TPA3110 and TPS3116 amps

rhing · « **Reply #60 on:** 26 Dec 2013, 04:24 pm »

Basically, I inserted the 33kohm resistors between the signal inputs and grounds as shunts to channel away extraneous noise. In effect, I just created a filter. Anyway, I found some 47kohm PRP 1/4 watt metal film resistors in my parts bin (not sure why I didn't see them before), and put those in. The sonic effect was a shrill quality with an exaggeration of the treble, and of course noise. The lower octaves were rolled off, and the soundstage lost all depth. I then tried some 20kohm PRP resistors and this is much better. It really doesn't seem to be different than the results I obtained with the 33kohm resistors, which were KOA Speer Carbon film resistors. I'll stick with the 20kohm PRP resistors for now.

Perhaps someone with an electrical engineering mind can better explain what I did and present some mathematical explanations. I would if I could, but I don't have the savvy or enough sophisticated tools to characterize the amp's input circuitry.

raysracing · « **Reply #61 on:** 26 Dec 2013, 04:35 pm »

The one I posted a pic of has some hiss and on my new/used Omega speakers there is so much treble that I can't even play a record because it accentuates the surface noise and what used to be barely audible pops to the front. Even setting the needle down sounds different. All my digital media is detailed, bright and unlistenable right now. Like an amazing tweeter/mid with the woofer removed and the treble turned up to boot. The speakers should settle down (and are), but sadly I had no idea an amp/speaker combo could be so far off. I even switched this am to an old integrated Yamaha DSP amp. It helped a bit. Man I wanted to use the 3116 with these Omegas.

I'll add some 47kohm resistors when I get home from vacation on several of my chip amps and see if there is any chance of finding a happy marriage with these Omegas.

rhing · « **Reply #62 on:** 26 Dec 2013, 05:09 pm »

Quote from: raysracing on 26 Dec 2013, 04:35 pm

The one I posted a pic of has some hiss and on my new/used Omega speakers there is so much treble that I can't even play a record because it accentuates the surface noise and what used to be barely audible pops to the front. Even setting the needle down sounds different. All my digital media is detailed, bright and unlistenable right now. Like an amazing tweeter/mid with the woofer removed and the treble turned up to boot. The speakers should settle down (and are), but sadly I had no idea an amp/speaker combo could be so far off. I even switched this am to an old integrated Yamaha DSP amp. It helped a bit. Man I wanted to use the 3116 with these Omegas.

I'll add some 47kohm resistors when I get home from vacation on several of my chip amps and see if there is any chance of finding a happy marriage with these Omegas.

I thought your amp had the tone controls. Have you tried using them?

As for the resistor value, you may need to experiment as I did to determine the best values for your system/tastes.

raysracing · « **Reply #63 on:** 26 Dec 2013, 05:12 pm »

Rhing when I turned down the treble it sucked out more than just the unrealistic sound of the cymbals, etc. It sucked out the fullness of the sound. I messed around with the EQ and parametric in Jriver too. Same effect as the tone controls once you affected the brightness to any degree you made it sound flat or thin.

rhing · « **Reply #64 on:** 26 Dec 2013, 05:24 pm »

That's unfortunate. I really like Omega speakers for their huge sound and high efficiency. I met Louis Chochos at a Rocky Mountain Audio Fest in 2007, and he was one of the nicest people I met at the show.

All the tweaking I've done still hasn't completely eliminated the hiss noise. On the other hand, my TDA7297 Class A/B chip amp is dead silent. My SMSL SA-S1 Tripath TA2020 T-amp is not dead silent, but significantly less noisy than the TPA3110D2. Still, with the little bit of noise I hear, the sound quality of the TPA3110 amp is more pleasing to my ears than either of the other two amps.

lacro · « **Reply #65 on:** 26 Dec 2013, 05:39 pm »

I finished the case for my "Little Ass Kicker" Connected it temporarily to a laptop in my living room, driving the C&C earlier version of the Pioneer's ($39 shipped). This little low powered amp fills the room with great sound.

rhing · « **Reply #66 on:** 26 Dec 2013, 07:45 pm »

lacro,

Very beautiful work there. Thanks for sharing your photos. Which caps did you put on the power supply? They look like they might be 2,200uF caps or something else significantly higher in value than the stock 220uF power supply caps.

I'm considering the Pioneer speakers for my office, and it's good to know that the TPA3110 has enough power to drive them.

wushuliu · « **Reply #67 on:** 26 Dec 2013, 08:15 pm »

Quote from: rhing on 26 Dec 2013, 07:45 pm

lacro,

Very beautiful work there. Thanks for sharing your photos. Which caps did you put on the power supply? They look like they might be 2,200uF caps or something else significantly higher in value than the stock 220uF power supply caps.

I'm considering the Pioneer speakers for my office, and it's good to know that the TPA3110 has enough power to drive them.

