SKU145225 - TDA7492 Class D amplifier board measurement - China reviews
Вставка
- Опубліковано 12 вер 2024
- A case study in out-of-band noise and QA issues, this $9,35 Banggood amp module is pretty crap. We take a look at the construction of the output filter, and modify it in order to actually work properly.
Performance figures (SPOILERS):
Test conditions: 26 V supply, 10 Hz-30 kHz bandwidth limit
Noise floor: -64 dBV unweighted
Power output into 8 Ohm
@ 0,1 % THD+N 20 W/ch.
@ 1 % THD+N: 35 W/ch.
Damping factor: 34
Frequency response: 10 Hz - 23 kHz +0,8/-0,2 dB
Gain: 33,4 dB
Temperature protection: Yes
Note that amplifier board ARRIVED BROKEN and needed repair before being safe to use. Soldering quality is very poor. Screw terminals are of poor quality, and their solder joints fail very easily.
--
THIS VIDEO IS RELEASED TO THE PUBLIC DOMAIN
To the extent possible under law, I have waived all copyright and related or neighbouring rights to this work, according to the Creative Commons CC0 affirmer. This means that anyone is free to use this content in any manner they see fit, without my objection. However, the video may still contain copyrighted content (images, video, audio, software, etc.) that does not belong to me. WHILE I DO NOT RESERVE THE RIGHTS OF ANY PART OF THIS VIDEO, THE COPYRIGHT OF THIRD-PARTY CONTENT STILL BELONGS TO ITS RESPECTIVE OWNERS. THE COPYRIGHT OF THIRD-PARTY CONTENT IS NOT WAIVED. THE VIDEO IS PROVIDED AS-IS, AND I DO NOT PROVIDE ANY GUARANTEE OF ITS QUALITY.
I have had a lot of these cheap class D boards, and while most do output pretty good power levels, they just don't come close to even a cheap class AB or B linear amplifier when it comes to sound, they have a really weird soudn to them that I cannot explain and upon further investigation, I found that all of mine output the same horrible sinewave as yours did unmodded!... And I have figured out the same thing you figured out here, they lack a proper filter on their outputs quite often. THey also cause interference on my antique radios, if I switch to the longwave band I can actually hear what the amplifier is playing through the radio! This makes me ponder the possibility of making a longwave transmitter using a class D amplifier... This is a great video, with great measurements, and I am now confident that my chinese amp boards can probably all become pretty decent if I modify their output filters a bit :)!
5.6Output low-pass filter
To avoid EMI problems, it may be necessary to use a low-pass filter before the speaker. The cutoff frequency
should be larger than 22 kHz and much lower than the output switching frequency. It is necessary to choose the L
and C component values depending on the loudspeaker impedance. Some typical values, which give a cutoff
frequency of 27 kHz, are shown in Figure 24. Typical LC filter for an 8 Ω speaker and Figure 25. Typical LC filter
for a 6 Ω speaker below.
The output LPF in the TDA7498 data sheet should be set much lower than 330kHz and higher than 22kHz.
There is.
Assuming the impedance of the speaker is 6-8 ohms, and 4 ohms is in the range of 6 ohms, you can use 4-8 ohm speakers.
however;
Inductance: 22μH
Capacitance: 2.2uF
So the cutoff is 23993.5104Hz, so this is actually the limit, but the impedance to the speaker is 9Ω, which is higher than the board speaker's 8Ω, so it's useless.
You want to recommend a 6Ω speaker, right? .
1uF in 22uH cannot filter 27kHz. Fc=33931.9479Hz;4.2Ω
0.68μF to 33μH is also useless 33597.6348Hz; 6.4Ω
Apparently, one board must be dedicated for each speaker, but there is no ideal for commercially available ones.
At 22μH, up to 1.5μF is fine for a 6-8Ω speaker.
