GaN Class D Review Gain/Phase pt1 with EPC GaN FETs
Вставка
- Опубліковано 5 жов 2024
- This video I review a GaN Class D Amplifier. It is a part 1 of the review of a Class D Amplifier using a GaN FET. This is what EPC calls a eGaN FET for enhancement mode FET. I'll show the schematic and Bode Plot of the Gain and Phase over the audio frequency band. #ClassDAmplifier #GaN #eGan #GaNFET #ClassDAmplifier
Future of Power Electronics GaN MOSFET • Future of Power Electr...
Please use my Amazon links to help support this channel. Below is an example Class D amplifier that you can start with then find the one you like. There are non right now with GaN, so I'll have to design one on this channel;)
amzn.to/36QhsNK
You can support this channel with this link to Patreon
Patreon.com/KissAnalog
That is very good phase response for class D. Nice!
Hi Taylor! I totally agree with you. Great to hear from you.
It seems a very good tutorial how to use this good scope. The Class-D is also not bad. Thanks for this great video!
Cheers! Stay safe!
I really appreciate you! Thanks for this feedback!
Hi Eddie, excellent video, thanks for the information.
Thank you for the nice feedback!!
Again very interesting and fun to watch! And just great that you won plus you were allowed to loan that amplifier:-)
I have a few comments, you do tell about some interesting details on the board but it is hard to follow, may I suggest that you start as you always do but then lay the board on the bench and fill the frame with the board, I think your point would be clearer in that way.
Later do you have a subscribe to "our" channel, I think it may sound a bit like you and your team are asking. :-)
I like the almost invisible transitions with a hint of special effect and the sign where you tell what to expect in the next video, great idea!!
All in all, I think you do make your videos a lot more smooth, natural flowing and relaxed, then you did, just half a year ago. Nice to see how fast you are getting comfortable and fluend in being an UA-camr:-)
Great feedback - thank you! I like your ideas and will try to incorporate them.. Thanks for your support!!
very informative, excellent.
Thank you - I appreciate you!
The roll-off is due to the limit of the output inductor -- a material designed to have the highest power density/efficiency at 100kHz. That will be what the HF is running at, the triangle wave that is combined with the sine-wave to create the pulse width switching and nearly 99% efficiency. The series inductor passes the lower frequencies, and the output capacitors integrate the output back into a sine-wave or musical tone and remove RF noise.
Thank you for this great feedback Dunc - much appreciated!
Again, another great video!
Thank you!
Do you have more on how and when to use differential probes? Great video btw!
I did do a comparison between diff probes. See below. But I think I need to do one on the when and how... Thank you!! ua-cam.com/video/892RcsIUAtU/v-deo.html
I am interested in the input part. The input is an Analog signal, which is generated by Gen1, and before driving the GaN devices, somewhere it should be turned to Digital signal or PWM signal. Where does this happen? Another interesting part is, you set up the load impedance to 50 Ohm, that is for input signal, right? Shouldn't it be set up to High Z load so that most of the input signal voltage dropped at the input of the amplifier? I mean most of the signal is delivered to the amplifier input. Thanks. Great video.
Thank you and great insight! I am working on reverse engineering this board, I asked for a schematic but had trouble. Maybe I'll get it... I'd like to do the Bode plot from input to where the DSP takes it in, and then do one from the output of the DSP to the output of the amp. I'll see if I can do it.
The low z setting on the generator allows it to put out the most signal. The high Z on the generator expects to see a high Z output. I actually tried it both ways but didn't see any difference;)
Great input and questions! Thank you!
Do you have a video discussing when and how to use differential probes. Kind of a newbie question but thats me.
That's a great idea Brad! I actually have a new set of probes coming so I'll do that. I also have a variety of low cost probes that I want to look at the cost vs quality. I'm happy with what I've seen.
@@KissAnalog Cool. Looking forward to learning about differential probes and some buying possibilities. Look at what you and learnelectronics caused me to set up. I like reproducing your experiments.
www.dropbox.com/s/e82lc8l4zv70mgh/Bench.jpg?dl=0
Brad
Wow! Nice bench! Now the fun begins Brad;)
I have the same AB amplifier you discussed in this video. Where can I find the heat sink you are using?
Hey Brad, I had this one from a tear down that I did awhile back. I'll show a video on where you can get a heatsink! Thanks!!
Thank You!!🙏
I appreciate you!
"...and here are the GaN FETs..." Me: Whaaat.?!? What sorcery is this!? Im going to need to douse holy flux all over to counteract these dark arts! 😄
LOL - thanks for the feedback!
Hey Eddie, really enjoying your channel. You're a wealth of knowledge. I'm learning so much, thank you for sharing/teaching your years of hard work.
I have a question about the GaN Fet. How does it pass so much current without generating the heat? (related to Conor's comment). I tried to figure this out by looking at the datasheet but I'm just too new to this to understand it from reading the datasheet. Maybe you could explain this a little in an upcoming video or here in the comments, possibly some measurements around this or even comparisons to a traditional Fet?
Oh derp, I didn't see your video GaN vs Mosfet till just now. The datasheet makes a lot more sense to me now and nice job describing where the power losses are. I'd still be interested in measuring some of these differences though. Thanks much Eddie
Thank you Jared - I really appreciate you! I think I do need to do a video breaking down a GaN FET and a MOSFET and comparing them. It is very interesting how GaNs seem to be so good at their small size. Great question!