The higher PS cap values improve low end response for a more fuller sound in my experience. Others have reported the same. I've also found PS caps to have a more tonal impact than say input cap mods, etc. I prefer Nichicon Muse myself for clean and clear and microdynamics. Elna Silmics are very warm. It's easy to try out on the 3110 board since it has surface pads making it very simple to solder and desolder. And PS caps are so inexpensive. Worth trying out.

rhing · « **Reply #68 on:** 26 Dec 2013, 08:41 pm »

What capacitance value do you recommend?

The lead spacing for the power supply caps is 3.5mm. The largest radial-leaded electrolytic cap size with a 3.5mm lead is 680uF / 25V if I get Panasonic FR caps. The Nichicon Muse KZ and Elna Silmic II caps in larger values are only available with 5mm lead spacing. Are you bending the leads to fit the thru-hole spacing, or are you using the SMD pads to mount the Nichicon caps? With having to bend the capacitor leads to fit either the smaller thru-hole spacing or the SMD pads, it seems like there would be some parts of the leads that are susceptible to picking up RFI from the chip.

wushuliu · « **Reply #69 on:** 26 Dec 2013, 08:50 pm »

Quote from: rhing on 26 Dec 2013, 08:41 pm

What capacitance value do you recommend?

The lead spacing for the power supply caps is 3.5mm. The largest radial-leaded electrolytic cap size with a 3.5mm lead is 680uF / 25V if I get Panasonic FR caps. The Nichicon Muse KZ and Elna Silmic II caps in larger values are only available with 5mm lead spacing. Are you bending the leads to fit the thru-hole spacing, or are you using the SMD pads to mount the Nichicon caps? With having to bend the capacitor leads to fit either the smaller thru-hole spacing or the SMD pads, it seems like there would be some parts of the leads that are susceptible to picking up RFI from the chip.

I cut the leads as short as I can get then bend them to fit SMD pads. Hm, I never had RFI problems with my 3110 when replacing the caps. The leads are pretty short once u cut them and bend. Can't imagine 1/4in or less leads wreaking RFI havoc. No recommended value per se. I think I used 1500uf, some use higher. I think 1000uf would be fine.

lacro · « **Reply #70 on:** 26 Dec 2013, 10:50 pm »

Quote from: rhing on 26 Dec 2013, 07:45 pm

lacro,

Very beautiful work there. Thanks for sharing your photos. Which caps did you put on the power supply? They look like they might be 2,200uF caps or something else significantly higher in value than the stock 220uF power supply caps.

I'm considering the Pioneer speakers for my office, and it's good to know that the TPA3110 has enough power to drive them.

Thanks for complement, but I am just a first timer with NO electronics knowledge that doesn't really belong here

I just assembled this with a bunch of scraps, and a $1.79 wooden box I got at A.C. Moore. The caps are these: http://www.digikey.com/product-detail/en/EEU-FM1E222L/P12384-ND/613745 This was my first time desoldering/soldering on a board.

wushuliu suggested a cap swap from stock which was quickly noticeable in SQ

The 5MM lead spacing was not a problem as the difference between 3.5MM and 5MM is only about 1/16" total. That's not difficult to bend, and make the leads almost as short as the stock caps. Remember this is my very first soldering job on a board

However, it was much easier than I thought..

I have tried this amp so far with my Paradigm Atoms, the Pioneer's, and right now I am listening to my XL-S Encores powered by this little amp. I think you will be surprised it will drive the Pioneer's with authority even though they are not that efficient, and should need more power.

I put the Pioneer's which were not even broken in, on stands, and moved them into the room away from any wall. My living room is 16x24, and these little speakers, and amp filled the room with wonderful sound. 80db SPL at 15' with peaks to 85db. This was at 50% volume.

My wife didn't like it when I removed the little system. To her it sounded better than her floor standers

Good thing I ordered another 3110 from PE. I may even build it into the Pioneers

It's driving these right now...

http://i.imgur.com/4waDmZ3l.jpg

raysracing · « **Reply #71 on:** 27 Dec 2013, 01:23 am »

Quote from: rhing on 26 Dec 2013, 05:24 pm

That's unfortunate. I really like Omega speakers for their huge sound and high efficiency. I met Louis Chochos at a Rocky Mountain Audio Fest in 2007, and he was one of the nicest people I met at the show.

All the tweaking I've done still hasn't completely eliminated the hiss noise. On the other hand, my TDA7297 Class A/B chip amp is dead silent. My SMSL SA-S1 Tripath TA2020 T-amp is not dead silent, but significantly less noisy than the TPA3110D2. Still, with the little bit of noise I hear, the sound quality of the TPA3110 amp is more pleasing to my ears than either of the other two amps.

Just a couple more hours after posting they were producing more
mid bass already. I may spring for the RS5 drivers. Thanks Rhing for all the R&D and sharing. Really helps a greenie like me and gives me confidence to try more things.

rhing · « **Reply #72 on:** 27 Dec 2013, 02:35 am »

There's no shame in any DIY project. I got started in DIY audio when the first Sonic Impact T-amps were the rage. Low voltage Class D amps are great for first-time DIYers as 12V DC is not lethal and if you screw up, the amps are affordable. On top of that, the parts count is relatively low and you have more options for enclosures since these amps don't radiate much heat, if any.

