DIY amplifier kits - Any experience?

[Slanski62]

I'm intrigued by the various amplifier kits on the 'bay. I think it would be fun to give one a try. Has anyone every built one? I have some very basic newbie questions:

Which modules do I need to build? There are various power supplies, amplifier designs (look to be copies of commercial designs) pre-amps, etc.

How do you connect all the parts together?

How do I control the volume and inputs?

Most important, would this sound any good?

Any suggestions are appreciated.

[marscoast]

You're thinking along the lines of a solid-state amp, correct? If you're a newb, steer away from those expensive tube kits for now.

Google "gainclone" and you'll find a lot of helpful information.

The Decibel Dungeon website is a great source of info Linky (http://myweb.tiscali.co.uk/nuukspot/decdun/index.html)

Here's an excellent chapter that focuses on the power supply:
http://myweb.tiscali.co.uk/nuukspot/decdun/gainclone_psu.html#psuq3

PE sells the Velleman 30W Stereo Audio Amplifier Kit that is a simple, fun place to start.

[philiparcario]

It depends on price levels and kit designs.

I have had very good luck with gainclones

and ska-audio greg ball makes them. the ska-audio amps are more costly link to him

http://www.ska-audio.com/

I have some gainclones got to find a link for them here is a transformer link

http://www.antekinc.com/

here is a case link

http://www.par-metal.com/

same building and same owner for the the transformer company and the cases company.

see thumbnails of a ska-audio mono block in a par-metal case ,

the black case is a 5 channel gainclone with a 5 channel volume control. you can make a very good 2 channel gainclone under 150 . that counts a nice case.

[dickwest]

Phillip,

Any comments about the comparison of the sound of a typical gain clone vs: a Hafler amp, such as the DH-120?

Thanks......

Dick

[jclin4]

I have not built any gainclone kits, but bought one off of Phil, then rebuilt the case. This one is based on the LM1875 chip, about 20 WPC I think. The sound in the high and mid-range is sweet and clear.

The bass could be more robust, but that might due to the low power rating (or even my setup). Higher power gainclones, like those based on LM3886 might be more balanced.

http://techtalk.parts-express.com/attachment.php?attachmentid=5339&d=1256505231

[mikejennens]

I built this one from Parts Express:
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=320-212
(pics attached)
I went to a smoke shop and paid $3 for the case, bought the transformers from Parts Express, and the tone/volume control from Tequipment.com.
The support by the Parts Express techs was awesome!

I also built one from audosector.com which I like and the support by Peter Daniel was awesome too!

IMO, either one is a good choice.
Mike

[philiparcario]

I have not built any gainclone kits, but bought one off of Phil, then rebuilt the case. This one is based on the LM1875 chip, about 20 WPC I think. The sound in the high and mid-range is sweet and clear.

The bass could be more robust, but that might due to the low power rating (or even my setup). Higher power gainclones, like those based on LM3886 might be more balanced.

http://techtalk.parts-express.com/attachment.php?attachmentid=5339&d=1256505231

nice job with the case. that chip is the low power one. the 3886 has better bass. remember the 3886 can do 68watts tops. If you want a lot of bass 2 or more LM3886's can be used per channel.

[philiparcario]

Phillip,

Any comments about the comparison of the sound of a typical gain clone vs: a Hafler amp, such as the DH-120?

Thanks......

Dick

dick I have been very lazy I have the Hafler sitting and waiting for a long tryout. I need to run it. I have a nice pair of scanspeak 2 ways I should try out my best gainclone against the hafler.

[charlielaub]

Not sure what level of power you need or desire...

If some LM3886's will work for you, then I can recommend these "kits" (power supply board plus two LM3886 boards) for a 60 WPC at 4 ohms stereo amp:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=310177847303

$44 for all that, and then you need to buy a 22V-0-22V transformer (I recommend buying an Antek AN-2222 from Antek (http://www.antekinc.com/) for $30), a case, fuses and an on-off switch for the most basic of amps. You only need to connect the wires between the parts and you are off and running. That is about as simple as it can get!

-Charlie

Ooops: I forgot to mention that you will need a heatsink. DO NOT run the LM3886 ICs without attaching them to a heatsink or they will auto-shut-off after about 30 seconds!

