Has anyone brought one of these. I should get one in the next day or two to assemble. It took a long time to get delivered because the 1st one shipped got bounced back to the seller.
I was looking for a low power reasonable performing 2 channel amp to run my tweeters in tri-amp speaker. The specs aren't the best, but seem ok if you don't push the limits of its power. I doubt if I would use more than 10w peak. The reason for this choice it is seems to have speaker protection that I hope works during power on and power off cycles. Going to use a 100VA 36vct transformer with it. After I bench test it if it works well I have to find a case and heatsink for it.
2018 GC version LM4766 dual channel HIFI power amplifier board diy kit | eBay
I was looking for a low power reasonable performing 2 channel amp to run my tweeters in tri-amp speaker. The specs aren't the best, but seem ok if you don't push the limits of its power. I doubt if I would use more than 10w peak. The reason for this choice it is seems to have speaker protection that I hope works during power on and power off cycles. Going to use a 100VA 36vct transformer with it. After I bench test it if it works well I have to find a case and heatsink for it.
2018 GC version LM4766 dual channel HIFI power amplifier board diy kit | eBay
Hi,
at first sight, this amp looks OK and no major components are missing. The real LM4766 is an amp just sounding perfect. You can discuss the speaker protection, as the real LM4766 chip has an anti pop and fade in fade out function. So why add one at extra cost?
I write "real" because you can trust no Chinese seller to use the expensive ($3) TI chip. There are millions of fake LM3886 (comparable chip, just in Mono) around. Some even work, most not as expected. What the chip fakers are unable to copy, are the advanced safety features and refined sound of the original. So the corresponding pins are connected to similar working, cheap circuits, that is all you get. Most blow, sooner or later.
Anyway, if you have such an amp and the circuit is close to the reverence design, throwing away the LMxxxx is easy and Mouser or RS will sell you a real one, making you a happy camper.
Please, before you connect an expensive tweeter, get the data sheet and compare the values of the parts used, no one in China cares about 1.000 or 10.000 Ohm. They care only for money.
A real DIYS amp...
at first sight, this amp looks OK and no major components are missing. The real LM4766 is an amp just sounding perfect. You can discuss the speaker protection, as the real LM4766 chip has an anti pop and fade in fade out function. So why add one at extra cost?
I write "real" because you can trust no Chinese seller to use the expensive ($3) TI chip. There are millions of fake LM3886 (comparable chip, just in Mono) around. Some even work, most not as expected. What the chip fakers are unable to copy, are the advanced safety features and refined sound of the original. So the corresponding pins are connected to similar working, cheap circuits, that is all you get. Most blow, sooner or later.
Anyway, if you have such an amp and the circuit is close to the reverence design, throwing away the LMxxxx is easy and Mouser or RS will sell you a real one, making you a happy camper.
Please, before you connect an expensive tweeter, get the data sheet and compare the values of the parts used, no one in China cares about 1.000 or 10.000 Ohm. They care only for money.
A real DIYS amp...
Any way to tell if it is a fake? So a real LM4766 wouldn't need a speaker protection relay as those function at internal to the chip. I saw some assembled LM3886 amp that had speaker protection relays and other didn't. When I looked that the data sheet for the LM3886 I also thought why do you need a relay of the chip does it for you.
I have some old Hafler DH220 that don't have speaker relays and they have an bad spike on power up and a less severe noise on power down. I'm making a power cycle speaker relay for them. If I connected the Haflers directly to a tweeter they would likely be damaged during one on cycle!
To be on the safe side I'm planning to use an 1st order cap as a high pass filter between the amp and tweeters.
I have some old Hafler DH220 that don't have speaker relays and they have an bad spike on power up and a less severe noise on power down. I'm making a power cycle speaker relay for them. If I connected the Haflers directly to a tweeter they would likely be damaged during one on cycle!
To be on the safe side I'm planning to use an 1st order cap as a high pass filter between the amp and tweeters.
yep and according to FF experience the UTC 2050 are survive this test too!
a little bit off topic because this amp is BTL but you need a tweeter amp but....:
LM4766 BTL bridge 120W mono power amplifier board|Amplifier| | - AliExpress
chris
The marketing errors
Product specifications:
● Output power: 120W (maximum undistorted power, 4 ohm load)
● Working voltage: 18V+0+18V (AC communication)
●Magnification 22 times
● Product size: 68mm X 70mm
●The power amplifier board is mono
● Power amplifier board weight: 100g
I don't know how this would get 120Wat 4 ohms. The LM4766 is current limited to about 4 amps. So at best I2R = 4x4x4=64 watts for 4 ohms, 4x4x8=128 watts into 8 ohms. So they should say 120W into 8 ohms not 4. It wouldn't fit my needs as I need 2 channels and to have no use for more than probably 10 watts per channel.