Like your reviews of the cheap Chinese amplifiers like it even better when you come up with fixes in modifications that make them perform better
I got one recently and the built quality has improved a lot, also proper capacitors installed at the LC filter speaker output. Even at 12v & 8-ohm speakers, most of the time I use less than 50% volume. Ofc it is no match for expensive systems but I only paid £6 for the board and it was interesting playing with it.
Great measurements! Would be nice to see how the other slightly more expensive eBay boards (with the higher end TDA_ chips and much nicer looking output filters) compare to this.
+qTelec I'm glad you found the video useful. A shootout like that would indeed be interesting, especially since there doesn't seem to be too many people around UA-cam with proper audio test gear, who are willing to spend their time testing cheap Ebay stuff. Perhaps some time far in the future, I'll have gathered enough ad revenue/donations to purchase a few boards for review. If you're acutely curious, you can send me a board/donation for a board, and I will test it for you.
This is how I discovered your channel!
I'm planning on doing a portable boombox with this amplifier.
If you want updates on the projects tell me!
I will only use it to power 15W maximum at 4ohm speakers, which means I'll have to change the output filter.
I have already made AutoCAD sketches for the enclosure and I have the bill of materials made. I just need to buy and wait for the component to arrive!
Good luck, I hope it turns out hi-fi.
which amplifier you can reccomend in this class?tpa3116 or tda7492?
papaVasyaA Definitely TPA3116
why?i read it has high distortin
papaVasyaA Depending on what voltage and power you are going to supply to the amp of course. For < 20W and 12V I would choose TPA7492P but for 20-40W 19-24V, the TPA3116 is way better.
I felt like an idiot during this video, I tried reading his mirrored meter without parallax on my monitor! That just doesn't work.
Hello, greetings from Mexico. Great videos, very in-depth and technical.
Which amplifier is better in your opinion? The TDA7492 or the TPA3116D2? They're almost identical in input power requirement and output power but I've found no direct comparison of them. I'm seriously thinking about getting the "breeze audio" amplifier you already reviewed and modded in one of your videos with the TPA3116 chip in it, though the gain seemed a little high and needed messing a little bit with the board and soldering skills to get it to 20db. On the other hand, the TDA7492 board I found has easy-to-adjust gain by just turning on and off a dip-switch to adjust it from 21.6db to 31db and has also Bluetooth connectivity. But I still haven't got the opinion of someone who knows in depth and has dealt with both of these amplifiers. Thanks in advance
The chips themselves are pretty similar, what matters is the quality of the surroudning design.
I got a cheap 50 watt amplifier/reciever, it has bluetooth CD player, am/fm, two way speakers and remote and I'm pretty sure it's a Class D amp as it doesn't give off any heat, It's by Innovative Technology just a cheapo unit from Wal-Mart only 100 bucks Canadian to amplify my laptop . It's incredible the features and performance when you gotta think my Pioneer component system from the 90s has less features and was over 1000 dollars. It honestly sounds good , even better since I have some Polk audio bookshelf speakers hooked up. This board looks like its pretty crusty but I wouldn't expect too much for so cheap.
Hi! I like your videos very much! Especially when you measure cheap ebay amps that i can afford! I was looking into the irs2092 chip, because a lot of people recommend amps based on this chip, like the L15D pro, and are talking about "hifi" sound, could you please look into them, if you are still interested in amplifiers?
I do flavour-of-the-week china amps whenever I have the time to do so.
Väldigt intressant
Tack!
Very detailed breakdown. I was wondering if you could help a lot of us DIY speaker builders. Adding a crossover to small speaker projects is really too expensive. Any way you could show us how to change the bandwidth on a little amplifier? Could it be as simple as adding an output capacitor to eliminate low frequencies below 100hz or so? Thanks in advance for any advice.
You can replace the input series capacitor, or just add another one in series according to the input impedance of the amplifier (has to be measured).
+FFcossag thanks for responding so quickly! You da man.
hi did you manage to re test this amplifer with a 470 n capacitors fitted and the filter to ground in place and removed I would be very interested in your results great video by the way
+jonas Dogue de Bordeaux I haven't modified this amplifier any further since this video, sorry.