[philiparcario]

yeah antek an-2220 or antek-2222 will both work

[WmAx]

Be sure to consider your reasoning behind this chore. If it's for fun and learning - then yes - it's worth doing. To save money? No way. You can get high power, superbly performing amps for super cheap these days. In higher power ones, you probably can't even build for the price of the completed commercial unit, much less build it cheaper.

-Chris

[tpremo55]

Be sure to consider your reasoning behind this chore. If it's for fun and learning - then yes - it's worth doing. To save money? No way. You can get high power, superbly performing amps for super cheap these days. In higher power ones, you probably can't even build for the price of the completed commercial unit, much less build it cheaper.

-Chris

For instance?
You make a good point, I'm just curious as to what you list into cheap, high power, superbly performing amps these days...

[WmAx]

Behringer Ep2500. Very good parts used through out, high grade glass epoxy PCBs - with redundant traces/PCB (mil spec). Very high power output (1000W+ x 2, both channels driven, 2 Ohms, 450W, x2, both channels driven, 4 Ohms and very low distortion. Can find for as low as $270 new.

-Chris

[Æ]

How about an inexpensive tube amp kit. Complete except for enclosure/housing. http://www.s5electronics.com/thome.html


You're thinking along the lines of a solid-state amp, correct? If you're a newb, steer away from those expensive tube kits for now.

Google "gainclone" and you'll find a lot of helpful information.

The Decibel Dungeon website is a great source of info Linky (http://myweb.tiscali.co.uk/nuukspot/decdun/index.html)

Here's an excellent chapter that focuses on the power supply:
http://myweb.tiscali.co.uk/nuukspot/decdun/gainclone_psu.html#psuq3

PE sells the Velleman 30W Stereo Audio Amplifier Kit that is a simple, fun place to start.

[JRT]

Behringer Ep2500. Very good parts used through out, high grade glass epoxy PCBs - with redundant traces/PCB (mil spec). Very high power output (1000W+ x 2, both channels driven, 2 Ohms, 450W, x2, both channels driven, 4 Ohms and very low distortion. Can find for as low as $270 new.

-Chris

Are you using that full range, and are you suggesting it for that application?





.

[philiparcario]

Behringer Ep2500. Very good parts used through out, high grade glass epoxy PCBs - with redundant traces/PCB (mil spec). Very high power output (1000W+ x 2, both channels driven, 2 Ohms, 450W, x2, both channels driven, 4 Ohms and very low distortion. Can find for as low as $270 new.

-Chris

does it have no hiss? is it dead quiet? Just curious. Gainclones have limits 1 chip is only about 60 watts(lm3886).
But you can make it in a very small case and the amp is silent no hum or hiss. It is not going to do a 15 inch sub. maybe the best of both worlds is the behringer for the bass and the gainclone for the mids/treble.

[tom_s]

AE, Did you build that amp? That's very nice.

[evilskillit]

Yeah things like the gianclone have their place. If you want to go active 3 way something like the gainclone can be good to drive the mids and highs. They have very low harmonic distortion, are small, don't have to have a noisy fan to drive them and are probably more than capable of driving tweeters and a lot of mids way louder than most people would ever want. Sure the Behringer amps are great for driving subs or maybe big passively crossed over speakers but they're not the best solution for every problem maybe even most problems.

Gainclones would probably also be great for powering desktop pc speakers if you needed more grunt than a t-amp can offer, or your background music speakers for your bedroom if you want something hifi and small. You can see where there could be plenty of applications where these amps might be good. I would like to take a shot at building one some day just to see what its like.

[Æ]

AE, Did you build that amp? That's very nice.

Yes. Thank you.
It comes in a kit, with everything you need except the housing/case.
I did the wood frame and aluminum top fabrication on my table saw and drill press.
The wood is Bubinga and Walnut from Woodcrafters.
I took my little amp to Deete's Sound Room in Carmichael for a listen. It's a "high end" audio salon here in suburban Sacramento. He sells a lot of kilobuck tube equipment. We hooked it to a large very expensive pair of Zingali Loudspeakers. My amp performed admirably. About 89 percent of the performance for 6 percent of the cost.

[djg]

AE, did you mod that amp or build it stock? Very nice work.

[Æ]

AE, did you mod that amp or build it stock? Very nice work.