Hello ! my dear friend, how are you, I watched a video on your channel for a while and since that day I haven't stopped watching your videos are fantastic congratulations, but I'm going to ask you a question what do you think about the new GaN technologies
Thank you! I appreciate you! I think GaN will be prevalent in the near future. I did a number of videos around GaN, but this is the first: ua-cam.com/video/Ej_P4_2xS4c/v-deo.html
@@KissAnalog I believe that GaN will be applied in power systems such as switching sources and class D audio systems and embedded systems, I can even say that in the construction of IC's chips
Do these things work on magic smoke or black magic?
LOL They do seem to be magical. I think I need to do a video to talk about how these are made vs other MOSFETS. Thanks for the feedback!
@@KissAnalog That'd be cool. Guys at ZeptoBars would also accept some GaN product, so we might get a die shot of them.
zeptobars.com/en/support-us
Found it on the "send us some interesting chips" parts because I was searching for a GaN die shot on their website.
Thank you! This is very interesting:)
Minute 12 in the video, I just realize: the guys placed a RED banana for MINUS... nice... no comment...
LOL! That's awesome - thanks for that!
Just another detail: the GW INSTEK scope, minute 18: why channel colors are shifted comparing to Agilent or LeCroy? Should that be something unusual or the design guys were just at christmas party when they did this? Should not be a standardisation of the things in the engineering world, anyway?
That's a great observation! You are right, but I think there isn't much cooperation or standardization with these folks;)
WOULD YOU PUT PRESENTATION ON PHASE MARGIN, LOOP GAIN,STABLE,AND UNSTABLE OPERATIONAL AMPLIFIERS?
Good to hear from you David! Yes, this is something that I want to do. It is a very important subject.
Why not just use an appropriate sized single input power supply?
The dual isolated and symetrical power supply complicates the overal design. I don't get it. I think some type of top secret multi gate controlled graphene doped nano hyper switching input will replace that power supply design.... Probably a crutch by design at this point.
Great question! Funny answer;) I think one reason is that the input signal comes in referenced to the ground so to swing the output Plus and Minus from ground, a negative reference helps. But it is just trade offs vs doing it another way. The split supply isn't really complicated, once you decide to use a transformer to isolate your power;)
It will be interesting to see how it acts under full load temperature-wise. No heatsinking specced to 150W impresses me! Could these (GaNs, maybe not this specific instance) be used, for example, as part of the loading circuit in a DC load? I'll have to hit a few data sheets.
Did you see pt 2 where I show the temperature? Very impressive! Thanks for the great question Andrew!
I did, yes. Was that at full load and for how long? When I was testing a supply I had it running at it’s nominal 15V and 3A for an hour sampling temperature rises to see what it would do. I was wondering if you had tried the same with these GaNs to see if the ‘no heat sinking’ was more than just a marketing statement but held up :). I know you have to send the board back so can’t destroy it so you would need to keep an eye on it but that’s ok. Aborting a test is as informative as making it: I’d be rooting for the latter though, it would be impressive to do that!
Do you sell/make any amps? I want you to build my next amp. :)
Hi Valentin, I am working on a design that I will do videos on covering schematic capture and board layout. I have a Class A kit that I bought that I plan to do a video on very soon. In the future I’ll have kits and maybe just plans and boards too. Thank you for asking - you’re the first;)
Hi, which brands do you think are the best for Class D Amplifiers? It's for a project. Thanks!
Thanks for the question! Are you referring to the GaN FET or the Class D controller?
The problem is that almost all cheap chip class-d amplifiers are built the same way! They also all use the same output filter design. None is better than another cheap class-d.
If you are going to have a class-d that can accept different loudspeaker loads then hypex ucd180 or ncore class-d is MUCH better as they are built in a slightly different way
Thanks for your feedback! I agree - they do follow a very similar design. Some of the newer chips have some interesting features. One thing to look for is the quality and size of the output filter parts - and that is where a big difference can be found.
It really depends - I had some Hypex NCore NC500 modules, they sounded and measured like the worst crap ever.
@@qddk9545 I have only used the UCD180 HG. From Hypex. frequency response independent of load.
You can NOT measure a cheap class-d amplifier ONLY in 4 ohms or 8 ohms! Speaker varies in ohms. and a regular cheap class-d does not behave the same in different OHMs.
I have compared the Hypex UCD18N HG with many cheap class-d amplifier cards and several older mid-range class-ab amplifiers and found that the hypex ucd180 hg has incredibly good sound quality!
@@KissAnalog It doesn't help how many functions an amplifier ic. has!
A class-d design with a series-connected coil without feadback behaves the same as a series-connected coil on a loudspeaker. The frequency will not be flat with different OHM load! Have also seen that between 20hz and 20 khz there can be 6db. difference!
This is on some of the most common class-d amplifiers you can buy on ebay and aliexpress.
I have only used the UCD180 HG from Hypex.
You can NOT measure a cheap class-d amplifier ONLY in 4 ohms or 8 ohms! speaker varies in ohms. and a regular cheap class-d does not behave the same in different OHMs.
I have compared the Hypex UCD180 HG with many cheap class-d amplifier cards and several older mid-range class-ab amplifiers and found that the hypex ucd180 hg has incredibly good sound quality!
The NCore 500 is certainly NOT a cheap amp, and of course I can listen and measure it with an 8 ohm load. If it can not drive an 8 ohm load, it cannot drive anything. I don´t know what it is you want to say?? I have made a UcD180 with a SMPS supply, without HXR, and it sounded terrible, and I have made a UcD400 with HXR and an analog supply, and it sounded pretty good, but the NC500 sounded like the worst crap.