From here:
http://dogbreath.de/GainCard/GainCard.html
it says that Gaincard's humpty has no capacitors. What are the possible benefits? Has anyone compared, side-by-side, LM3875/3886 amps powered by PSUs with and without lots of capacitance in the PSU section?
http://dogbreath.de/GainCard/GainCard.html
it says that Gaincard's humpty has no capacitors. What are the possible benefits? Has anyone compared, side-by-side, LM3875/3886 amps powered by PSUs with and without lots of capacitance in the PSU section?
Many theories are cursing around on that issue. Most poeple think that the Gaincard setup with only 1000µF per rail next to the chip gives the sound a certain "magic in mids and highs". The drawback seems to be a lack of punch. The latter is of course due to the lack of capacity.
About the cause for the "magic in mids and highs" you will find many opinions. One is that the lack of capacity makes the PSU faster, which in consequence makes the amp faster as well. Wouldn't any capacitor be faster than the LMs slew rate anyway? And aren't PSU capacitors extreme low pass filters, making the PSU as slow as possible to filter out any noise? Contradicting, isn't it?
Another opinion is that 1000µF capacitors have the best ESR/ESL-to-capacity ratio of all capacitors (Do they?) and that you cannot improve ESR and ESL by paralleling capacitors. That is possibly true, if you look at the equivalent circuit of a real capacitor.
A third opinion is that having the capacitors next to the chips avoids inductance and resistance due to long pathways so much that any further capacitor is superfluous. Wires with bigger diameters or wider PCB paths could however be designed to the same effect.
Google for more, if you are interested.
My opinion is that the lack of punch (=bass) results in more pronounced mids and highs, thus giving the impression of "magic". No proof of that however, so in the end you have to judge for yourself.
About the cause for the "magic in mids and highs" you will find many opinions. One is that the lack of capacity makes the PSU faster, which in consequence makes the amp faster as well. Wouldn't any capacitor be faster than the LMs slew rate anyway? And aren't PSU capacitors extreme low pass filters, making the PSU as slow as possible to filter out any noise? Contradicting, isn't it?
Another opinion is that 1000µF capacitors have the best ESR/ESL-to-capacity ratio of all capacitors (Do they?) and that you cannot improve ESR and ESL by paralleling capacitors. That is possibly true, if you look at the equivalent circuit of a real capacitor.
A third opinion is that having the capacitors next to the chips avoids inductance and resistance due to long pathways so much that any further capacitor is superfluous. Wires with bigger diameters or wider PCB paths could however be designed to the same effect.
Google for more, if you are interested.
My opinion is that the lack of punch (=bass) results in more pronounced mids and highs, thus giving the impression of "magic". No proof of that however, so in the end you have to judge for yourself.
two completely opposing designer viewpoints.
You will never find agreement.
Try both and decide which suits your ears, your speakers and your room.
If you find that the conventional high cap PSU sounds better in the bass and that the gaincard low cap no PSU suits the highs then passively bi-amp your speakers and build the better version for each driver then you get the best for your installation.
You will never find agreement.
Try both and decide which suits your ears, your speakers and your room.
If you find that the conventional high cap PSU sounds better in the bass and that the gaincard low cap no PSU suits the highs then passively bi-amp your speakers and build the better version for each driver then you get the best for your installation.
but most of those who have properly thought through the shortcomings of low capacitance also use very efficient speakers of the order of >=96dB/W/m.jackinnj said:With only 1,000uF............. going to seriously modulate the power supply rails if you try to run the LM3886 to its maximum --
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.