Basically it's stock. I did upgrade some of the components. I used gold plated binding posts instead of the cheapies they provide. Same with the RCA input jacks, I went with heavy duty gold plated. And I found myself a slightly better volume potentiometer. The volume pot they provided was absolutely lousy. My only serious complaint would be about the lousy volume pot.
S5 originally provided you with plastic tube sockets, instead of ceramic sockets. I upgraded to ceramic at additional expense. Now though, I believe ceramic tube sockets are standard issue.

[WmAx]

does it have no hiss? is it dead quiet?

The Ep2500 is dead silent; it's unweighted SNR was measured at -110dB by Chasw98 over at AVS.

The only issue is the fan - which is easily fixed by exchanging it for a specific model Matsushita (and adding a pot or a couple of resistance networks with a selectable mode switch as well to adjust speed on top of that is my suggestion).

Also, if you have only unbalanced RCA to feed it directly from your pre-amp, I recommend a device such as the Samson S-Convert to quietly increase the signal voltage to an ideal level and connect to the balanced XLR or 1/4" TRS inputs, as it's generally a bad idea to crank the amp input dials all the way up IME(usually results in some small amount of hiss/noise) to compensate for low voltage consumer RCA output feeds.

Yamaha also makes some very high quality pro amps (P2500S, P3500S, P5000S, P7000S) that are a real bargain for the power vs. price. And these have no fan noise to deal with, so no exchange of the internal fans is needed. Also, the rack mount ears come off the Yamahas so that they fit in a regular home audio rack/cabinet if so desired.

-Chris

[djg]

Thank you, AE.

[DDF]

Not sure what level of power you need or desire...

If some LM3886's will work for you, then I can recommend these "kits" (power supply board plus two LM3886 boards) for a 60 WPC at 4 ohms stereo amp:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=310177847303

$44 for all that, and then you need to buy a 22V-0-22V transformer (I recommend buying an Antek AN-2222 from Antek (http://www.antekinc.com/) for $30), a case, fuses and an on-off switch for the most basic of amps. You only need to connect the wires between the parts and you are off and running. That is about as simple as it can get!

-Charlie

Ooops: I forgot to mention that you will need a heatsink. DO NOT run the LM3886 ICs without attaching them to a heatsink or they will auto-shut-off after about 30 seconds!

Charlie, thanks for the lead, I just ordered this kit. Had a Plitron trannie, nice case and abig hairy heatsink from when I blew up some old power modules. I'll post once I get it up and running.

BTW, been carrying some of these LM parts around with me for well over 10 years but was always too lazy to wire something up. This makes it brain dead simple. Now I have to design some new small speakers for my boy's room. Thanks for the excuse for a project. :)

Dave

[charlielaub]

Hey, glad to help with your decision! One thing I have discovered with these particular chip amps is that there is a resistor in series with the output of the IC. I think that this is their way of increasing stability to high frequency oscillation, although the need for that is really dubious. This just puts a series resistance in the mix, which reduces electrical damping and increases driver Q. The LM3886 datasheet shows a better method - an inductor and resistor in parallel, made by winding 10 turns of wire around a resistor. I plan to just delete this component (remove it from the pre-assembled PC board) in my applications. It's the rectangular white thing on the board. Other than that, these kits are super, AND super easy to use because you do not need to solder everything together.

-Charlie

[DDF]

Hey, glad to help with your decision! One thing I have discovered with these particular chip amps is that there is a resistor in series with the output of the IC. I think that this is their way of increasing stability to high frequency oscillation, although the need for that is really dubious. This just puts a series resistance in the mix, which reduces electrical damping and increases driver Q. The LM3886 datasheet shows a better method - an inductor and resistor in parallel, made by winding 10 turns of wire around a resistor. I plan to just delete this component (remove it from the pre-assembled PC board) in my applications. It's the rectangular white thing on the board. Other than that, these kits are super, AND super easy to use because you do not need to solder everything together.

-Charlie

Thanks for the heads up, ditching the resistor is a must. The 3886 doesn't like any load much below 6 ohms and it it has that current limiting circuit in it. I wonder if this is a way for them to avoid too many returns when people mismatch with low impedance speakers?

My old Bryston hooked up to the computer speakers uses an output inductor for stability, pretty common in the old days. The National datasheet calls up 0.7uH//10ohms.

