Regarding RF noise from amps (or any electronics for that matter): wouldn't the amount of power going into the emitter make a difference? Be it an air-core inductor or an actual explicit broadcast antenna, it seems like how much RF you generate must be proportional to power. What's the power "threshold" for when the RF actually becomes a problem?
The tpa3116 is a very efficient chip, so there's not a lot of "waste power" left over to use for RF pollution, right?
For example, in my setup (nearfield, moderate listening levels), I estimate I'm only putting about one watt (usually less) into each speaker. So that's two watts total output from the amp, and let's say it's only 50% efficient (overly conservative estimate), so that means two watts are going to "waste". If all of that is converted to RF noise, what does that mean in practical terms? What's the worst-case scenario for RF pollution that one can do with two watts of power?
Let's say someone's pushing a constant 20 watts each into two speakers, so 40 watts total output from the amp. At this level the tpa3116 should be operating closer to 90% efficiency, but let's say 80% for the sake of argument. That implies 40/.8 = 50 watts input power, and 50-40 = 10 watts of waste. How much "broadcast power" do you get with 10 watts?
These are honest questions, by the way, not rhetorical or subtle arguments.
***
10 watts of broad spectrum RF energy that is transmitted outside the unit would wipe out all kinds of communications within at least hundreds of yards - cell phones, police, fire, etc. Even two watts would create big problems. Nothing like that would comply with FCC emission standards.
There's no way this kind of energy is getting outside the SA-60's case.
This amp can output a maximum of 4A per channel. That's 2A per inductor. Most inductors used for the TPA3116 handles at least twice that if not 3 times.
Saturation is not an issue.
Air core inductors however typically have higher ESR. Often twice or 3 times as much. That would limit output and affect audio quality.
***
Everything in electronics design (and life) is about trade-offs, pluses and minuses.
So, the first question is whether the air core inductors in the SA-60 create radiated RF noise problems - my tests say it does not.
As to ESR: core inductors have several kinds of losses that contribute to ESR that do not affect air core inductors: hysteresis, eddy current, proximity and skin effect. Because of all these factors, core inductors often have higher total ESR than air inductors.
As I see it, using air inductors as in the SA-60 is a reasonable approach taking into account the pluses and minuses. The bottom line is that it works great in my system without any noise issues. I'll be running more tests and trying it out in other setups, and if there are any problems I'll report back. Also, I plan to try out other 3116 boards which include core inductors and will be able to compare them. By the way, I've have Class D amps here using solid core inductors that have radiated noise problems - there are lots of design issues that affect the noise issue.
... there are lots of design issues that affect the noise issue.
I know. That's why my comments are generalized. You can design class D amps with air core inductors that don't have EMI issues and that do sound good. They would be rare. On the other hand, you also can design class D amps with toroid or shielded core inductors that do have EMI issues and that do not sound good (or at least as good as they could).
Since this is DIYaduio the general advice is to not use air core inductors in class D amp designs but use toroids or shielded core inductor specifically designed for this use.
I tried doing some trace finding for Sure board. I know picture is a mess, but it gives some idea where everything is going.
I find parts marked as TVS1 and TVS2 very interesting, because the way they connect to inputs. On the part is reads 5A6Σ (or 546Σ). Might this part been used by Sure in hopes it would reduce turn off pop? I didn't find any datasheet for the part, but using multimeter I can find two diodes with forward voltage of 1,174V.
I guess Q1 (near power jack) is just used for LED, Q2 is for mute function when MUTE pin is connected to GND (absolutely not sure about this).
All traces might not be correct and I could be wrong about the part, but what you guys think?
Better pictures for the board are here.
I find parts marked as TVS1 and TVS2 very interesting, because the way they connect to inputs. On the part is reads 5A6Σ (or 546Σ). Might this part been used by Sure in hopes it would reduce turn off pop? I didn't find any datasheet for the part, but using multimeter I can find two diodes with forward voltage of 1,174V.
I guess Q1 (near power jack) is just used for LED, Q2 is for mute function when MUTE pin is connected to GND (absolutely not sure about this).
All traces might not be correct and I could be wrong about the part, but what you guys think?
Better pictures for the board are here.
Attachments
There is a listing on shenzhenaudio, and also on Amazon (with shenzhenaudio as the seller):
http://www.shenzhenaudio.com/smsl-sa-60-60wpc-tpa3116-class-d-digital-amplifier-hifi-air-core-inductance-desktop-amplifier.html
http://www.amazon.com/SMSL-SA-60-Digital-Amplifier-silver/dp/B00PA3QIOE/
SMSL SA-60 60WPC TPA3116 Class D Digital Amplifier HiFi Air Core Inductance Desktop Amplifier
Apologies for the bad links in my post -- here they are fixed.
