Adding larger capacitors to Audiosector LM3875 kit

[ratza]

Because people don't know better and copy from somebody whom they believe to be competent.

Answer of the day. No, wait! That's the answer of the week!

 

[jordyyy]

jordyyy, why don't you try and hear the results yourself? Adding more capacitance won't matter after a certain value.

I'm still in the research stage of my audio system
Maybe ill give it a try after i order/mod my gigaworks dac,
order/build my lighter note kit passive pre
and order/build my lm3875!

but i think im about ready, so it wont be long.

 

[danielwritesbac]

. . .
What is good bass for you? That extremely clean and well-defined but not punchy bass of OB speakers? Or that powerful bass you get from PA and horn speakers that kicks you in the chest and robs you of your breath or affects your heart rate? Or rather that boomy sound you sometimes get from bass-reflex speakers or boom cars?. . .

Oh, yes please!

All of those are necessary, not droning all the time, but rather in response to the music source, on an as-needed basis.

Ah, and for bass, the input load for the amplifier matters a lot less than the output load for the source. Fortunately, in the case of a computer, its approximately 11.5k DC load at the amplifier. The audiosector specified model of input cap goes in-between pot and rca jacks.

Insufficient load at the input can make the source like a floorboarded car with the clutch pushed in.
The part of the signal most reliant on current is the larger size signal--the bass. So, if the source isn't pushing as much current as intended, then chances are that it, likewise, won't push as much bass as intended. That's especially true of computer sound cards in general, although it sure doesn't apply to every model of sound card.

P.S.
An excellent PA prosound effect from LM3875? Oh, I haven't the strength today. Try motional feedback. That'll put the smackdown on room noises, spike system noises, and the horrible acutance (fake clarity via ringing like a photoshop sharpening filter) screeches of the non-inverted LM3875. Those noises must be controlled before it can be enjoyable to run the amplifier at full output. Otherwise, any decent 1 watt amplifier would be just as good as a power amp that you wouldn't crank up anyway.
And ixnay that 2-way speaker design because its no good using full power for flapping the same driver that outputs the voice band, shakin' up the singers. They really hate that, just ask them. Its unanimous amongst singers that your music is never supposed to sound like a 2-way speaker cranked up loud. And, while we're shedding off clear sounding screech fakery, might as well throw out those dome tweeters, especially 1" size and get some electrostatic (or at least a decent 3/4" dome restricted to the range that it can do best).
Ah, sorry, prosound on LM3875 would probably need at least nested design if not go full out with motional feedback, before it would be fun loud, and there would also need to be speakers designed to run pleasantly at full output. On traditional design, those would be the size of soverigns by the time you stacked series-parallel big woofers up to get 45hz@101db with 1w@8ohms, as a big head start for the little chip. Maybe we could make them skinny with folded half wave design to greatly decrease expense. Going without motional feedback, we could use nested amp design to knock out that screech, a quality regulated supply on the pilot amp for really good bass, and 4-way speaker design with l-pads to serve for low-distortion room compensation. Okay, its do-able. But, those folded half waves are going to punch a big hole in the ceiling.

 

[danielwritesbac]

You know, a pilot amp could be a great new audiosector product.
Its a preamp that attaches to both the audio input and the speaker output of the power amp so that the power amp is inside of the preamp's feedback loop--nested amplifier design drops the gain on noise but it will also help produce sufficient bass thanks to the pilot amplifier's independent regulated voltage.

 

[doggy]

I have the LM3875 also. Seems I am the only one here, who does not worship Panasonic fc caps supplied. I have a bunch of them, and sonically to me are on the bottom. I replaced mine with Elnas and also paralled one extra 330 uf-200 v on each board. This cap adds more bass. I also use a CRC pre power supply using 4-1500uf caps,2 220uf 200v after the rectifiers.

The LM3875 is my main system amp right now, and it has as the best bass of all my amps.

 

[danielwritesbac]

I have the LM3875 also. Seems I am the only one here, who does not worship Panasonic fc caps supplied. I have a bunch of them, and sonically to me are on the bottom. I replaced mine with Elnas and also paralleled one extra 330 uf-200 v on each board. This cap adds more bass. I also use a CRC pre power supply using 4-1500uf caps,2 220uf 200v after the rectifiers.

The LM3875 is my main system amp right now, and it has as the best bass of all my amps.

Wow! I'd sure like a sketch of that power circuit. Would you mind posting that?

 

[doggy]

I use 2- 5W ceramic r on the center 0V rail. The 1st off the rectifier to the 220uf 200volt caps,from there a 2nd ceramic to the 4 main ps filters.

