TPA3116D2 Amp

Jump to Latest
#5,522
kp93300 said:
jk44,
what material do you use to glue the heatsink to tpa3116?
kp93300
Click to expand...

I used some thermal paste, OCZ ultra 5+ silver compound. It is normally designed for CPU heatsinks and overclocking. It is not an adhesive, so the heatsink is not permanantly fixed. As others have mentioned, the TPA3116 doesn't really need the heatsink, I've run it without the heatsink and was only a little warm to the touch
 
#5,523
irribeo said:
On MarkAudio speakers:
Panasonic FM has a little lifted midrange causing some sibilance sometimes or you could say they sound snappy if you like it. They attenuate treble a little and lose detail there, I mentioned cymbals becoming just noise befor. Oscons attenuate treble, but do not lose detail, but do lose mid and high dynamics. They sound restrained, that is different from FM that pushes focals a little forward as a manner of speaking, mid sounds very dynamic. Now combine and see what happens 🙂 Oscons need 100 hour to sound the way I tried to describe.
Click to expand...

The midrange sounds a little restrained after putting the OSCONs in, upper bass and upper midrange sounds a little lacking in emphasis. I found the FMs on their own didn't have much sibilance like you described.

It sounds like they could work together, but I guess it needs some burn in? Any other things worth trying? I can try trimming down the legs a little erik777, although it will make it trickier to solder in there 🙂
 
#5,524
I left ampboard powered 24 hours for a couple of days and playing music during daytime🙂 6 Oscons did improve. Maybe you put Oscon sound better into English than I do 🙂 6 Nichicon FP RNU are quite a different solid, more detail and better dynamics everywhere, and no sibilance or harshness heard, but that really amazed me, because mid/highs have so much more energy it is asking for trouble I thought LOL 6 Chemicon PSG 390uF solids are in between Oscons and FP RNU's, so less energy treble than Nichicons but a little harshness, sibilance is added. Somehow Nichicon avoided that. Both these 2 others also need quite some time. I think it could be shortened if rated voltage is applied for an hour or so befor use like datasheets suggest to get leakage down to already very high datasheet values (compared to wet electrolytics), at least if it is related??? I used 12V PSU these comparisons.
 
#5,525
irribeo said:
I left ampboard powered 24 hours for a couple of days and playing music during daytime🙂 6 Oscons did improve. Maybe you put Oscon sound better into English than I do 🙂 6 Nichicon FP RNU are quite a different solid, more detail and better dynamics everywhere, and no sibilance or harshness heard, but that really amazed me, because mid/highs have so much more energy it is asking for trouble I thought LOL 6 Chemicon PSG 390uF solids are in between Oscons and FP RNU's, so less energy treble than Nichicons but a little harshness, sibilance is added. Somehow Nichicon avoided that. Both these 2 others also need quite some time. I think it could be shortened if rated voltage is applied for an hour or so befor use like datasheets suggest to get leakage down to already very high datasheet values (compared to wet electrolytics), at least if it is related??? I used 12V PSU these comparisons.
Click to expand...

Yes I recall you mentioning earlier the big difference between the OSCONs and the FPCAPs. I will try to find but digikey (which i prefer over mouser) doesn't have the 25V/330

I have a SMSL SA36A Pro (TPA3118) that I put 8 panasonic FM, What a delightful sound. The difference is 10uF metalized polyester input caps I think, I think the amp I'm working on (Hifi 2.0 by WLX) needs to get rid of those tiny electrolytics on the input (behind the WIMAs)
 
#5,526
jk44 said:
I used some thermal paste, OCZ ultra 5+ silver compound. It is normally designed for CPU heatsinks and overclocking. It is not an adhesive, so the heatsink is not permanantly fixed. As others have mentioned, the TPA3116 doesn't really need the heatsink, I've run it without the heatsink and was only a little warm to the touch
Click to expand...
I was wondering why the heatsinks on these 3116 boards are so tiny.

you can get heat conductive double sided tape. Bought on ebay, Ive used that extensively in the past on everything from PC RAM modules to amps. Works surprisingly well.

its impressive how efficient these boards are that there is that little waste heat generated.
 
#5,527
Evan55 said:
I was wondering why the heatsinks on these 3116 boards are so tiny.

you can get heat conductive double sided tape. Bought on ebay, Ive used that extensively in the past on everything from PC RAM modules to amps. Works surprisingly well.

its impressive how efficient these boards are that there is that little waste heat generated.
Click to expand...
I was definitely considering that, or a heatsink marketed for raspberry pi chips, that use the adhesive tape as well. But I realized that it runs only warm to the touch that I could just thrown the little spare heatsink I had, even if it's just not the tightest adhering
 
#5,529
irribeo said:
Situation isn't very clear to me. I read dac-B1pre-3116 but also many references to a pot and "direct" and/or all kind of volumecontrols in different places being used. Was this an existing situation that played with another poweramp?
Click to expand...

Yes, I tried eliminating some of the stages to see if I could identify the offender. The noise is absent only with direct connection from (low volatage output) ipod and without a physical pot. Whenever a pot in inserted either via B1 or as passive it seems to pick up some noise. If connected straight to external DAC without volume pot (so at full volume) the noise is a loud background hiss and and a not so loud buzz. They both go away if music is played but come back if for example music is paused (but source is still connected). I guess I will keep trying moving things around. I suspect a ground issue, although I don't get how a simple passive pot can create a ground problem.
 
