Pavel Dudek's (Upupa Epops) LM4780 amp

Jump to Latest
#221
I think you would just use one big one if you are going mono or two if you were going 2-channel.

I doubt that there would be a benefit for separate transformers for each module, since they are all processing the same audio. There could even be some weird interactions between multiple transformers but I'm not enough of an expert on electronics to answer that issue.
 
#222
Each modul, connected in parrallel, give 120 W/ 2 ohm. If you need ( theoretically ) 240 W/4 Ohm, you must to use two modules, which each will be connected in parallel, but one with " inphase ", second with " reversed phase " and load connect between both outputs. So for subwoofer you need two modules, for stereo operation four modules. I don't make kits, because in this amp are used commonly available devices, which you can to buy in every stores.
 
#224
Upupa Epops said:
So for subwoofer you need two modules, for stereo operation four modules.
Click to expand...

Thanks for clearing that up, Pavel.

It's great that you included balanced inputs into the design, my new main amps (under construction) will be run balanced and sit next to my speakers ~30 feet(10m) of cable away from my preamp and sources.

Is it better to feed the PA03 inputs(as subwoofer amp) a signal from the preamp or the output terminals of my main amplifeirs? Linkwitz crossover and equalization boards for his Thor subwoofers will precede it.

I don't make kits, because in this amp are used commonly available devices, which you can to buy in every stores.
Click to expand...

No $$$$ designer components? As Peranders posted seven months ago at the beginning of this thread "This can't be high-end. In fact, it must be rubbish." 😉

My bank account thanks you for using low priced parts.
 
#225
To PMA : Typical peak current of each channel ( guaranted by manufacturer ) is 11,5 A , so in parallel mode it is 23 A. I mean, that for 240 W/ 4 Ohm it is enough 😉. To darkmoebius : Better is to drive it from preamp ( with crossover on the way, certainly ) than from PA - you get lower distortion and better SNR. And any expenive components ? Good results make mainly good circuit design and sophisticated PCB, not " for Titan's developed " devices, what is often declared on amateur pages and kit's magazines. 😉
 
#226
Pavel,

we both are circuit designers, and as friends and colleagues I assume we should speak in a serious manner. The value 11.5Amps is output current limit for tON=10ms, typical. The guaranteed value is 7A (just look at the datasheet column at right side). 240W/4 Ohm means 7.74 Arms and 10.94Apeak, for purely resistive load. This will be increased for the complex load as speaker is. And the peak value you speak about is meant for output limitation in 10ms, not the continuous output current value.

There is another diagram in the datasheet, page 13. It shows 160W/4 Ohm in the parallel mode for 1% THD+N. Also, page 14 shows the safe operating area of the chipamp.

Frankly speaking, I would not recommend this 240W/4 Ohm, regarding the thermal load and safe area of the chip.

Regards,
Pavel
 
#227
To PMA : Sure, you have certainly thruth, this circuit isn't " heavy duty tank " 😉 , 240 W of output power is rather theoretical value. For long time reliability is better in this case to use this circuit up to cca 200 W ( in parrallel / bridge configuration ), maybe little bit less ( I haven't long time experience ). 😎
 
#228
Upupa Epops said:
Thanks to PMA you can see measuring results of distortion LM 4780 in PA 03 module, measured by his multitone method. You can compare distortion of generator with distortion of complete amp. All is measured by output power 5 W / 8 Ohm and results are surprisely good. At first distortion of generator by 10 kHz :
Click to expand...

5W into 8ohms... have you heard the expression "downhill
with a tail wind" 🙂

How about a stress test, say 50W into 4ohms.

Cheers,

Terry
 
#229
To Terry : Here is measuring results by classical method on PC ( peak at 15625 hz is relict from TV and 19 kHz pilot signal from tuner ) - you see, home method 😉 . This is distortion at 6 kHz by 50 W / 4 Ohm ( " basic " harmonic isn't good filtered ). You can see still good results 😎 .
 

Attachments

#230
moamps said:


100dB with 16bit measuring resolution?:scratch:

Click to expand...

100dB of resolution is 1 part in 10^5 . You have to do some range switching (or use a very good logarithmic amplifier) to get 100dB of resolution.

Take a look at manufacturer's spec sheets -- you very often have a case where there is a 50% probability that the smallest bit of an instrumentation ADC being correct. If you want to measure a microvolt referenced to 1 volt you need 20+1 bits of resolution. (So you make your work easier by changing the ref or amplifying the signal, all of which take a little time.)

Linear Tech has some good application notes on implementing very high resolution ADC's and DAC's -- these seem to me to be a lot more practical than Analog Devices or TI. Perhaps the crowd at LLTC has better editors, or more entertaining writers...
 
#234
This is photo of PCB. Do you like it ?
 

Attachments

  • frontpa03_small.jpg
    frontpa03_small.jpg
    89.4 KB · Views: 1,556
You must log in or register to reply here.
Top Bottom