[Folsom]

Alright guys, here it is! First the parts you need. *C4 amended.

From Mouser.

Input caps, order 10x because it's cheaper than 4x (yes you'll need 4x)

647-UES1H010MDM
UES1H010MDM
50volts 1uF

Power cap 1x

661-EKZH250E222MK30S
EKZH250ELL222MK30S
25Volts 2200uF

Shunt/Mute Power cap

661-EKY500ETC220ME1
EKY-500ETC220ME11D
50volts 22uF

Building

Ok, so first I recommend getting to this point, the naked board of all the parts you don't need. That is unless you want the volume pot, but I wouldn't recommend it.

Here's a clear board, I've labeled the R and L inputs, and their grounds. R1, D1, D2, C1, C2, C3, C4 (not shown removed) C5, CN4, and RP 50k have been removed.



Install a wire for the bypass of the diode (D2), I did this on the bottom.



Install the Shunt/Mute cap (C3) 22uf, negative side (short lead) goes to the striped pattern on the board.





Install the upside C1 and C2 capacitors, 1uf Nichicon Muse. Long leads go toward the input (as circled)



Install on the bottom 1uf capacitors with the reverse orientation of the first two (long leads towards the chip), on C1 and C2. This is very important, and can only be done with non-polar caps. Without doing this your amplifier won't be that great unless you got some very nice film caps for the inputs.



Install C5, the big 2200uf cap, again stripe to stripes.



That's it! My next post here will be about advanced options, this is the simple, but very good build.

Use only DC power supply after removing the diode.

*do not replace C4, it's a worthless snubber that causes ringing.

[Folsom]

Advanced options

Adding capacitance to the DC input, multiple capacitors. You can use whatever size you want 470-2,200uf. I think 1,500uf, 1000uf, are decent sizes. Panasonic FC's are good, but I'm really into the low impedance ones offered from Mouser now, the United Chemi-con, KZH and others. whatever low impedance one. For an amplifier of this size 9000-12000uf is as much as you'll probably need. That's for 15w amplifier.

[Folsom]

Burn in time isn't too short with these mods. Expect much sweeter sound somewhere after 20 hours or so...

as in don't judge it too fast!

[rhing]

Your first amp stopped making sound--presumably from using an SMPS that might have triggered a fault or something in the TDA7297 chip. Now that you've modded another unit, do you still support the idea of removing the polarity protection diode and replacing it with the jumper wire between the Vcc input and the rest of the amp's power supply path? You mentioned before that do this increased soundstage depth. Also, which power supply are you now using in terms of DC voltage and current?

[Folsom]

I switched to a 12v. This is probably the smart way to go. I can guarantee my chip failed for a particular reason even, but this new one has been playing for a few days. Besides that if it was a voltage problem the diode wouldn't stop it.

All the datasheets don't indicate any need for a diode. The diode is purely so you can use AC power. But it probably sounds like poop with AC.

Do I support it? I wouldn't use it with it, literally, because my different amplifiers I've previously used were better before removing the diode. But I was in a little bit of a panic when the first one failed since I didn't bother to order two or more, ha! They take awhile to arrive.

[Folsom]

I have a little bit of bad news for everyone... I tried listening to this amplifier without it being plugged into my power conditioner... It loses the magic. But that's true with all amplifiers. Trouble is it loses too much.

Looks like I might have to develop a linear supply that can provide comparable results without a conditioner (have it built in). Such as it goes... not a big surprise since I've never heard an SMPS sound very good in anything non-constant current supplied; baring a good power conditioner.

[S Clark]

Why not skip the power supply and just go straight to a big amp/hour battery? 

[Folsom]

Maybe. I haven't been totally convinced of them. The sound is usually good, but their's an edge I don't always get. Not the bad one, a good one.

Plus it's not unusual that I want to listen longer than 6 hours or whatever.

[wushuliu]

I have a little bit of bad news for everyone... I tried listening to this amplifier without it being plugged into my power conditioner... It loses the magic. But that's true with all amplifiers. Trouble is it loses too much.

Looks like I might have to develop a linear supply that can provide comparable results without a conditioner (have it built in). Such as it goes... not a big surprise since I've never heard an SMPS sound very good in anything non-constant current supplied; baring a good power conditioner.

Or just get a TPA3116 instead. Can't believe you're going all on the TDA without having heard the 3116.

yep, hijack!

[Folsom]

I'll give it a try, but I'll be very surprised if it's too much better. I can tell you the 3875 amplifier can't come even close with a nice linear PSU and power conditioning.

[wushuliu]

I'll give it a try, but I'll be very surprised if it's too much better. I can tell you the 3875 amplifier can't come even close with a nice linear PSU and power conditioning.

3875 is totally unrelated to the TI TPA3116. The 3116 is Class D sibling chip to the one used in the TBI Millenia.

