LM1875 in parallel configuration and used in a composite amplifier.

[00940]

If you have lm1875 in stock, why not. I'm genuinely curious to see how you will manage the composite things. That's something I've recognized above my own skills.

Good luck with the project.

[FauxFrench]

Originally Posted by anti

.. and even then, there is the the LM1876

Was it to put only two in parallel, the LM1876 could be used. Very good chip. However, we have a lot of LM1875 in stock.......

[chermann]

Originally Posted by FauxFrench

You are right in that this LM1875 project is not about how to make a composite amplifier in the most rational way. Had it been, we would have used LM3886 or TDA7293. Your level of skills has been noticed and this project will hardly teach you anything.
We are a number of persons having LM1875 (more or less genuine) in stock or the option to buy LM1875 at a good price. The aim is to show how such simple but well performing chips can be arranged in parallel to compensate for their moderate current capability and allow BTL use. On top of that, use in a composite amplifier configuration will bring down the THD and allow feedback from after the load sharing resistors. I am convinced we will reach a good result but we will neither take away the interest for the Fremen design, nor put Neurochrome out of business. Due to only 5 pins, the LM1875 is easy to use in a Vero-board. Further, as you already point out, the heat losses will be spread out.

Yes 00940... i read a lot and with this "project" you will not learn much by me yes FF we try the LM1875- i 100% agree with your comments.


chris

[chermann]

Hi



to be clear which supply voltage we should have for 4 R load i did a test again.
to get 3Amps aout we need at 4,459R a Voltage rms about 13,37V at the resistor.


i use 25V supply but with the with 450mVrms i get this current limiter. switch on_off_ on off --> i show this in the LM1875 kit thread.
i go dwon with the supply 20V - this is the max off Voltage of the chip to do not go into the limiter. and here are the results:
pic 1 +/- 20V supply 500mVrms in (G=27dB) 11,2 Vrms at the load of 4,459R -- clipping
pic 2 +/- 20V supply 480mVrms in (G=27dB) 10,6 Vrms at the load of 4,459R -- clipping
pic 3 +/- 20V supply 460mVrms in (G=27dB) 10,2 Vrms at the load of 4,459R -- no clipping - this give us a current of 2,287Amps continuous

pic 4 +/- 15V supply 460mVrms in (G=27dB) 9,3 Vrms at the load of 4,459R -- clipping - supply voltage too low
pic 5 +/- 18V supply 460mVrms in (G=27dB) 10,1 Vrms at the load of 4,459R -- clipping - supply voltage too low - closed to 20V supply but not enough this give us a current of 2,265Amps continuous

pic 6 is the test enviroment - idle - no signal

so we should calculate per chip at 4R with 2,2 continuous?
chris

[chermann]

here is a summarize of that test


if you look at pic 3 and pic 5 you can see that the formula - look at the Vpp

Vout = Vsupply - 10V is correct!!!

[FauxFrench]

Good work as usual, Chris. The current limit kicks in at 2.3Arms corresponding to 3.25Apeak. I assumed use to 3Apeak in the initial estimations so that wasn't completely wrong.
Was this an Amazon chip or a genuine TI chip?

[chermann]

Originally Posted by FauxFrench

Good work as usual, Chris. The current limit kicks in at 2.3Arms corresponding to 3.25Apeak. I assumed use to 3Apeak in the initial estimations so that wasn't completely wrong.
Was this an Amazon chip or a genuine TI chip?

FF i will never say/write that you made a mistake....
if i did no mistake at my workbench....
original TI chip...to confirm this test i will do this test with my amp boards = amp 1_2 the nichicon UES board....

[chermann]

Hi FF
according to post 9

If we limit the supply voltage to +/-25V such that we can expect output signal clipping at 20V, the output power calculations become:
SE-coupling: 25W in 8 Ohm.
SE-coupling: 50W in 4 Ohm.
BTL-coupling: 100W in 8 Ohm.
BTL-coupling: 133W in 6 Ohm.
BTL-coupling: 200W in 4 Ohm.
Still respectable values.

what about the Transformer ratings?
2x15VAC - 160VA-200VA -230VA for both channels

chris

[kokoriantz]

Originally Posted by 00940

Sorry but that's just nonsense. The 0.95A max is for power transmission. Rating for chassis rating is already at 7A and even that doesn't apply to such short leads in open air. TI engineers will have determined the proper sizing of the leads based on temp raise and voltage drop.

PS: Eventual problems with low notes at high power come from the SPIKE protection system with the lm3886.

7A chassis is valid if the wires have their magnetic field constrained by a steel chassis or the leads are magnetic as the case for electrolytic capacitors and was with metal case transistors.

[kokoriantz]

here is a first approach for composite .