400w DARLİNGTON AMPLİFİCATOR 2 ohm STABLE

norazmi

diysmps Senior Member
yup this amp work good, also have nice thd also need good heatsink if you drive lower load, you can substitute bd911 and bd912 with tip41c and tip42c.
 

Spare_Parts

New member
Simulators are not accurate when it comes to darlington in package like the tips 142 and 147.

What causes the inaccuracy? thermal coupling of the devices being in the same package is not included in the model perhaps? I have used the TIP142/47 SPICE3 .MODEL files that can be found at element14.com in other circuits... I will sim this circuit if I find the time before anyone else gets around to it :) This amp is class B, so crossover distortion is the main issue... Also, if anyone works on the PCB, the 3 1N4007 diodes should be located where they can be thermally coupled to the main heatsink, if possible.
 

Lira

Member
What causes the inaccuracy? thermal coupling of the devices being in the same package is not included in the model perhaps? I have used the TIP142/47 SPICE3 .MODEL files that can be found at element14.com in other circuits... I will sim this circuit if I find the time before anyone else gets around to it :) This amp is class B, so crossover distortion is the main issue... Also, if anyone works on the PCB, the 3 1N4007 diodes should be located where they can be thermally coupled to the main heatsink, if possible.

There is two resistor inside the package, one of 150ohms and another one of 5Kohms, but it varies from one manufacturer to another, even with the part number of the same factory you find very different values. That's the problem. I worked on this PCB, the fact is that i don't use the diodes the way is on the pcb, I use wires to put the diodes on the heatsink. But i'll redesign as soon as i have time.

Best regards

Lira
 

Spare_Parts

New member
Understood. I had previously been creating my own darlingtons by using discrete transistors and resistors, so that may be the best way... can apply "monte carlo" component drift properties to the resistors and see how the changes affect the transfer function. I bet the negative feedback will compensate and that it won't really change anything audible though. It may be easier to use TO-220 (transistor packages) diodes so you can mount them with screws rather than clamping diodes to the heatsink:)
Best regards
 
Last edited:

Lira

Member

Você ja montou este amplificador? para você conseguir 400w rms sobe a fonte para 37v + 37v (AC), se for usar apenas fonte de 28V + 28V pode até tirar 2 dos tips e trabalhar com apenas 2 pares por canal.

Did you assemble this amp? To get the 400W RMS you have to use a symetrical psu of 37V with a minimum of 8A per rail. with such a low PSU with 28V you only need 2 pairs per channel, and you will get about 180w RMS with 5% of distortion.

Best Regards

Lira
 

uranium

New member
i tried assembling this amp ..changes made : 1ohm emitter resistor in place of .33 , tip41/42c in place of bd911/912 ....
supply 35-0-35V / 8A
sound was very distorted
T1 was hot,tip147s were a little warm,tip142s cold..
diodes were mounted on main heatsink ..
I tried replacing all transistors but no difference...
searching net provided very little info on this amp...
many dont favour darlington amps ..
i was interested since lira made it work..& tip147/142 are cheap & easily available at my place...
 

Attachments

  • IMG_1511.JPG
    IMG_1511.JPG
    139.3 KB · Views: 80
  • IMG_1516.JPG
    IMG_1516.JPG
    132.8 KB · Views: 52

Lira

Member
i tried assembling this amp ..changes made : 1ohm emitter resistor in place of .33 , tip41/42c in place of bd911/912 ....
supply 35-0-35V / 8A
sound was very distorted
T1 was hot,tip147s were a little warm,tip142s cold..
diodes were mounted on main heatsink ..
I tried replacing all transistors but no difference...
searching net provided very little info on this amp...
many dont favour darlington amps ..
i was interested since lira made it work..& tip147/142 are cheap & easily available at my place...

Hi, Can you take some more pictures of it??? try to do a shortcut on the input and see if there is any tension on the output, something like 200mV is ok , but higher than this, check the BC`S. If you need anything I'll try to help you.

Best Regards

Lira
 

Lira

Member
UPDATE: I found some issues with new tips142/147, for some reason the last amplifier i mounted one week ago is not stable. I had this problem 1 year ago, the output is very distorted, the pnp transistors are warm and the npn cold, i'm using a +25 0 +25 dc power supply, i'm checking it, as soon as i find the problem i'll post here and how you can fix it.
 

Lira

Member
Hi sir Vedmitra, yes, but no good news yet. What i found out is the first units of this amplifier i used bd911 and bd912 and now i'm using tip41 and tip42, and also, i used the bc327 instead of bc556, using only 2 diodes for thermal compensation is more stable and generate less heat, and, to top it all, the latest Tips 142 and 147 has changed, i don't know why but they tend to be more unstable than the old ones and the quality of the output is horrible... I will retest it on the Multisim and change some values. i think i'll change the diodes with a vbe multiplier, and some resistors. i still have 80 boards of this amplifier.

Here is the pic of the one i was using on the first units:
images.jpg

And the ones that is creating all that troubles:
images (1).jpg

Sorry for the late answer, i was busy with my final tests of the semester and the final project at the university. but now i'm on vacation, and i have some more time ... i want to make this amplifier stable and with low distortion like the first units i build.

I'll post some tests in multisim with some modification and the pics of the board i'm using.

Best Regards

Flavio Lira
 
Last edited:

Lira

Member
Update: It is finally stable. Here is the mods:
This schematic is stable in simulation, but generates a lot of distortion in real life:
tip142_147.jpg
First it will only be stable with 2v the collector of the two driver transistors (bd911, bd912 or tip41,tip42). The gain is set in a crazy way on the original schematic, it uses a 3.3K and a 47R to make the gain , i changed it to 62K and 1K, it is far more stable. Here is the schematic that is really stable:
tip142_147_stable sch.jpg
And here is the simulation on 20Khz, its not perfect but it works... far less distortion!
tip142_147_stable 1.jpg
and with square waves:
tip142_147_stable 2 20KHZ.jpg

hope this helps
View attachment darlinton_sim.zip

Best Regards

Lira
 
Last edited:

Lira

Member
Here is an example using VBE Multiplier:
Fixing_using_VBE_multiplier.jpg

You can use this small board:

vbias board-5.png

and just put this instead of the 3 diodes, and replace the 470uf cap with a 100uf cap (C7 on the Schematic).

Best Regards

Lira
 

vedmitra

Member
Thanks Lira for multisim file,first thing i noticed is it needs a delay start as at switching on it throws lots of d.c., I will advice to simulate circuits with 1 volt input,and accordingly adjust feedback resistances, I will poke further to get the best out of this circuit in simulation.
 

Lira

Member
I think the capacitor of 470uF in the VBE multiplier is causing the inicial DC, I lowered it to 47uF and it is better, but i always use a dc protect board with delay on the outputs. One thing that i noticed, is that this amplifier absolutely doesn't work well with 2 ohms output and 50V its a bit high for the darlingtons, that is why i'm simulating with 40V. I tried yesterday this amplifier with tip35 and tip36, and works well, with the same problems, but well.
If you want the simulation, here it is:
View attachment tip35_36_VBEMULT.zip

The DC protection i'm using in almost all of my projects it is this one:
View attachment protector_salidas.pdf


Best Regards

Flavio Ribeiro de Lira
 
Top