#5,530
marcel5996 said:
Yes, I tried eliminating some of the stages to see if I could identify the offender. The noise is absent only with direct connection from (low volatage output) ipod and without a physical pot. Whenever a pot in inserted either via B1 or as passive it seems to pick up some noise. If connected straight to external DAC without volume pot (so at full volume) the noise is a loud background hiss and and a not so loud buzz. They both go away if music is played but come back if for example music is paused (but source is still connected). I guess I will keep trying moving things around. I suspect a ground issue, although I don't get how a simple passive pot can create a ground problem.
Click to expand...

If things dont work out, just buy the sure tpa3116 for $ 23 and add the oscons. And youll like what u h.ear
 
#5,531
jk44 said:
I was definitely considering that, or a heatsink marketed for raspberry pi chips, that use the adhesive tape as well. But I realized that it runs only warm to the touch that I could just thrown the little spare heatsink I had, even if it's just not the tightest adhering
Click to expand...
I'm surprised that I haven't yet seen a back side brace to bolt the heat sink to.

The data sheet spec heat sink is probably enough to run at full power for several minutes in the Sahara desert without exceeding the maximum temperature spec but cooler is better and the little bit of excess from a back brace would only take a few minutes work.

An added bonus is that it would eliminate any bend from the heat sink fasteners in the circuit board.
 
#5,532
marcel5996 said:
Yes, I tried eliminating some of the stages to see if I could identify the offender. The noise is absent only with direct connection from (low volatage output) ipod and without a physical pot. Whenever a pot in inserted either via B1 or as passive it seems to pick up some noise. If connected straight to external DAC without volume pot (so at full volume) the noise is a loud background hiss and and a not so loud buzz. They both go away if music is played but come back if for example music is paused (but source is still connected). I guess I will keep trying moving things around. I suspect a ground issue, although I don't get how a simple passive pot can create a ground problem.
Click to expand...

Sounds very much like grounding issue.
 
#5,533
A poweramp with open input has hiss. Here if I connect a finished product, preamp or source it is gone, source doesn't need to be powered on. YJblue has nothing on input and I think most preamps/sources do have something like 47k to ground on output. So add like 47k resistors from signal to ground somewhere I guess could help??? Never needed it myself, so just guessing 🙂
 
#5,534
irribeo said:
A poweramp with open input has hiss. Here if I connect a finished product, preamp or source it is gone, source doesn't need to be powered on. YJblue has nothing on input and I think most preamps/sources do have something like 47k to ground on output. So add like 47k resistors from signal to ground somewhere I guess could help??? Never needed it myself, so just guessing 🙂
Click to expand...

Indeed, all the various tpa3116 amps I've played with have a hiss with no input connected. I tried putting a 100k (I think?) from input to ground, but that didn't help. But simply connecting a source fixes it, so I stopped worrying about it.
 
#5,535
Panasonoc OSCON's vs Nichicon's ???

erik777 said:
Because of Irribeo i tried nichicons. So I pulled the oscons off my sure board. And replaced with the nichons caps, a very nice noticable improvement. 🙂
Click to expand...

erik777 said:
If things dont work out, just buy the sure tpa3116 for $ 23 and add the oscons. And youll like what u h.ear
Click to expand...

Oscon's or Nichicon's???
Mu Sure3116+6OSCONs sounds great. Is Nichicon's beater and which?
 
#5,537
okay Guise,

I finally tried to do a build worthy of the experts on here. This is an amplifier I made as a gift for a friend's wedding. It features YJ blue/black with input cap, power supply cap, bootstrap cap, and snubber mods...

Basically every good mod except the inductors. And, the volume pot is just a cheapy 50k from YJ.

I also added a board that runs the power-on LED indicator as well as some spare capacitance using 2x1500uF Panasonic FM caps.

Mounted in a cedar souvenir box I found in a thrift store.

A very good friend with a drill press helped me mount all the connectors without making it a total hack job.

Interesting thing here is that when I finally completed it and hooked it up it sounded terrible. I have a set of the Pioneer BS22 speakers with the Dennis Murphy mods and all I heard was serious distortion and crap.

So, I started to troubleshoot only to realize that I was actually hearing the amp burn in or something. All the distortion quickly faded away and in the midst of A/B testing with my desktop version of this amp--same mods plus coilcraft inductors and ALPS blue velvet volume pot--I quickly coudl no longer discern the difference.

So, I guess I literally heard those FM caps and the other stuff burn in ...

Within 20 minutes it sounded great.

I want to thank Rhing and all of the other experts on here. Without your pictures and work I would have never been able to put this together.

Now, I need to build one for myself. It pains me to have to give this away.

https://www.flickr.com/photos/45066974@N08/15586342906/in/photostream

https://www.flickr.com/photos/45066974@N08/14989263424/in/photostream/

https://www.flickr.com/photos/45066974@N08/15586339716/in/photostream/

https://www.flickr.com/photos/45066974@N08/14989258944/in/photostream/

https://www.flickr.com/photos/45066974@N08/15423897828/in/photostream/

https://www.flickr.com/photos/45066974@N08/15610840632/in/photostream/
 
You must log in or register to reply here.
Top Bottom