LM3875 - newbie gainclone project

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have a few questions about lm 3875 gainclone design shown below.HERE is the scheme
-
1.Why are that 100nF caps needed, do I have to include them in circuit and what can happen if not?

2.How to properly test the amp output(DC voltage)?
*I have first measured the DC voltage without load-20mV, than I have connected 5.6Ohm/5W resistor and the DC voltage was raising or jumping quickly (I think it was more than 1V) and became very hot, so in moment I turned it off

NOTE:
Im using +/- 34V DC power supply, 2x1000uF caps per channel BUT I did not include that 100nF caps in circuit

Thank you very much
 
The small caps are to bypass any noise from the ps around the amplifier. Depending on the quality of your ps filter caps they may not be needed and you might not hear any difference. In the very worst case the amplifier could be slightly unstable without them, but if the amp is being operated with a gain above 11 this is unlikely except to very poor circuit layout.
 
Hi,
the 100uF decoupling on board are probably mandatory.
The 1000uF smoothings caps would need to be very close to the board to replace the 100uF, possibly less than 50mm (flow and return) of wiring.
The heating when connecting the dummy load sounds like oscillation.
Houston, are you there?

BTW,
I have a big disagreement with the gainclone builders.
Inverted and omitting part of the DC blocking is asking for offset problems.
Passive attenuators add to these problems.
Omitting adequate smoothing cripples the continuous power ability and low bass response. But what do I know? It is only my opinion.
 
Zobel is mereley the description of the topology. The purpose of a zobel in a crossover is to flatten the impedance of a driver in order for the crossover to see a constant impedance, making the crossover more predictable. The purpose of the zobel with a gainclone is to dampen high frequency (above audible range) response. It is merely a capacitor and resistor in series with each other, and placed across the output terminals of the amp.

And yes, you can go with a non-inverted design. Most builders actually recommend it.
 
Grandma´s_SUB said:
Is it wrong if I have forgotten to put input capacitor? :rolleyes:
no, I was just wondering...
perhaps you would get more help, if you would share more information
showing a circuit you didn't build, is not very helpful...;)
anyway, there must be still something wrong. Even without dc-blocking cap, dc should be low and shouldn't change very much with shorted input
regards
 
Hi Grandma,
are you now sinking the input offset current to ground?
Was there no route for the offset current in your original schematic?

[Input offset current] times [Rs] = [input offset voltage].

[Input offset voltage] times [amp gain] = [Output offset voltage].

If one were to omit the Zin defining resistor the effective resistance seen by the input terminal is infinity and the output offset becomes infinity (well, limited to near supply rail voltage).

Whose schematic did you copy?
 
Its hard to say but I cant answer you on first two questions since I'm n00b but I can say you when I added both resistors and capacitor at input I got it working it.

If one were to omit the Zin defining resistor the effective resistance seen by the input terminal is infinity and the output offset becomes infinity

If I understand properly when I didnt have the 10k resistor I got infinitive gain (the +Vcc at output)

*The used schematic is in attachement
 

Attachments

  • nixgccdl.gif
    nixgccdl.gif
    15 KB · Views: 378
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.