I do have the YJBlue 2.0 board - I repeat - cool to the touch.
What is ambient room temperature? Is the board in a chassis? Are there other heat sources in the chassis? Active chassis cooling or passive? If passive, how it is it designed?
All of these things have a direct effect on the temperature of the IC and other components. If I let the board sit by itself on a table, it would run cool, but I do not intend to use it this way.
In my case, ambient room temp is ~30C. The board is in a chassis with an SMPS (which gets "warm" but not "hot"). The chassis has series of holes drilled in appropriate places to allow heat convection, but this is not sufficient to keep the board cool. It isn't warm, it is very hot. I added active cooling to the chassis and it solved the problem. Next problem is, I don't want active cooling.
black & blue
Replaced 2 DC caps, and 4 inductors with Elko's 1000µF/35V BC-Components, and Coilcraft's on YJ black board, few hours ago. Original inductors were warm-to hot before, and all the time, so was heatsink of the chip. Now everything is dead cool, after two hours of playing music! These issues fixed
ON/OFF 'pops-cracks' remain the same 😡
Replaced 2 DC caps, and 4 inductors with Elko's 1000µF/35V BC-Components, and Coilcraft's on YJ black board, few hours ago. Original inductors were warm-to hot before, and all the time, so was heatsink of the chip. Now everything is dead cool, after two hours of playing music! These issues fixed

ON/OFF 'pops-cracks' remain the same 😡


Replaced 2 DC caps, and 4 inductors with Elko's 1000µF/35V BC-Components, and Coilcraft's on YJ black board, few hours ago. Original inductors were warm-to hot before, and all the time, so was heatsink of the chip. Now everything is dead cool, after two hours of playing music! These issues fixed
Nice!
This gives me hope! I'll be changing my inductors right after work (and checking my gain resistors).
It's too bad that YJ is cheaping out their components to the point that the SQ suffers and the parts run hot. To be expected from a low cost board selling for $20. But not all the boards are that way - mine never got hot and did not sound awful stock. The Ybdz Wiener (http://www.aliexpress.com/item/Free-Shipping-1pc-TPA3116-Class-D-digital-amplifier-50W-50-W-with-Silent-Sleep-Design/1586764360.html)which sells for $13.66 appears to use high quality parts as it is not even warm, sound good, and has near ideal layout with SMT parts very close to chip pins. I may be ordering a few more of the Ybdz just to have on hand. They sound and work well straight out of the box.
@Jadran: those Coilcrafts look nice, and seem to have fixed the running hot issue. Which model Coilcraft is that?
@Jadran: those Coilcrafts look nice, and seem to have fixed the running hot issue. Which model Coilcraft is that?
Last edited:
So which parts need to be changed to keep it running cool? I just got the YJ blue/black board and have ordered replacement bootstrap caps. I'm assuming power caps will need to be changed? I'll be creating a one chassis integrated along with the Little Bear 6N3 preamp board so heat will be an issue.
Interesting. I guess I could always stick some 10uH coils on it for 4ohm load.The Ybdz Wiener (Free Shipping! 1pc TPA3116 Class D digital amplifier 50W +50 W with Silent Sleep Design-in Other Electronic Components from Electronic Components & Supplies on Aliexpress.com)which sells for $13.66 appears to use high quality parts as it is not even warm, sound good, and has near ideal layout with SMT parts very close to chip pins. I may be ordering a few more of the Ybdz just to have on hand. They sound and work well straight out of the box.
So which parts need to be changed to keep it running cool? I just got the YJ blue/black board and have ordered replacement bootstrap caps. I'm assuming power caps will need to be changed? I'll be creating a one chassis integrated along with the Little Bear 6N3 preamp board so heat will be an issue.
I've changed a lot on my board (power caps, input caps, bootstrap caps) and the heat issue persists. Looks like inductors are key here, and as someone suggested earlier; double-check gain resistors. I'm doing both after work today and will report back.
Room temperature around 23 degrees C. The boards sits in a simple chassis stolen from another class D amp - but without cover.