BTW, very helpful when choosing power supply and heatsinks:
www.national.com/appinfo/audio/files/Overture_Design_Guide15.xls Overture_Design_Guide15

[Æ]

If you can get a hold of the very 2nd issue of Speaker Builder magazine, there was an article by Nelson Pass and someone else about output damper circuits, resistor and inductor.
For amplifier stability, but not always needed. I modify them or remove them.


Hey, glad to help with your decision! One thing I have discovered with these particular chip amps is that there is a resistor in series with the output of the IC. I think that this is their way of increasing stability to high frequency oscillation, although the need for that is really dubious. This just puts a series resistance in the mix, which reduces electrical damping and increases driver Q. The LM3886 datasheet shows a better method - an inductor and resistor in parallel, made by winding 10 turns of wire around a resistor. I plan to just delete this component (remove it from the pre-assembled PC board) in my applications. It's the rectangular white thing on the board. Other than that, these kits are super, AND super easy to use because you do not need to solder everything together.

-Charlie

[joekraska]

Behringer Ep2500. Very good parts used through out, high grade glass epoxy PCBs - with redundant traces/PCB (mil spec). Very high power output (1000W+ x 2, both channels driven, 2 Ohms, 450W, x2, both channels driven, 4 Ohms and very low distortion. Can find for as low as $270 new.

-Chris

This is interesting. I didn't know they were going this low in price. Do you own one? What's the fuse rating on the back?

Many manufacturers of amplifiers rate their amps at the peak discharge of their AMP, not according to constant output. One can tell the true output ability of the AMP by multiplying the fuse by the wall power (e.g., 115v x 20A = 2300W ).

Since your average house only has 15A circuits, and the electrical code requires that such devices be derated to 12A, it's probable that this device is no more than about 1400W or so. So what's the real truth? What's the fuse say? It cannot lie.

(Actually, reading the manual, I see it has a breaker of unknown rating; still it seems clear to me this manufacturer is inflating its figures, which is already off to a bad start with me, albeit I am willing to accept it may be a good amplifier regardless of the misrepresentation).

Joe.

[lhwidget]

This is interesting. I didn't know they were going this low in price. Do you own one? What's the fuse rating on the back?

Many manufacturers of amplifiers rate their amps at the peak discharge of their AMP, not according to constant output. One can tell the true output ability of the AMP by multiplying the fuse by the wall power (e.g., 115v x 20A = 2300W ).

Since your average house only has 15A circuits, and the electrical code requires that such devices be derated to 12A, it's probable that this device is no more than about 1400W or so. So what's the real truth? What's the fuse say? It cannot lie.

(Actually, reading the manual, I see it has a breaker of unknown rating; still it seems clear to me this manufacturer is inflating its figures, which is already off to a bad start with me, albeit I am willing to accept it may be a good amplifier regardless of the misrepresentation).

Joe.

That's a good way to see what the absolute maximum output would be if the power supply could pull that much amperage, but most transformers require a slow-blow fuse that will pass more current than the transformer is rated for to cope with the inrush current when turning the amp on. Most amps will be 60% to 80% efficient, and their power supplies will collapse under a constant load greater than the transformer's VA rating. For a quick minimum estimate, I'd look at the transformer's VA rating and divide by two. The maximum continuous power estimate would be about 80% of the transformer's VA rating.

Charlie,
The 1/10 Ω 5W resistor in the output is there to force the amp to share the load when two or more amps are run in parallel for higher current capability. It certainly isn't needed on a single amp. The output Zobel in the National Semi guides works fine on mine for stability.

[joeybutts]

I've built two amps from 41hz.com. works awesome.

[riverrat373]

How about an inexpensive tube amp kit. Complete except for enclosure/housing. http://www.s5electronics.com/thome.html

I have built a K-502 tube amp from a kit I purchased at Antique Electronics Supply and it sounds very good. I am now in the process of building the K16LS amp from S5 Electronics. These kits are inexpensive, easy to build and sound excellent!

[joekraska]

For a quick minimum estimate, I'd look at the transformer's VA rating and divide by two. The maximum continuous power estimate would be about 80% of the transformer's VA rating.

Noted. And that was very helpful. I'll remember. 80% VA.

What I should have said was "not in your dreams does it pull more than 115v x the fuse rating" along with "not in your dreams does it pull less than 115v x 12AMP" if it fits a standard NEMA 5-15 (standard house wall plug) While I think I accidentally said that the fuse reveals the maximum power of the amp, what I meant to say was that one can be totally sure that this is the ceiling....