+FFcossag that's a shame as I changed the caps 100 nf you suggested i replaced ghem with 470 and it sounds a better plus the signal caps were replced,
jonas Dogue de Bordeaux While it was a while ago I was toying with this thing, I recall seeing some odd frequency response characteristics when you change the filter around that much. I would not be surprised if you've got quite a lot more or less treble than it originally had.
+FFcossag it was more smoother I don't have the test equipment to test for distortion. I would love to buy the perfect class d amplifer I look forward to future tests on classd d thanks for sharing your input
I really wanted to build an amp for years. But after Topping and that one French company made what is considered to be #2 amplifier in the world for less then 1,500 bucks? I gave up on the idea. I do like those 10 to $15.00 amps quite a bit for basic things I got. But I just gave up on all these cheap dreams of big power from these amp boards.
The Brazilians got the Sub woofer amp market down pat with power output. and the cost and the amount of work you have to put in to these amplifiers to surpass anything that's in the car audio world is a losing proposition.
the only kits to bother with are the ice units for home audio mains voltage gear.
DAC'S? Topping or Schitt. Few others. But those are the tested and proven buy and try and be done with it.
these days, if something doesn't have a balanced input it just doesn't pass muster
Looking on ebay for a cheap 50w+50w, 12v amplifier board - preferably with bluetooth. Anything you would recommend?
Got one of these to amplify the headphone output of my tablet.
Unfortunately it seems to pick up some noise from the screen/led backlight driver - a fizzle/hiss.
I can reduce this down if I connect a 22uF electrolytic cap and a 1M resistor over each input - but I can completely get rid of the remaining hiss by touching any ground point.
This is running off a 12v lead acid, the noise definitely appears to be coming from the (ungrounded, on battery) tablet. Any further suggestions?
+Marineio use a damn serious Inductor-Capacitor filter between the battery and Vcc, make any GROUND wire you connected thicker, use those thick and hard ones that come in a phone line, or thicker. not those of a mains cable, those are messy.and inductor that can be salvaged from inside a fluorescent lamp is one of the best choices if you supply your board with a high voltage, like >20V, otherwise, you can exploit a secondary wire of an old step-down transformer (leave it in the transformer, forget the other wire), choose it thick enough and long enough (resistance must be low ( < 2 Ohms))in this way you can enhance the supply rejection, which is a problem in class d amps more than class ab ones
FFcossag I just watched this and I noticed one thing, you are feeding this amp 26v !! Ok well your electrolytic cap is rated at only 25v !! I guess you didn't notice this and I am not sure if it would have affected your results but you might think about it !!
You're right but it doesn't matter, at such little over-voltage, it's just a question of the cap not really lasting as long as it's supposed to.
+FFcossag what would you get if you used 18-19vdc @ 6ohm
Would love to see you upgrade a class D amp by putting the PWM signal into a large mofet and filtering the mosfet's output.
That'd be a pretty cool idea, but it does put even more stuff outside of the feedback loop, which isn't good.
I thought Class D amp didn't have a feedback loop?
Of course they do, however the output stage is not included in that.
Then make it a sub amp and don't worry about high frequency distortion or something. I think you're clever enough to make it work well, especially with a very limited frequency response.
Sure, but there isn't much point to it since a proper, higher power china amp costs nothing anyway. While figuring out your own high-power output stage could be a fun exercise, I don't think it has much practical application.
why do you crowing?
Hi Which caps need to be 470nF for 4ohm speakers?
have you done tda7498 ?
Nice man!
Nice Video!
I got the same Amp and want to modify it as in your setup.
I was going to change C3 and C24 to 220nF as you suggested and get new 100nF for C1, 2, 23 and 22 (ceramic or foil?)
Any suggestions what else to change? Are there better values? Because I have to order new components anyway.