The boards have the 1500uf panasonics removed and replaced 220-330 uf caps-63v, the outboard caps are added to one cap only on the board; I actually had too much bass and fine tuned this one-it works for me.

I would like to try inductors one day instead of R's

Power supply is 1 important link in the audio results chain-I like "rolling" caps;0

 

[danielwritesbac]

I use 2- 5W ceramic r on the center 0V rail. The 1st off the rectifier to the 220uf 200volt caps,from there a 2nd ceramic to the 4 main ps filters.

The boards have the 1500uf panasonics removed and replaced 220-330 uf caps-63v, the outboard caps are added to one cap only on the board; I actually had too much bass and fine tuned this one-it works for me.

I would like to try inductors one day instead of R's

Power supply is 1 important link in the audio results chain-I like "rolling" caps;0

2 ohms~2 ohms padding on the 0v rail? That's a lot like putting a 4 ohm resistor in series with the speaker's negative terminal. . . except that equal and opposite voltage at the the 330uF caps (stereo amp?) which is 330uF//330uF, total 660uF (because of 2 amp boards) is creating a path of least resistance (maybe?) at the amplifier board ground planes (marked "CG" on audiosector boards) and, sure that 660uF ought to be able to reference an 8 ohm speaker down to 60hz solid with roll off possibly occurring at 30hz. Those amplifier boards have a pair of 330uF, each and there's two amplifier boards, and there's no padding in between the speaker and the amplifier boards, so that's pretty interesting! They're making your speaker ground.

That's fine if you like the sound; however, there's a safety caveat in that the neutral at the amplifier boards isn't as "attractive to current" as the firm ground of a computer sound card. You might consider a groundlift resistor to create a signal star ground on future designs.

Well, that's an interesting approach. You've padded the only power supply cable that the amplifier doesn't use. Since the speaker does use that line, you may have created some form of hybrid current drive. Those sorts of arrangements are normally used to take the shout out of full range speakers and sometimes used to increase the bass of sealed box speakers.

See also variable current drive ("dial-a-tube"): Lenard Audio - Education - Guitar Valve Amps.

 

[AudioEngineer]

I just wanted to give an update because I just added some 10,000 uF capacitors to the power supply board and added the snubber components. I sat down and listened to some cd's and I honestly can't tell the difference or if it made any improvement at all. I thought a snubberized power supply with larger caps was supposed to give better bass and a more dynamic sound but it still sounds the same as before I think. Its hard to say since I couldn't do a direct comparison before and after the changes. I think there's a a bit more bass but maybe I'm just hearing things because I was expecting an improvement. Oh well it didn't cost too much for the parts so I guess it was worth trying it to see if it made any difference.

You should use about 20,000 uf per rail, 4 x 10,000 uf caps would do nicely followed by a 100nF cap on each rail. There won't be a huge difference but the bass will be very solid and the amplifier will not distort if pushed to near its limit where the protection circuitry kicks in. I have had very good results using this method.

 

[gootee]

Yes, at least 20000 uF per rail, and maybe twice that, as a guess.

More reservoir capacitance will enable output of strong bass to lower frequency, without clipping, up to maximum rated output power, but after some point more capacitance will make very little difference. That point depends mainly on the rated peak output power, the load impedance, the capacitors' ESR, and the power supply.

You should make sure that the capacitance is large enough that pulling the maximum rated peak (not RMS) load current out of it, continuously, for a full rectifier charging period, will not pull the rail voltage too low. What "too low" is would have to be calculated. But we know that the rail voltage dip divided by the peak current can never be lower than the net capacitor ESR (equivalent series resistance), or even very close to it since that would require excessive capacitance. (So use low-ESR caps.)

But the rail voltage should never dip more than the nominal rail voltage minus the peak signal voltage minus the amplifier's clipping voltage, Vclip, which is in the datasheet and is usually at least 3 volts or more.

After the maximum Vdip_max is calculated (and remember to account for the effect of low AC Mains voltage), just use I = C dv/dt, re-arranged as C_minimum = Ipeak_max x charging interval in seconds / Vdip_max, where the charging interval would be 1/120 for 60 Hz mains or 1/100 for 50 Hz mains, if using a full wave bridge rectifier. Multiply the result by one million to get uF. (ESR is not included but if you search for my other posts, you will find the more-accurate equation, with ESR included, and also a downloadable spreadsheet that does it all for you, and calculates how close the caps get you to the theoretical max rated power without any chance of clipping.)

It can help many aspects of the sound quality if you make sure that you have enough capacitance very close to the chip amp's power pins. You must have at least 470 uF as close as possible to each chipamp power pin, but more can only improve its performance. I would install up to 10000 uF if it would fit, there, but 1000uF to 2200 uF would probably be enough. (The farther away from the power supply caps the chip amp is, the greater the effects of the rail conductors' inductance and resistance, and the more I would worry about adding more capacitance right at the chip amp. But the really-low bass frequencies probably aren't affected by the distance, unless it's many feet.)