This subject somewhat thwarts me by the way--for example, how powerful are Pyle's amps on this site, any way?

Joe.

[DDF]

Also received strong reliable recommendations for:
http://www.chipamp.com/lm3886.shtml

Don't want to start a flame war, so if you don't subscribe to the following, OK, but lets not "get into it". Word on the street is that SQ from these Overture devices is unusually sensitive to component selection, and the chip amp . com devices get a lot of thumbs up in this area.

[Deward Hastings]

http://www.chipamp.com/lm3886.shtml

the chip amp . com devices get a lot of thumbs up Certainly do from me. The 3886 amp boards are simple and clean. I haven't used their power supply board, as my builds have gone into GFA535 chassis and used the existing Adcom power supplies.

[DDF]

Certainly do from me. The 3886 amp boards are simple and clean. I haven't used their power supply board, as my builds have gone into GFA535 chassis and used the existing Adcom power supplies.

How'd it compare to the Adcom?

[Deward Hastings]

How'd it compare to the Adcom?I've never done a direct comparison, and of course the point of the swap is to make a four channel amp for use with electronic crossover, which of course sounds better :). I just don't hear anything from a 3886 that I can identify as distortion . . . within its output limits and loaded by a single driver it qualifies, I think, for Doug Self's "blameless" description. But used thus I doubt that I can hear a properly biased 535 either . . .

[lhwidget]

Noted. And that was very helpful. I'll remember. 80% VA.

What I should have said was "not in your dreams does it pull more than 115v x the fuse rating" along with "not in your dreams does it pull less than 115v x 12AMP" if it fits a standard NEMA 5-15 (standard house wall plug) While I think I accidentally said that the fuse reveals the maximum power of the amp, what I meant to say was that one can be totally sure that this is the ceiling....

This subject somewhat thwarts me by the way--for example, how powerful are Pyle's amps on this site, any way?

Joe.

I thought that's what you were thinking. Just as you are thinking, there are caveats though. I used a 200 VA transformer in a LM3885 stereo chip amp. The caveat here is that the internal protection circuitry of the chips will limit its output to 45 W/channel even though the power supply will provide more amperage. Also, most amp topologies will not deliver peak voltages equal to the power supply's output voltages (the "rail voltages"). Most loose 2 to 4 V, so with a ± 35V power supply, the amp may only be able to deliver ± 31V @ what ever amperage the transformer can supply @ its secondary outputs.

Then, the ability of the power transistors to handle the power dissipation must be taken into account. This is usually the limiting factor controlling the amperage the amp can deliver at the voltages it can swing through.


For small amps like the Pyle (and the much cleaner T Amps), probably the best way to look at it would be to decide how much distortion you can tolerate, and look at the amp's THD vs power plot. With the little chip amps like the Pyle, they are usually rating the amp @ 10% THD. When you look at the power output @ 0.1% or less, it will be more in line with the amp's size & price.

[Æ]

The 1/10 Ω 5W resistor in the output is there to force the amp to share the load when two or more amps are run in parallel for higher current capability. It certainly isn't needed on a single amp. The output Zobel in the National Semi guides works fine on mine for stability.

Reminds me of an emitter resistor (for common emitter amplifier).

[joekraska]

Then, the ability of the power transistors to handle the power dissipation must be taken into account. This is usually the limiting factor controlling the amperage the amp can deliver at the voltages it can swing through.


I've read some of the LM* chip manuals. But in addition to apparent circuitry protection I've wondered about the above. This is in part because I build all my own PC's, and am aware of extreme cooling options in PC space such as water cooling (and obviously quiet fan based cooling) that permit circuitry to operate at higher output than normal (often to include a voltage increase to the CPU). I have therefore been idly wondering what was possible with amps here, although as with my home pcs, I would generally avoid anything like this in extremis.


or small amps like the Pyle .

