sheet.

I added the RF-filter to the input and used the values from the guide. If you think it's not good, I can change it to the values you suggested.

Also changed the component names to R1,R2 etc. so it's easier to make changes, and I added the Zobel network. I'm posting only one channel schematic because they're identical...

[AndrewT]

Hi,
your new values don't work.
The NFB DC block is 22uF & 2k7. This is ~60mS RC time constant.
The input DC blocking filter must be <=60/1.4 ~40mS.
You have 22uF & 47k giving >1000mS. You must set the input filter narrower than the NFB filter.

The gain is now 120/2.7+1=45times, whereas before it was 19times.
which do you need?

The RF filter is missing. use between 0.5uS and 1.5uS RC time constant.
For 2k2 this requires a parallel cap to ground of between 220nF and 680nF.

Don't mix up the grounds which have the same symbol.
You have a signal ground, a NFB loop ground, a speaker return, a Zobel return, a decoupling ground. These all handle completely different currents and pulses.

[Nenola]

Thanks Andrew!

I've made the changes and now it should be ok? I've understood that a gain of about 40 is recommended and would be ok. The previous 19 was a mistake.

I have to get back to that grounding issue when I first get the values correct.

and thanks again!

[danielwritesbac]

Speaker zobel:
You may wish to illustrate the speaker zobel in a different location--right at the speaker terminals is most understandable. It may share ground with the speaker.