I have also just uploaded a review of this amplifier, slightly different approach. I've used 48 volts from a SMPS. I thought a linear one (unregulated) would suffer quit a bit of voltage drop. I've not uploaded part 2 yet, where I look at the figures. My results were 91 watts x 2 into 8 ohms with 48 volts supply. Just prior to clipping. As for the sound, it does sound different to my favorite amplifier the L12/2. I find all class D amplifiers sound different to class AB, not in a bad way, just different. After 2 weeks of listening, I'm still not sure which I like the most. Most annoying!!
Thanks so much Michael! I did see your video just now - nice work! I'm still working on this series but hope to clean up the noise;) I'll be interested in your part 2. I'll be sure to give you a shout out on my next video;)
@@KissAnalog Thanks, I really appreciate that, I need all the viewers I can get. Kind Regards, Mike. PS I'm listening to the class D right now. Probably the best so far.
Been doing electronic starting in the early 70's as a hobby and in my military and commercial employment. To Date I still do not own a decent scope. Never needed one as I always had access to the best at the times. 64 years old and retired I guess I need to buy one. LOL I never made plans of getting this old.
LOL Time does go fast - and it is hard to believe how old I am already;) I started early too and still having fun! There are some great options for scopes these days - that will blow away what was available just 10 years ago, but at a fraction of the price;)
@@KissAnalog lol yep it do's . Here I Am sick as a dog and still have the need for one . At this stage I really don't think I will buy one. Your age sneaks up on you. I had all these plans and woke up on morning finding I was too old to implement them UN FUN .
If you are aiming to get a low budget start with this “digital amplifier” project, a couple of old Lenovo, HP, or Dell laptop computer power adapters is an option. These power adapters put out between 18 and 20 volt each at 4.5 to 5.5 amp, and are isolated from input ac power and so could be cascaded to 36 to 40 volt. I have been using this type of power adapters for the prior generation of class-d amplifiers using TI TPA3116 D2 and have a satisfying results. It would interesting to compare your class-d amplifier with low budget class-ab, such as from prior generations of Yamaha or Denon home theater receivers. Thank you for this presentation!
Wow, I love having the part number right on the circuit board! Would have been nice to measure the class D switching frequency and how far down the leakage is after the output filters.
@@KissAnalog Please do both, because you're uniquely qualified among RU-vidrs in this arena to do it right and give us accurate empirical analysis of your results.
IMO, linear supplies for audio are kind-of over-represented on YT because of audiophile romanticism. In double-blind style testing, I think a correctly spec'ed and adequately designed switcher would be audibly indistinguishable from a good linear supply for this kind of amp... and most amplifiers that don't rely on things like coloration from insufficient filtering for their character (lookin' at you, vintage tube amps!) 😅 I'm still interested to see the measurable differences, because those should be much easier to find. Of course, if there did end up being a definite non-subjective audible difference, I'd be even more interested!
I like the ranges of this MicSig, but I do like them all: Micsig's new DP10007 Differential Probe & Why we need them ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-QQcZR_KV0P4.html
I don't really hear much about class D used as a power supply for driving amplifiers mainly valve amplifiers, is a better approach! the question? is what sort of power class D amplifier would you need to drive say a 20 watt valve amplifier? I'm assuming the more powerful is the cleaner will be so we looking at 2000 watts or greater?
@@KissAnalog no I mean class D as the power supply for the amplifier, it being a clean form of a smps! running over 120kHz this eliminates low frequency what is ideal for valve amplifiers, where it suffers distortion at it's greatest in the audio band especially mains 50/60 Hz and 100/120Hz.
That's a great point! I actually think a good stiff SMPS might be just the thing for any amp that takes a lot of power to run. As you say - the switching noise is well above the hearing range and compared to the linear which has the diode commutation noise - as well as transformer humm.
awesome! I am eyeing TPA325x series for my class-d amp adventure and this video get me closer to like the chipset. Is there an access to power down the module (M1 or M2), what's the deep sleep current consumption of the module? video 12:28 min, I am guessing that those LEDs are the FAULT and CLIP_OTW pins, if both are logic 1 is in normal condition : Junction temperature lower than 125°C and no Overtemperature, undervoltage or junction temperature higher than 125 degrees (, these 2 pins must have been driving the LEDs). Having said that, if any one of the LEDs goes off, it means an error occurs.
Thanks John for the great feedback! Like the CD, as it most likely has a digital control that filters this. There is an analog feedback path, but the beauty of having digital is that it can operate around the analog setting boundaries, like temp, voltage, current, and even frequency;) The switcher is much faster (500kHz) so there's plenty of digital bandwidth for Nyquist (if that was the question).