SMSL SA-60 60WPC TPA3116 Class D Digital Amplifier HiFi Air Core Inductance Desktop Amplifier - Professional Audio Store - Shenzhenaudio.com
Amazon.com: SMSL SA-60 60WPC Class D Digital Amplifier silver: Electronics
***
I just purchased the SMSL SA-60 amp. I'm an extra class amateur radio operator, have been for the past 46 years, and am active on the HF bands (1.8 MHz through 30 MHz). All kinds of electronics have been thrown out of my home (figuratively, if not literally) because they caused a miniscule increase in RF noise level (which can make the difference between making a DX (distant) contact and not). Light bulbs, computers, toys, coffee makers, remote controls - out the window.
With the SA-60 turned on vs. turned off, I can hear no change in noise level on the HF bands, with the HF transceiver plugged into the same AC line at the amp, sitting 10 feet away from the amp, and using a dipole antenna that runs twenty feet above the amp's location.
I hear no noise at all introduced into the audio system in which the amp is being used - and I'm acutely sensitive to audio noise. For example, I recently found that my HTPC was introducing a very low level audio noise because one screw was not fully tightened on the outside of the case.
The fully enclosed metal case of the SA-60 is doing a good job of blocking any radiated noise and noise is not getting out through the AC line or the line input or speaker lines.
At the same time, the SA-60 sounds wonderful. It's just a couple of hours old and seems to be improving with time. I have no other 3116 amps at this point to compare it to, but have LOTS of amps, both solid state and tube (high, mid and low end) to compare it to, and it's an excellent amp (within its power output limits, which are after all about 25 watts at reasonable THD levels). I'll be experimenting with better power supplies (inlcuding high current 12 volt battery) to see what difference that makes.
If this amp had noise problems you would be advised to duck while driving past my home, because it would be flying at high speed after having been thrown out the window.
Les
Thanks for that feedback. I might have to buy this amp🙂 looks like a winner.
I do have a couple of issues with the SA-60 amp - but not about its sound quality or freedom from radiated or internal noise (in both respects I think it's excellent). My complaints are about functionality issues.
First, the speaker binding posts on the back are reduced size. This was a necessity because of the small size of the case. However, it's dangerously easy to short out the speaker connectors with normal size spades or bare wire. Insulated banana connectors are probably the safest connection to use - though they are not my favorite. The electronics could be put in a larger custom cabinet with standard size binding posts - but that defeats the point of buying a finished amp (and the finishing on this amp is quite nice).
Second, the amp is really set up to be used as an integrated amp rather than a standalone basic amp. It has two inputs, front and back, which are switchable (a nice feature for an integrated amp). However, I want to use it as a basic amp with my HTPC analog output acting as preamp. I also want to turn it off when not in use (I'm pretty careful about minimizing unnecessary electric power usage - though I realize the 3116 draws only about a half watt when turned on but not in use, I'd rather save that power).
When the amp is turned off, and then turned back on, the volume is automatically reduced to a very low level (a good feature to prevent accidentally blasting away one's speakers or ears). However, that low gain setting in insufficient for use as a basic amp. So I need to manually turn up the volume all the way every time I turn on the amp to use it. That's a little bit of a pain - but it gets much worse if I want to use 7 of the amps (bi-amping each speaker) in a home theater system (it would require running around the room and manually turning up 7 amps each time). I emailed the distributor, Shenzhen Audio, asking if this feature could be defeated, so the amp remains at the last set level, and they answered no (at least not without modifying the unit, if that is feasible).
So this unit really works best as a 2 channel integrated amp. With banana plugs or VERY careful use of bare wire.
First, the speaker binding posts on the back are reduced size. This was a necessity because of the small size of the case. However, it's dangerously easy to short out the speaker connectors with normal size spades or bare wire. Insulated banana connectors are probably the safest connection to use - though they are not my favorite. The electronics could be put in a larger custom cabinet with standard size binding posts - but that defeats the point of buying a finished amp (and the finishing on this amp is quite nice).
Second, the amp is really set up to be used as an integrated amp rather than a standalone basic amp. It has two inputs, front and back, which are switchable (a nice feature for an integrated amp). However, I want to use it as a basic amp with my HTPC analog output acting as preamp. I also want to turn it off when not in use (I'm pretty careful about minimizing unnecessary electric power usage - though I realize the 3116 draws only about a half watt when turned on but not in use, I'd rather save that power).