Room temperature around 23 degrees C. The boards sits in a simple chassis stolen from another class D amp - but without cover.
Under those conditions, mine would likely run cool as well.
@Jadran: those Coilcrafts look nice, and seem to have fixed the running hot issue. Which model Coilcraft is that?
RFS1412-223KE
Yes, nice and cool 😉
My next mod will be the bootstrap caps - will replace the original with ceramic caps - not sure if I shall hear any difference. But I'll report back.
So which parts need to be changed to keep it running cool?
The Chinese boards are built down to a price, meaning very cheap inductors are fitted. These coils are very lossy at HF (read - above 20kHz) so get hot. Being lossy they demand more current from the TPA3116 making that hot too. The solution is relatively simple - install low-loss coils.
I said I don't have a tweezer style iron - not that they don't work.
Ooops! I guess I read your message wrong. Thank you for the clarification. I am going to place an order with Circuit Specialist - pretty inexpensive.
Regards,
De-soldering the inductors was easier than expected - but that was using one of those desoldering hot air stations used for smt parts.
Warm heatsink/inductors means something is wrong - they are dead cool with my amp - no matter if idling or playing music.
I tend to agree with you. I have both the YJ red and the Audiobah green. I was checking for high temperature at the heatsink when I used them and have not encountered any heat issue.
Regards,
The Chinese boards are built down to a price, meaning very cheap inductors are fitted. These coils are very lossy at HF (read - above 20kHz) so get hot. Being lossy they demand more current from the TPA3116 making that hot too. The solution is relatively simple - install low-loss coils.
May I respectfully disagree? Even before I installed CC inductors there was no heat detectable. A friend of mine had indeed the heat problem and - as mentioned above - the gain setting resistors were unequal - thus too much heat - after replacement of the wrong resistor, the board (also YJ Blue) started to operate flawlessly.
May I respectfully disagree?
So you're disagreeing that cheap inductors were fitted to yours and your friends' boards? Or disagreeing that lossy (cheap) inductors result in inductors and chip getting hotter than necessary? Or some other disagreement? 🙂
Even before I installed CC inductors there was no heat detectable. A friend of mine had indeed the heat problem and - as mentioned above - the gain setting resistors were unequal - thus too much heat - after replacement of the wrong resistor, the board (also YJ Blue) started to operate flawlessly.
Is there a mechanism known whereby fitting the wrong resistors leads to hotter inductors and chip? Does it for example lead to a change in the switching frequency?
I replaced the inductors with the ones I mention a few posts ago.
No sound from either channel. Tired of messing with it for now.
I put the Sure TC-2000 + STA508 board back in place. It sounds as fantastic as it always has. Perhaps I'll simply buy another one of those for my bi-amping plans.
No sound from either channel. Tired of messing with it for now.
I put the Sure TC-2000 + STA508 board back in place. It sounds as fantastic as it always has. Perhaps I'll simply buy another one of those for my bi-amping plans.
Last edited:
Incorrect resistors just result in incorrect gain but should not have anything to do with making a chip run hotter or an inductor run hotter. The TPa3116 reads the voltage value set by the resistor voltage divider and sets the gain to one of 6 values I think. It doesn't check gain again until it is power cycled. There is no "wrong" gain because on the EVM they use a pot to set the gain at different settings.
I replaced the inductors with the ones I mention a few posts ago.
No sound from either channel. Tired of messing with it for now.
I put the Sure TC-2000 + STA508 board back in place. It sounds as fantastic as it always has. Perhaps I'll simply buy another one of those for my bi-amping plans.
The Tripath just doesn't measure up to a properly working TPA amp. I usually don't make declarations but I've listened to enough tripath and different TPA chips (3116 s not the only one just newer) feel comfortable saying that Tripaths don't hold a candle. And there are many earlier posts from others testfying as such.
Just get a quality board from hiampmini if you don't want to mod according to the recs that have already been given. They cost what $35? That's nothing considering the performance.
- Home
- Amplifiers
- Class D
- TPA3116D2 Amp