Folsom · « **Reply #73 on:** 27 Dec 2013, 04:08 am »

1-1.5k caps are fine, but you'll want enough to get 8-12k total per voltage supply. Meaning probably not on board. It's good to keep them near the same or the same uf to prevent frequency ringing.

For input caps 50/r = uf . So if the input resistance of the chip is 25, 2uf would be good. I'm on my phone so i can't look the amp chip up.

For input caps you want higher impedence, power supply lower; generally.

My use of Nichicon Muse ES, two caps one in reverse orientation, is probably impossible to beat for the money.

rhing · « **Reply #74 on:** 27 Dec 2013, 06:47 pm »

Quote from: lacro on 26 Dec 2013, 10:50 pm

wushuliu suggested a cap swap from stock which was quickly noticeable in SQ The 5MM lead spacing was not a problem as the difference between 3.5MM and 5MM is only about 1/16" total. That's not difficult to bend, and make the leads almost as short as the stock caps. Remember this is my very first soldering job on a board However, it was much easier than I thought.

I happened to have a few Panasonic FM 1,500uF / 25V in my parts bin, and I guess I was worried about nothing with the lead spacing. These caps have 5mm lead spacing and they fit fine in the thru-holes. The TI datasheet mentions that low ESR capacitors should be used, and the Panasonic FMs fit this description. What's interesting is that the low level noise has been decreasing. I'm not sure why this is the case, but I took off those 33kohm resistors and heard no changes. The Panasonic caps are still burning in. I'll share my listening impressions later. Thanks for the help guys.

shadowlight · « **Reply #75 on:** 27 Dec 2013, 07:48 pm »

What happens if you get something like Nichicon UHW1V222MHD (2200uf, 35V) with 5mm spacing work? Just trying to understand what happens when you change stuff around. I am going to get some 2200uf like lacro mention also.

http://www.digikey.com/product-detail/en/UHW1V222MHD/493-6871-ND/3664337

lacro · « **Reply #76 on:** 27 Dec 2013, 08:36 pm »

Quote from: shadowlight on 27 Dec 2013, 07:48 pm

What happens if you get something like Nichicon UHW1V222MHD (2200uf, 35V) with 5mm spacing work? Just trying to understand what happens when you change stuff around. I am going to get some 2200uf like lacro mention also.

http://www.digikey.com/product-detail/en/UHW1V222MHD/493-6871-ND/3664337

Not sure 2200uf is the right value, most are using 1000-1500uf. Like wushuliu said, it's pretty easy to swap the caps around trying different ones. I would listen to other members with real electronics know how. My cap choice was just a nubee playing around, but it sure sounds better to my ears than the stock caps, YMMV. On my next board I will try something different to compare.
BTW/ while writing this I am smiling

listening to this amazing little amp that was the price of a lunch.... I think I may even buy Texas Instruments stock.

shadowlight · « **Reply #77 on:** 27 Dec 2013, 08:39 pm »

Lacro,

I am a newbie with swapping parts out and replacing them with something else. The only diy work that i have done is building the class-d amp (the long thread) but that was easy since all i needed to do was hook up thing and little bit of soldering

. I taking a step forward with the 3110 and 3116 boards.

rhing · « **Reply #78 on:** 27 Dec 2013, 10:50 pm »

I'm sure the Panasonic FM 2,200uF / 25V caps would be fine. The Panasonic FM caps I installed earlier today are still burning in, but I can already hear a huge difference with this Sure amp. Interestingly, the noise that was annoying me when I first got this amp has significantly decreased to the point where I removed those 33kohm resistors. I don't need them anymore. The bass is significantly deeper and tighter. The soundstage has grown in all three dimensions. I am really impressed with the dynamic range and detail too. Basically, the only mod/tweak that is on the amp is the power supply cap swap. I also think moving the RCA connectors closer to the board's inputs helped too. I wish I did this cap swap from the beginning, and I could have avoided the anxiety over the noise.

Folsom · « **Reply #79 on:** 28 Dec 2013, 12:53 am »

Smaller caps discharge quicker. They're rather effective, where as giant caps won't be except in lower frequency extension. That is when comparing lower to moderate priced capacitors. You can spend $119 a cap or whatever, and get similar performance from a 50kuf to a 1kuf.

It's easier to just use multiple caps to get very low ESR, and keep them small and fast. (define small, I know, you're choice, play around and see what you like?)

As noted here by another member, what I'm talking about is real;

Here . You'll notice some skeptics, people suggesting more complicated (and to my ears often inherently bad sounding) filtration ideas, but at the very end the real skeptic tried it out and now is a total convert.

It's just the behavior of class D, why fight it?