When the amp is turned off, and then turned back on, the volume is automatically reduced to a very low level (a good feature to prevent accidentally blasting away one's speakers or ears). However, that low gain setting in insufficient for use as a basic amp. So I need to manually turn up the volume all the way every time I turn on the amp to use it. That's a little bit of a pain - but it gets much worse if I want to use 7 of the amps (bi-amping each speaker) in a home theater system (it would require running around the room and manually turning up 7 amps each time). I emailed the distributor, Shenzhen Audio, asking if this feature could be defeated, so the amp remains at the last set level, and they answered no (at least not without modifying the unit, if that is feasible).
So this unit really works best as a 2 channel integrated amp. With banana plugs or VERY careful use of bare wire.
Les, thank you very much for your ongoing reviews.
Quick question: what's the best scenario to use with those T-amps? Set full volume on pre-amp (say iPhone, computer or actual pre-amp) and use the volume rotary on the T-amp for actual output, or set the t-amp full volume and use the pre-amp (again, iPhone, computer, actual pre-amp) to manually set the listening volume?
Quick question: what's the best scenario to use with those T-amps? Set full volume on pre-amp (say iPhone, computer or actual pre-amp) and use the volume rotary on the T-amp for actual output, or set the t-amp full volume and use the pre-amp (again, iPhone, computer, actual pre-amp) to manually set the listening volume?
Les I also want to thank you very much. I was just about to order one of these amps when I read your latest post. I also planned to use them in a bi-amped HT system driven from our HTPC, although I only need 6 channels total.
What a bloody shame on the volume issue! I guess their market is much more for standalone use than in a multi-amped system so I can understand their adding that 'Feature'.
I thought I had finally found THE amp that would work well and at it's price is a good value for fully assembled in a case with supply. So back to the drawing board!
I am looking for high SQ, so without doing lots of mods that my 73 year old hands and eyes are no longer good at I'm not sure what the best option now is. Some have posted that the green Audiobah board is one of the best unmodded. There are also the HiAmplifier boards Ultimate V1 TPA3116 Class D Stereo Power Amplifier Board w Bootstrap Caps | eBay They are certainly a lot pricier than any of the other bare boards and there has been little talk about them on any of the 3116 forums.
My preference would be a packaged amp such as the SMSL60 but I could install in a case if needed.
Any suggestions would certainly be appreciated.
Thanks,
Rod
What a bloody shame on the volume issue! I guess their market is much more for standalone use than in a multi-amped system so I can understand their adding that 'Feature'.
I thought I had finally found THE amp that would work well and at it's price is a good value for fully assembled in a case with supply. So back to the drawing board!
I am looking for high SQ, so without doing lots of mods that my 73 year old hands and eyes are no longer good at I'm not sure what the best option now is. Some have posted that the green Audiobah board is one of the best unmodded. There are also the HiAmplifier boards Ultimate V1 TPA3116 Class D Stereo Power Amplifier Board w Bootstrap Caps | eBay They are certainly a lot pricier than any of the other bare boards and there has been little talk about them on any of the 3116 forums.
My preference would be a packaged amp such as the SMSL60 but I could install in a case if needed.
Any suggestions would certainly be appreciated.
Thanks,
Rod
Also, the SA-60 carries both FCC and CE certifications, which include RF noise emission testing and limits.
I don't know about FCC certification on this amp but it is not CE certified. CE logo on it is China export logo not Conformité Européenne.
I tried doing some trace finding for Sure board. I know picture is a mess, but it gives some idea where everything is going.
I find parts marked as TVS1 and TVS2 very interesting, because the way they connect to inputs. On the part is reads 5A6Σ (or 546Σ). Might this part been used by Sure in hopes it would reduce turn off pop? I didn't find any datasheet for the part, but using multimeter I can find two diodes with forward voltage of 1,174V.
I guess Q1 (near power jack) is just used for LED, Q2 is for mute function when MUTE pin is connected to GND (absolutely not sure about this).
All traces might not be correct and I could be wrong about the part, but what you guys think?
Better pictures for the board are here.
5A6 diodes? Maybe inputcapacitors for DCblocking aren't necessary? I mean these seem to block DC towards source too, not from source?
My preference would be a packaged amp such as the SMSL60 but I could install in a case if needed.
Any suggestions would certainly be appreciated.
Thanks,
Rod
Rod, you might check the HIFI 2.0 amps - I've got a couple, and I must say that they sound good - I didn't even feel the need to try any mods on them:
New Silver HiFi 2 0 Stereo Output Digital Power Amplifier TPA3116 50WX2 | eBay
You get a cased amp, ready to use - and, like I said, the sound quality is really OK for my ears 🙂
Even if you don't like it, at that price you haven't lost much...