Sweeet! I still have some TPA3255 ICs left over from an old project and picked a couple more up recently. They're really excellent. It looks like your board used about 20x more capacitance across the supply than what even TI recommends, which is nice to have on high power class D, IMO. With some external RC feedback added, you can get even better distortion and noise performance from these devices, as well as making the amplifier specs somewhat less dependent on speaker impedance. Some potential downsides are an inevitable reduction in gain, the active pop & click suppression gets defeated, so you're likely to get some pops on power-up (I don't remember if delayed RESET sequencing helped), and you need to test it thoroughly to make sure it's stable across the operation spectrum. I think TI recommended that most of us mortals use the 'step-response - estimate phase margin from the overshoot - check for ringing' technique. I really liked the idea of lower noise and distortion, but with the pops and potential stability issues I didn't have the time or patience to test to my standards, so I DNP'd the post-filter feedback RCs. The specs are already so good that nobody's likely to ever hear a difference, and I think it ends up being more spec bragging rights than anything. I'll probably give it a better shot on my next project. Also, to keep them from getting hot I used two of them, one for each speaker in parallel BTL. Not the economical route, but for one or two off type projects .. you know! :)
LOL - thanks for the awesome feedback!! You hit a bunch of great points and you have experience with this part - so I'm very interested in you ideas and comments. I will do some follow ups as I think this will make a nice project;) Let me know where I'm going right and wrong;)
@@KissAnalog That's funny, I was hoping you'd make a more detailed video about the part so I could find out if *I* did anything wrong in my class-d designs that used it. 😂
FYI this board is an exact copy of the TI Evaluation board. If you want to drive high power into 4 Ohm loads, I would suggest using a really good non-electrically conductive heat sink compound like arctic silver ceramique. There is a high power density through the small heat sink pad of the IC.
Sorry I misspoke- this is not an exact copy of the TI EVM, since it does not have the different modes like 4X SE that the TI EVM has. But there is a lot in common, like the switcher for 12V gate supply and the opamp setup for inputs. I have one of the EVM's as my workshop stereo, and it rocks. Thanks for the detailed review!
Thanks Frammis! I did see the similarity to the TI board. I was considering making a version using the best options available. I'd like to look closer at the LC components and the OpAmp;) I like the idea of the arctic silver ceramic!
I've tested more thermal compounds than I can count in different applications, and Ceramique (now Ceramique 2) always come out on top. We buy it in bulk (rather than in tiny tubes) directly from Arctic Silver.
Learned about power supply from your channel for my amp using lm3886 few years ago. Now this might be my new project. What's your opinion between this and lm3886?.
Great video Eddie! I didn't realize that micsig is 4 chan. That's cool. I'm looking for a scope with 4 channels that won't break the bank. What's your opinion on it other than using it to protect your gee dub ah ya from harm. Hahaha I think we have to have a funeral for that amplifier at over 100. Active cooling is definitely the way to go. I wonder how far something like that could be pushed with water cooling.
Thanks Joey! Great question regarding the MicSig. This is one of my favorite scopes. I like the larger screen - and the touch display is actually nice to use. I have a screen protector on it which I don't know if I like. It is less reflecitve for videos, but it does add a layer of plastic. Here's a link to a review - and please use the links;) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-NtEt8Zc8oVA.html I think to get the full power - it would take a larger heatsink and/or active cooling;) I'll have to test this with a quite fan first;)
hello, I don't understand your video, I only know Spanish, I'm using the translator, I wanted to know what the real power is without distortion and how much impedance was measured
I have some of those amplifier boards. There is a 12v regulator - LM2575S-12 - on the board, I replaced it with an HV version because the one that is fitted is not good for 48v supplies.
@@MichaelBeeny Lucky. Component overlay stated LM2575S-12 too, so they fitted the correctly specified part, but the seller suggested up to 50v supply voltage.
The TPA amps seem to love to have perfectly balanced input signals. When using an OP1632 circuit on the input to the TPA chips in BTL and PBTL mode the distortion is lower and the overcurrent fault doesn't trigger so easily. Not sure what it is but the TPA3116 and TPA3255 seem to want a perfectly balanced input in BTL/PBTL modes. At least in simulation they do, need to get some boards made...
My board 4ohm speaker impedance 200w load power clip then amplifier off...what is the problem... Plz ans totally output this board 4ohm impedance 48v(12.5A) power supply
In the Texas Instruments datasheeet there is a diagram for a simple differential R/C filter that you need to make THD measurements on class D amplifiers.
Thanks John! I think I need to do a video just on this subject. If I used this filter without some discussion - I think it would just bring up too many questions of its validity. What are your thoughts?
@@KissAnalog When you 'scoped the outputs on the amplifier, you saw HF switching noise superimposed on the output waveform. I built a filter according to TI's suggestion, (100R and 4n7 from what I remember), in a small diecast box. The output waveform post filter had no visible switching noise. Really, THD measurements without a filter make no sense, and this filter is far cheaper than, say, Audio Precision's. They use a filter too, for THD measurements on Class D amplifiers, though theirs is multi-section, and very flat within the passband.