More here:

http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf

one of many versions here:

http://www.ebay.com/itm/Assembled-TPA3116-store-Power-amplifier-board-50W-50W-/400560249079?pt=US_Amplifier_Parts_Components&hash=item5d434058f7

So far from comparisons the general feedback is the 7297 has a more impressive soundstage but the 3116 sounds more refined to some with better top end. Granted the TPA3116 boards are a whopping $20-28 compared to the $6 7297, but it's certainly worth comparing.

[Folsom]

3875 was just for referance of overall. You could buy $1200 3875 monoblocks that sound like a dead dog fart compared to the TDA7297, if you ask me.

Perhaps someone should let me try theirs. I am fooling around with TDA7297 deciding if I want to make it a product, because I am so impressed. Unfortunately it appears it needs extra care in the power arena.

[S Clark]

... You could buy $1200 3875 monoblocks that sound like a dead dog fart compared to the TDA7297,

Hmmmm, now I know what was in that old NAD amp I used to have.   It did a great job creating realistic dead dog farts with a super soundstage.

[rhing]

So far from comparisons the general feedback is the 7297 has a more impressive soundstage but the 3116 sounds more refined to some with better top end. Granted the TPA3116 boards are a whopping $20-28 compared to the $6 7297, but it's certainly worth comparing.

Just returned home from a business trip, and discovered that my Sure Electronics TPA3110D2 amp module arrived from Parts Express. I will make the connections in my system tomorrow and begin some comparisons between the TI Class D amp and the TDA7297. I have big expectations for the TPA3110D2.

[Folsom]

Oh good, maybe I'll just wait for word from you. The surface mount parts don't look as fun to work with, but who knows on sound. rhing are you going to try battery?

I'm skeptical given how good the TDA7297 sounds with conditioned power. I might have to try an experiment plugging it into another socket not near my conditioner. It sounds pretty darn good at a friends house without a conditioner. But... well, we'll see.

[rhing]

Oh good, maybe I'll just wait for word from you. The surface mount parts don't look as fun to work with, but who knows on sound. rhing are you going to try battery?

After listening to this amp, I have to say I am very impressed with this Sure TPA3110D2 amp board. I have to qualify this statement with the fact that I connected this amp to my Audio Research LS7 vacuum tube line stage preamp, since I do not have any spare 50kohm volume pots except the cheap stock volume pots that came with my two TDA7297 amps. I used the Audio Research preamp to handle volume control duties. At this point, I cannot make a direct apples-to-apples comparison between the Sure Class D amp and the TDA7297 Class A/B amp, but I will say that the combination I have been listening to sounds better, overall, than the TDA7297 in stock or modified form.

• More detailed or articulate
• Deeper and cleaner bass--less boomy
• Higher frequency extension
• More natural tonality
• Soundstage is comparable with a large, deep soundstage
• Greater dynamic range

I feel like I'm splitting hairs with this, but I do hear the difference with careful listening with music material that I am very familiar with. Using my Alan Parson's "Sound Check 2" test CD which has a series of 1/2 octave tones from 20Hz up to 20kHz, I could verify that the bass response is lower with the TPA3110D2 than with the TDA7297. With the TPA3110D2, I could hear down to 31.5Hz in my room. I don't hear this with the TDA7297. In fact, the bass seems to roll off around 45Hz. What the TDA7297 does do is it seems to have a stronger bass output between 80 and 120Hz, which can give the impression that the bass is stronger.

I still have to sort this out, but I do hear some low level background noise with the TPA3110D2. The noise disappears when I engage the MUTE on my LS7, which indicates that the output filter-less TPA3110D2 amp may be emitting some RFI that's affecting my preamp. I never hear this noise, or any for that matter, from the TDA7297 or my Dynakit Stereo 35 tube amp. If this is something that cannot be reduced/eliminated, than an amp module such as the Yuan Jing TPA3116D2 2.0 amp might be a better solution since it does have the LC output filter to minimize RFI.

I still have more testing and listening to do, but the Texas Instruments Class D-proponents like wushuliu are probably onto something really good with these Class D amps.

[Folsom]

Interesting. Do you have any power conditioning on your smps?

[MLS]

I am getting some noise with TPA 3110 also when I connect it to a pre-amp.  It is dead silent when it is connected to an Apple AirportExpress.  I have not been able to figure it out yet.

[rhing]

I am getting some noise with TPA 3110 also when I connect it to a pre-amp.  It is dead silent when it is connected to an Apple AirportExpress.  I have not been able to figure it out yet.

Thanks for pointing that out MLS. I just connected my iPhone to the amp, and it's quiet just as you described.

[rhing]

Here's a clear board, I've labeled the R and L inputs, and their grounds. R1, D1, D2, C1, C2, C3, C5, CN4, and RP 50k have been removed.



Install a wire for the bypass of the diode (D2), I did this on the bottom.


....

Use only DC power supply after removing the diode.

I removed the diodes and installed solid core Copper wire jumpers in their place. The improvements in frequency extension, better dynamic response and detail were subtle, but worth the effort. If anyone has this amp and has a decent DC power supply, this is a worthwhile mod.