I don't know about FCC certification on this amp but it is not CE certified. CE logo on it is China export logo not Conformité Européenne.
It's not FCC qualified either. The FCC logo is a Chinese lookalike.
Or it's just a fake. The FCC logo holds no legal obligations. Anyone can smack one on to any device they like without any consequences. You cannot however put a fake FCC ID# on your device. The FCC ID# is the only thing that holds any legal obligation.
Last edited:
Rod, you might check the HIFI 2.0 amps - I've got a couple, and I must say that they sound good - I didn't even feel the need to try any mods on them:
New Silver HiFi 2 0 Stereo Output Digital Power Amplifier TPA3116 50WX2 | eBay
You get a cased amp, ready to use - and, like I said, the sound quality is really OK for my ears 🙂
Even if you don't like it, at that price you haven't lost much...
the price is really good!
what are the electros between the bigger film capacitors and the chip/heatsink for?
the price is really good!
what are the electros between the bigger film capacitors and the chip/heatsink for?
Local decoupling I'd imagine.
Hi everyone, I have recently bought an tpa3116 red board, and a one of 1uF(according to pcb) capacitors, near audio input is damaged, pins are broken. I have replaced it with same value trough-hole capacitor. Now when I turn it on, sound comes and goes away,it works for second and then it turns off for second. Is that because of capacitor or something else? Also, chip is running hot, i cant hold finger for more than few second on it.
I have resoldered every capacitor and resistor on board and problem is still there. I even tried few capacitors instead one that was broken(all 1uF, 50v).
i think what you have is a broken smd soldering pad.
Local decoupling I'd imagine.
so they tried to get that as close as possible towards the 'vcc-pins'?
too bad that they didn't place them left and right of the chip.
instead they could put the pins of the signal input right in front the input pins of the chip.
I made the following modifications to the standard YJ Blue-Black board:
Snubber Capacitors: SMD 0603 Multilayer Ceramic 330pF 50v C0G 5%
Snubber Resistors: SMD 0805 10ohm 0.25watt 1%
Bootstrap Capacitors: Multilayer Ceramic 0.22uF 25v X5R 10%
LC Filter Inductors: Coilcraft MSS1210-153MEB 15uH
LC Filter Capacitors: Multilayer Ceramic 0.68uF 50v X7R 10%
Output Filter: Multilayer Ceramic 0.01uF 100v X7R 10%
Input Capacitors: Elna SILMIC II 10uF 35v 20%
Power Supply: Panasonic 25SEPF330M 330uF 25v
I replaced all bootstrap, filter and snubber capacitors with multilayer ceramic capacitors as recommended in the datasheet and Evaluation Board Guide.
The snubber circuits were added on the bottom of the board using SMD components. By mistake I ordered SMD 0603 size capacitors instead of the slightly bigger SMD 0805 size, so soldering was a challenge.
Given the physical layout of the tracks and ground connections on the YJ Blue-Black board, the 1nF capacitors of the output EMI C-RC Snubber is directly in parallel with then 0.68uF capacitors of the LC filter. The 1nF (0.001uF) capacitor is completely masked by the much larger value 0.68uF capacitor. I have therefore omitted the four 1nF snubber capacitors and simply mounted the 0.68uF filter capacitors in the position of the 1nF capacitors. This leaves more space for upgraded inductors, if required.
The YJ Blue Black Board has a gain setting of 26db with an input impedance of 30kΩ. A -3db bass roll-off at 20Hz, requires a 3.3uF capacitor. I already had the SILMIC II 10uF capacitors on hand and fitted the 10uF and it sounds very good.
The inductors were changed to 15uH, which is a better mach for the 6Ω impedance of my speakers. The inductors also have the same footprint as the original inductors.
Power supply is a laptop SMPS brick delivering 19v at 4A. An additional 4700uF 35v decoupling capacitor is connected directly to the power supply terminals of the YJ board with the power supply wires.
The result is very satisfying and sounds very good, comparable to my NAD system.
i would have tried to place the caps you use for signal input to place right above the other pin, which is closer to the chip.
what are the electros between the bigger film capacitors and the chip/heatsink for?
I can tell what they are:
Elite 2.2uF 50V NP
so they tried to get that as close as possible towards the 'vcc-pins'?
too bad that they didn't place them left and right of the chip.
instead they could put the pins of the signal input right in front the input pins of the chip.
If the input signal has already passed the input capacitors it's not that subjective to noise being differential. Also remember that the GVDD pin annoyingly placed in the center of the chip also has to be decoupled by at least 1µF. 2.2µF could also be used if it means you can buy bulk and thereby save money.
Last edited:
- Home
- Amplifiers
- Class D
- TPA3116D2 Amp