Everyone seem to loves the LM3886, but not much is said about the LM4766. In reviewing the specs it seem that the LM4766 if not driven to high output levels has a reasonable low distortion level on par with the LM3886 for my old ears that can even hear 12KHz. Seem like a good chip for 2 channels if driven to less than 20 to 30 watts depending on the PS voltage.
I got the thing assembled. No instructions no schematics. Just a PCB and bag of parts. The silk screen has most of the parts labeled except for a few polarized caps. I had to look up the datasheets for the C1237HA and LM4766 to figure out what caps went where.
Last edited:
I assembled it and tested with a bode plot at 3 different input levels. The frequency response seems OK at each given input level, I see a slight phase shift, but of concern I see about a 4 db loss of gain as the input level is increased. I don't have a large enough resistor to test at higher levels, only went to about 8 watts. Should be able to go to 30+ watts in which case the gain loss would probably be even worst.
Amp[1] = input PPV
Amp[2] = output PPV
RMS[2] = output RMS V
Amp[1] = input PPV
Amp[2] = output PPV
RMS[2] = output RMS V
Attachments
Hi Blackice
i am not sure if i understand your plots correctly. the last - pic 3 is about 8Watt (7,5W) into 8R load. ok fine - you requested about 10WAtts-right?
you transformer is a 36CT - is 18-0-18VAC is after rectifier about 26VDC per rail.
so about the power - you have 7,73rms --> Vout peak is 10,9V --> so out pp 21,8V sin wave so you have some head room to go higher until clipping . i guess that 26VDC rail drop down to something like 23V or so- so with a bigger load resistor you should can go up.
pic 1 - 2 what do you mean with PPV-
purple pic 3
that means the purple line shows the frequency response and i can´t see 3dB loss? - at the low frequency you have a "loss" just about 1db 29,85 down to 28,85dB - this is not bad!
the amp input is 712mV rms and output is 21,8Vrms that give a gain about 30,6 and its about 29,6dB. i would change the gain setting down for less hiss. Rf in the datasheet (pin3-7 and pin1-12) is 20k ohms - i guess you have about 30k there...
chris
i am not sure if i understand your plots correctly. the last - pic 3 is about 8Watt (7,5W) into 8R load. ok fine - you requested about 10WAtts-right?
you transformer is a 36CT - is 18-0-18VAC is after rectifier about 26VDC per rail.
so about the power - you have 7,73rms --> Vout peak is 10,9V --> so out pp 21,8V sin wave so you have some head room to go higher until clipping . i guess that 26VDC rail drop down to something like 23V or so- so with a bigger load resistor you should can go up.
pic 1 - 2 what do you mean with PPV-
purple pic 3
that means the purple line shows the frequency response and i can´t see 3dB loss? - at the low frequency you have a "loss" just about 1db 29,85 down to 28,85dB - this is not bad!
the amp input is 712mV rms and output is 21,8Vrms that give a gain about 30,6 and its about 29,6dB. i would change the gain setting down for less hiss. Rf in the datasheet (pin3-7 and pin1-12) is 20k ohms - i guess you have about 30k there...
chris
Sorry my brain was working backwards I meant VPP not ppv or volts Peak to Peak. What I meant was at low input voltage in the first picture of the gain is approximately 34 db where 3rd photos the gain at a higher input voltage is about 30 Db. Test resistor was a 10-ohm 10 watt resistor. Another interesting fact about this kit the input impedance is only 1 K ohm.
Funny, I was just looking at this exact same amp board on Taobao,
Taobao GC Board
What is your overall impression?
I was going to buy the blank board and fill it with components from Digikey or Mouser.
Thank you,
David.
Taobao GC Board
What is your overall impression?
I was going to buy the blank board and fill it with components from Digikey or Mouser.
Thank you,
David.
Hi
As Turbo wrote. ......useless.....
if you dont need the power of a LM3886 then the 4766 is fine. relay and other things are not needed if you use original parts...
in post 5 i found a mono bridge board.
LM4766 BTL bridge 120W mono power amplifier board|Amplifier| - AliExpress
chris
As Turbo wrote. ......useless.....
if you dont need the power of a LM3886 then the 4766 is fine. relay and other things are not needed if you use original parts...
in post 5 i found a mono bridge board.