What about the two 1uF at the input? Replace, change value or just leave them as they are?
Thanks!
+Lobot321 The input caps should be fine as they are. You should use film caps whenever you can, ceramics aren't usually very linear and can induce distortion.
The only other real thing I'd change about this amp is to set the gain to a lower value, which shouldn't be too hard to figure out from a dig through the datasheet.
+FFcossag Thanks for your quick reply.
Gain0 and 1 are set by the resistors R8 and R28 (each 30k as 3.3V pullup). I am pulling them down each with a standard NPN transistor between resistor and gain-pin. This way I can activate gain should I need more volume with a Raspberry Pi.
Same goes for the standby function. This is controlled by R23 (300k pullup). I can set the amp to standby by pulling down between R23 and standby-pin with a NPN.
Would you suggest changing L1-4 from 22 to 33uh (I am using 8ohm speakers) and remove the additional filters with the 8R2?
+Lobot321 I can't really say; this amplifier performed about as well as you'd expect it to when I tested it on 8 Ohm. I don't think changing the filter is going to do you much good.
+Lobot321 hey,
how can i decrease the input gain? i didn't understand everything from your mod. Only that you can adjust it now :D
+sinoBLN
Just remove R8 and R28 if you want to deactivate gain permanently.
If you want to be able to switch gain on and off you need to pull the 3.3V between resistor and gain input to ground. For that you can use an NPN transistor or a mechanical switch.
Hi! Can you test the tda7498 ? :D
I have something like that in the works for some time in the future.
Have You changed values of ceramic capacitors?
+Pauliaus Elektronikos Projektai The small ceramics I used were the same 100 nF as were installed from factory. The two film caps I used are 1 µF.
+FFcossag thank You for replying. Interesting and useful video. Subscribed :)
I have a 2.1 amp board (bit.ly/2AaQzFM) using a 19V battery. My actual setup is a 6ohm subwoofer connected to the sub output and only one 8ohm (30W rated - 60W max) speaker connected to the left output, so the right output channel is unused. My doubt acordingnly to this table (bit.ly/2jlJavl ) is that at 19V the max power output on each L/R channel is 25W at 8 ohm. Actually the setup is working well but I really don’t know if the 8ohm speaker or the amp are actually getting stressed (if that is the word).
So my question is if there is a way to combine both L/R output channel to get one single mono output channel with twice the power (25W+25W), so the speaker and the amp can work more relaxed and so the amp will not blow up.
What would you suggest me?
I don’t know too much of electronics so I don’t know if what I exposed above is actually possible and correct.
Thanks.
That amp will always be stressed because those heatsinks are not big enough. Beyond that, it's fine. When it comes to audio, as long as it sounds good, you're good.
Also, no, you can't combine the channels, as they're already in a BTL configuration.
That fire truck tho..
lol..
COOL ROTOREY SOLDRE JOINTS HEHE
Stupid lead free solder! >:C
RoHS compliant
Class d stuff is not digital, it's switching.
+Bogdan Serban By being "switching", it's also digital; the signal is converted to a pulse-width modulated signal which is binary in nature. The entire output stage up until the output filter is, well, binary.
+FFcossag this is a common confusion. Digital essentially means binary, strings of 1 and 0 that can be translated. Digital machines work by using switching transistors, but what they do is compute numerical values. Here switching transistors are used because they are more efficient rather than in their linear state. But what they do eventually is modulate a voltage.
+Bogdan Serban Yes, that is entirely correct. The output stage is still digital up until the output low-pass filter turns it back into an analogue signal. Whether or not the "data" is machine readable or pulse-width modulated for the sole purpose of being filtered to analogue, it still only has two levels - it's still binary - and it's still digital.
+Bogdan Serban !'ve got a digital tripod for my camera, what does THAT mean? It is Chinese of course.
I look at digital as a sub-realm of analog ... because, when you get to high frequency digital stuff, PCB trace shape, length, position, size, and distance DO matter!
еееееее рооок