See this (http://www.parts-express.com/pe/showdetl.cfm?Partnumber=248-484). If that's a 3000W amp, I've uncled a monkey. I'm not sure how one goes about uncling a monkey, but if that's a 3000W amp I'll fess up in gory detail about how I managed my monkey uncling, I swear. :-)

[mattk]

......See this (http://www.parts-express.com/pe/showdetl.cfm?Partnumber=248-484). If that's a 3000W amp, I've uncled a monkey. I'm not sure how one goes about uncling a monkey, but if that's a 3000W amp I'll fess up in gory detail about how I managed my monkey uncling, I swear. :-)

Evidently the FTC no longer cares about power ratings. At one time if an amp had anything even remotely resembling a power rating in its name it had to live up to that power rating using the FTC's method for measuring RMS power. I just looked at the pyle audio site, and they don't have any specs/manual for the so-called 3000W amp. However, for the PT600A, which they call a 300W receiver, there is a manual, and it shows 60W/channel RMS into 6 ohms, which is a bit more believable. Even with that, there is no current draw in the specs.

Another example:

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=248-462

This is called a 2X40 watt amp, but in the product description it gives a more believable 2X8 watts, albeit at 1% distortion, and at 4 ohms.

There ought to be a law. Oh wait, there used to be.

[lhwidget]

I've read some of the LM* chip manuals. But in addition to apparent circuitry protection I've wondered about the above. This is in part because I build all my own PC's, and am aware of extreme cooling options in PC space such as water cooling (and obviously quiet fan based cooling) that permit circuitry to operate at higher output than normal (often to include a voltage increase to the CPU). I have therefore been idly wondering what was possible with amps here, although as with my home pcs, I would generally avoid anything like this in extremis.



See this (http://www.parts-express.com/pe/showdetl.cfm?Partnumber=248-484). If that's a 3000W amp, I've uncled a monkey. I'm not sure how one goes about uncling a monkey, but if that's a 3000W amp I'll fess up in gory detail about how I managed my monkey uncling, I swear. :-)

I was thinking about the little bitty Pyles, but you're right the Pyle you linked to has what appears to be a misprint in the bulleted list and does sound extremely optimistic in its "peak" (probably instantaneous) power ratings. But, instantaneous ratings do give insight into the amp's ability to handle high transient peaks, which is good. For continuous output ratings, the things we were talking about above all come into play.

The band boys don't care as much about the ability to handle high transient peaks, they will be looking for honest continuous power ratings so they can achieve live levels in performance venues.

Just as you're thinking, you can go a long way with water and forced air cooling, but water is complicated, and fans are often loud enough to be obnoxious in a nice home environment (check the second review on the Pyle amp).

With its fan, the Pyle amp probably produces about 100 to 150 w/ch continuous power and its continuous output into 4 & 2 Ω loads is probably about 150% & 175% of its 8 Ω output (I have no idea what the limitations of its "hybrid" amplifier are). It may not be able to maintain that for more than a few minutes before shutting down or blowing up.

But hey, it still sounds like it could be a really nice amp for the money if it's output is quiet and it produces clean, low distortion power. You'd be surprised to see the actual power you're using for casual listening (usually below 5 watts) and a serious session at 90 to 100 dB will probably not require more than 15 - 20 watts on average with 25 - 30 watts peaks and possibly some heavy bass peaks hitting 30 to 60 watts. (These are really broad esitmates.) Using a 100 w/channel (continuous power) amp will give most folks a fair amount of headroom.

Usually, we're reduced to large passive heat sinks for the power stages, and multiple output stages to spread the current load across several output devices. The problem is, even large passives just can't move heat fast enough to let the chips (or output transistors) run at their maximum output for more than a few seconds (sometimes even 10ths of a second) without damage if the amp has been operating near its limit long enough to saturate the heat sink.

[lhwidget]

Also received strong reliable recommendations for:
http://www.chipamp.com/lm3886.shtml

Don't want to start a flame war, so if you don't subscribe to the following, OK, but lets not "get into it". Word on the street is that SQ from these Overture devices is unusually sensitive to component selection, and the chip amp . com devices get a lot of thumbs up in this area.

I'll certainly second that. I've always liked the ChipAmp board designs. However, I built mine point to point with a typical filtered power supply, and it's great, no noise, and no distortion that I can hear.

I think most of the thoughts & theories about a chip amp's "sound" come from the caps (type and the actual value of the component) selected for the input and feedback filters. It's pretty easy to change the amp's frequency response well into the bass range with these components, and the output filtering for stability (output choke and Zobel) can reach down into the 10 to 20 kHz octave. This is probably where the minimalist designs are coming from.

The crusty and venerable AndrewT on the DIY Audio forums
http://www.diyaudio.com/forums/chip-amps/
can offer excellent guidance on the filter designs.