LM4766 BTL bridge 120W mono power amplifier board|Amplifier| - AliExpress
chris
I have assembled LM3886 and it does make pop sounds when turning off. When measured the power it was as it should be and thought this chip should be alright.
Is there a risk it is fake if it pops? In that case the test you mentioned should help.
Last edited:
Hi Taskitas
every chip LM1875,LM3886 should have a short protection mode if its an original chip!
So if you use a chip with unknown source it may be a fake chip without this protection and you short the output during a signal -then the fake chip will die...smoke.burn...what ever.
the lm3886 has a mute circuit . check with datasheet if you components on your board are as recommended.
if I used at the LM1875 the fake chip i get strange sounding pops a max power.
chris
every chip LM1875,LM3886 should have a short protection mode if its an original chip!
So if you use a chip with unknown source it may be a fake chip without this protection and you short the output during a signal -then the fake chip will die...smoke.burn...what ever.
the lm3886 has a mute circuit . check with datasheet if you components on your board are as recommended.
if I used at the LM1875 the fake chip i get strange sounding pops a max power.
chris
Last edited:
I have assembled following these instructions:
A Complete Guide to Design and Build a Hi-Fi LM3886 Amplifier - Circuit Basics
Although I'm running it on 24-0-24V toroidal, but did calculations accordingly as mentioned in the guide.
I'm wondering about details of the test, do you play something specific, like 1khz and high volume, or it doesn't have any difference?
And do you just touch the wires or hold for a few seconds?
A Complete Guide to Design and Build a Hi-Fi LM3886 Amplifier - Circuit Basics
Although I'm running it on 24-0-24V toroidal, but did calculations accordingly as mentioned in the guide.
I'm wondering about details of the test, do you play something specific, like 1khz and high volume, or it doesn't have any difference?
And do you just touch the wires or hold for a few seconds?
Last edited:
Hi
if its an original chip you can keep the short some seconds ! please use a current limiter or a bulb limiter at the mains to avoid damage your amplifier. put the amp board out from your housing and use a lab supply.
please read the datasheet and other documents.
As seen in Section 5, a short to ground with an input pulse applied to the amplifier will be current limited by the conventional current limiting circuitry for a few hundred microseconds. When the junctiontemperature reaches its limit, SPiKe protection takes over, limiting the output current further, as the junction temperature tries to rise above 250°C.
chris
if its an original chip you can keep the short some seconds ! please use a current limiter or a bulb limiter at the mains to avoid damage your amplifier. put the amp board out from your housing and use a lab supply.
please read the datasheet and other documents.
As seen in Section 5, a short to ground with an input pulse applied to the amplifier will be current limited by the conventional current limiting circuitry for a few hundred microseconds. When the junctiontemperature reaches its limit, SPiKe protection takes over, limiting the output current further, as the junction temperature tries to rise above 250°C.
chris
Last edited:
Actually, 4 A is the peak output current for a typical IC (i.e. the mean of the statistical distribution of a large sample size). 4 A, peak would give you P = 4*4*4/2 = 32 W into 4 Ω. The power into 8 Ω would be limited by the supply rails to the IC.
However, note that the LM4766 is only guaranteed to be able to provide 3 A peak, which lowers the output power in 4 Ω to: P = 3*3*4/2 = 18 W.
If you run the chip at the max recommended voltage of ±30 V, you get a peak output swing of 30-4 = 26 V, peak. The '4' is the drop-out voltage, which is guaranteed to be no larger than 4 V according to the data sheet. 26 V peak gives you 26*26/(2*8) = 42 W into 8 Ω.
With the 3 A peak output current limit, you're looking at P = 3*3*8/2 = 36 W into 8 Ω.
I would expect a typical IC to be able to deliver about 40 W into 8 Ω and 30 W into 4 Ω, assuming that it can actually run at ±30 V and deliver a clean signal to a 4 Ω load.
Or they just made it up. Or they grabbed the wrong number from the data sheet. Or any combination of the two. It's enough to raise some concerns about the competency level of the designer, that's for sure. That's not to say you can't use the board. It could be a fun platform for experimentation.
The LM4766 looks like a very capable chip. It looks like a dual LM3886 with their output stages scaled back (and it might indeed be exactly that). I bet the reason "everybody loves" the LM3886 is that it provides higher power. "Want" often wins over "need".
Tom
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.