Not sure if you have registered with Audio Precision's website, but if you do so you can download some excellent white papers on this subject. They make some specialized filters for class-D (AUX-0025) that TI uses for measuring all of its class-D products.
Thanks for the great question! There’s 2 front speakers, 2 surround, 1 center, and a subwoofer speaker, so that’s 6 speakers. So it depends on the type of amps you have.
@@KissAnalog hi things for your reply I have idea TPA3255 stereo for each channel configuration and mono for drive Subwoofer. It's total 5 ic which is 4 ic stereo and 1 ic for mono i dont know how match amps it's handle.
As a beginner, I am wanting to learn more about op amps, and what they can do. I am considering ordering a 741 build kit from Evil Mad Scientist to better understand how they work. Good idea, or not so much. I know this question doesn’t exactly pertain to this video. 🙃
I’ve just started an OpAmp series. Have you seen my inverting OpAmp? I then used MicroCap to simulate one. This is a way to learn like experimenting in the lab. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-2RYajCwoPjU.html
How about making both a linear (transformer) and a digital power supply that can supply the power for 120Wx2 into 8Ω and 260Wx2 into 4Ω (at these output, THD is ~%). Either will need to supply 48V DC and 16 amps of current, i.e. almost 800W of power! Actually smaller power, say 400 to 600W may be more than enough. Also would like to see if you can add a small fan which will only activate when the heat sink reaches certain temperature.
Ooooh, a Kiss Analog bed Comforter! That could be just the boost I need for my coefficient of courtship! "What's a Kiss Analog?" ... I'm so glad you asked, Nikki! ~*Lays down and seductively pulls up RU-vid on the bedroom HDTV* 😎
Hi! @14.54 be careful because the switching voltage regulator mounted in these cheap chinese boards is not the HV version, thus only can withstand 40V dc max! Also, in my board (I have one identical to yours) the opamps are not the good NE5532P raccommended by TI, but crappy TL082 which sound much worse. Not sure if the output chokes are good enough too, or their core saturates easily. Bottom line, mine sounds worse than an even cheaper TPA3116D, to be honest.
Thanks for that great feedback! My board actually came with the HV (high voltage) version of the 12V switcher;) The OpAmp is from ST Micro and has a designator GZ-316, but I don' t know which OpAmp that is.
Oke Eddy go for a switching power supply, it has a 12v switcher on board so i think a linear power supply will not be beneficial, its a D amp so do not put money in it, safe your money and buy a nice case for that Aleph 5, do some test on that 10pf on that board, and they say that the 10pf needs to be a silver-mica cap, so what do you think about that, you can go for more power to the irfp250n mosfet, inside that is like 2 times a irfp240, you also need to change the source resistors for more power, so go lower, experiment with that, more interesting video, also i did buy a focusrite scarlett 2i2 for 150 euro, so i do some measurements with that with the software REW, but i still need to build a dummyload at 8 and 4 ohm, and a voltage divider for safety for the input of the focusrite scarlett 2i2, or i will fry that input ;-), it will be a cheap way to go and mesure you amps
Thanks Jos Haas! I agree - a SMPS is the right choice for a Class D:) I'd like to find one with a PFC front end to keep the PF close to unity and for the lower noise. I'll have to get back to that Aleph 5. I'll look into the FET and the 10pF cap. Thanks for the feedback!
@@KissAnalog I use these power supplies all the time they're cheap and they work cleanly. I'm comparing it to a toroidal linear. Of course you have to modify the AC input as they are 220. PFC is a nice to have, but not a must have for me. I'm not sure what the price would be for that added feature.
@@KissAnalog The whole circuit is 100 percent analog. It's not even a litttle digital. The Pulse Width modulation principle used in class D is also commonly used in digital audio. This pulse width modulztion cirduit is however purely analog . CLASS D IS NOT DIGITAL. AT ALL. Not even a little bit. Really. It is called D because it came after C and before E. Class H is a special kind of class D, recently finding it's way into hifi. Why don't we call it Higital amplification?
There are class D amplifier chips with a digital input and a DAC on board, connected to the source via the i2s bus; such chips can already be classified as digital.
In fact, considering the large cost of the TPA chip, likley because of the large die required even for the crappy 85mO mosfets, a controller with extgernal mosfets would likley be a great deal by comparison. I believe IR/Infinion has such controllers. No idea on price/part number, but if decent then this might be way better.
Thanks for pointing this out. The integrated packages have a lot going on so there are a lot of trade offs. I have some other boards with different parts, so I'll have to get to reviewing them. Let me know if there's one you would like to see.