If you can also fit a smaller cap even closer to each power pin, you should do so. Even for the 470 uF, alone, it would be better to use five 100 uF in parallel, or two 330 uF or 220 uF in parallel. Putting them all on different sides of the power pin would be best, in terms of lowering the inductance by paralleling (which should give the best transient performance and therefore probably also the best imaging.)

A 0.1 uF from each rail to ground must be used, but ONLY right at the chip amp's power and ground pins. Anywhere else on the rails can't have any good effect at the chip amp power pins (which is the ONLY place that matters at all), due to the rails' inductances, and is just asking for trouble, such as oscillation or ringing, which might cause distortion that sounds like brighter, better highs.

Also, a snubber can only be effective if the resistance value is set to the characteristic impedance of the LC of the circuit. The only good place for a snubber for which the need has not been verified with measurements (e.g. diode turn-off ringing seen on oscilloscope) is probably across the transformer secondary. But again, using a generic resistor value will probably do almost no good. And a small capacitance alone should never be used on rectifier diodes or transformer secondary, since it will not damp the resonance but will instead lower its frequency, which might be worse than doing nothing. (A snubber is a resistor and the familiar series capacitor is only added in order to keep the resistor from seeing anything except the unwanted frequencies, so its power dissipation isn't unnecessarily large.)

 

[lanchile]

I know this is an old thread but, this is my conclusion about the the Audiosector LM3875 kits:
These kits come with "small" 10uf caps for the power supply, When I build these kits (long time ago) I try many caps in the power supply, here is the list that I remember putting there:
2.7uf
10uf
470uf
10000uf
20000uf
I did this using "alligator clips" to change the caps quickly and easily.
Using the 2.7uf, 10uf, 470uf did not change the sound too much at all. but, as soon as I put the 10000uf caps...the bass came alive!!!. I try the 20000uf caps but, I did not noticed an improvement over the 10000uf caps, only it took longer to the music to disappear after turning off the amp (discharging time from caps). I would strongly suggest to all those members using those kits, just try using at least 10000uf caps and you will "feel" the bass. Nothing to loose here but, you will gain a lot!. Now I do not have that amp anymore, I moved to solid stated class A/B amps and I can tell you one thing...I am not going back to Chip amps anymore

 

[danielwritesbac]

I know this is an old thread but, this is my conclusion about the the Audiosector LM3875 kits:
. . .as soon as I put the 10000uf caps...the bass came alive!!!. . .
Now I do not have that amp anymore, I moved to solid stated class A/B amps and I can tell you one thing...I am not going back to Chip amps anymore

Interesting!
Could you do us the favor of naming what solid state discrete parts class AB amplifier you're using now and what is the most important ways that it is more practical/useful than the chip?

I ask because, if given the schematics for both, plus a description of how one of them is more practical/useful, the electronic differences can be analyzed to find the cause of the difference in practicality/usefulness. That can be really fun, sometimes.

And, if there are several reasons to choose the discrete instead of the chip, could you list them in a prioritized list, like #1, 2, 3, 4, etc. . .?

 

[lanchile]

Interesting!
Could you do us the favor of naming what solid state discrete parts class AB amplifier you're using now and what is the most important ways that it is more practical/useful than the chip?

I ask because, if given the schematics for both, plus a description of how one of them is more practical/useful, the electronic differences can be analyzed to find the cause of the difference in practicality/usefulness. That can be really fun, sometimes.

And, if there are several reasons to choose the discrete instead of the chip, could you list them in a prioritized list, like #1, 2, 3, 4, etc. . .?

Well, after my first LM3875 chip amp, I went to discrete to try their sound. I build the NXV200 from Aussie amplifiers (Anthony) it sounded way better than my "boutique" LM3875 amp using Black gate caps etc. after that I started building many other discrete amps from many vendors. Right now I am using 2sk1058 2sk2221 irfp9240 mosfet amplifier schematic. Power supply is unregulated, Mundorf Audio grade 6800uf x4 with a 42vdc per rail. if I have to compare the sound of a chip amp to these discrete amps, I would say that chip amps do not have the "authority" to drive speakers and produce the right sound at the same time. I am trying to explain that the sound of these IC's are good, but will not compare to a well designed discrete amp. Build the best chip amp that you can and then build the best discrete amp that you can. compare sound and choose...I bet you will choose the discrete!. For a "main" amp in my living room I would not use a chip amp. But for a kitchen, garage, for the kids etc I would use a chip amp not doubt!!!