Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 17th October 2013, 12:47 PM   #1
diyAudio Member
Join Date: Feb 2009
Default Questions about a TDA2040 based amp

Hi all,
since I'm relatively new in building amps, I'm here for some advice.
My plan is to build a single channel amp around the TDA2040 chip, I don't really need too much power as the amp will be used to drive a near field mini monitor at low to medium SPL. I already have a 50VA 9V toroidal transformer with 2 secondary windings that will be connected in series with a center tap thus providing -9, 0 9V or roughly -12, 0, 12V after rectification. I know that I can use higher voltages with this chip but this is what I have and for now I don't want to buy another transformer, I can do that later.

First question: what should I use as a smoothing cap after the bridge rectifier. 2x 2200uF? 4x 2200uF? some other values?

Second question is about the heatsink. The tab of the chip is connected to the pin 3 and with a split power supply this pin is connected to -V. I don't have any experience on that but I'm pretty sure I do need to (electrically) insulate the heatsink. What is the proper method of doing that? Is thermal grease between the tab and the heatsink sufficient, or do I need to use a thermal pad, or other method that I'm not aware of?

Third question is about power-on or off pop up. Does this chip provide some protection on that or do I have to worry about? What I have to do?

Last question is about the PCB. Following the datasheet I've designed the attached PCB. Does it seems good, or are there flaws that I'm not aware of? Also what happens if I change C3, C4 and C7 to 0.15 or 0.22uF (instead of 0.1uF)? Or if I change R4 to 2R2 instead of 4R7? Or R2 to 560R instead of 680R?

Any other comment will also be appreciated.

The datasheet of the TDA2040 can be seen here: Datasheet

Thanks in advance, Ralf
Attached Images
File Type: png amp_tda2040.png (31.5 KB, 238 views)
  Reply With Quote
Old 17th October 2013, 02:12 PM   #2
AndrewT is offline AndrewT  Scotland
diyAudio Member
Join Date: Jul 2004
Location: Scottish Borders
You have correctly separated the Signal side from the Power side. Good.
But go further: add pads for a Signal Ground to Power Ground combination of paralleled resistor & two inverse parallel diodes. There's plenty area at the bottom of your proposed PCB. You can link out this option if you build a true monoblock.
You have connected the Speaker Return to the Power Ground. This makes the junction in there (on the PCB), the Main Audio Ground (MAG). You MUST use this as your MAG if you build a stereo or multichannel amplifier. Even a dual mono must use this MAG (this becomes impossible because you end up with two MAGs that are non-coincident).

Add RF attenuation to the input.
Reduce the loop areas of all the signal pair traces.
Reduce the loop areas of the Power input.
Reduce the loop area of the Zobel.
Reduce the loop area of the Speaker feed.
Leave room to replace the input electrolytic with a plastic film capacitor.
Push the chip closer to the edge of the PCB, so that the heatsinks fits.
Add multiple holes/pads (different Pin Pitch) for every capacitor.
Move the chip decoupling to right next to the Power Pins.
Replace the film decoupling with X7R decoupling caps.
Add diodes to limit back emf from the speaker.
Add diodes to short out accidentally reversed power connections.
regards Andrew T.
  Reply With Quote
Old 17th October 2013, 03:18 PM   #3
diyAudio Member
Join Date: Feb 2009
Thanks Andrew, I'll have to digest what you said. Maybe I'll have later some questions about that. The first is what you mean with "reduce the loop area ..."?

And do you have the answer about the heatsink insulation? For this one I don't have a clue.

  Reply With Quote
Old 18th October 2013, 08:38 AM   #4
AndrewT is offline AndrewT  Scotland
diyAudio Member
Join Date: Jul 2004
Location: Scottish Borders
A circuit has a source voltage.
From that source voltage comes TWO wires.
I (and others) refer to these two wires as Flow and Return.
EVERY circuit has this Flow and Return.

The AREA included between the Flow and Return is an aerial.
This aerial emits radiation and it receives radiation.

Your layout should strive to minimise this radiation.
You do this by minimising the AREA enclosed by the loop formed between Flow and Return.

There is an alternative:
form a figure of 8 with the Flow and Return.
The first half of the loop sends out one phase of radiation. The next half sends out the opposite phase.
If the two loops are exactly equal and they are exactly the same distance from a receiver, then the two out of phase radiations cancel. This is why we see twisted pairs so often.
Every loop just about cancels with the adjacent loop in the twist.

You can't do this twisting with traces, at least not easily. Instead you can swap the traces from one side to the other to try to improve the cancellation via opposite phases. This is rarely seen, but Cherry described the technique and showed an example PCB in an old paper, about mid 90's.
regards Andrew T.
  Reply With Quote
Old 18th October 2013, 11:45 AM   #5
diyAudio Member
Join Date: Feb 2009
Andrew, thanks for the detailed explanation.
  Reply With Quote
Old 8th November 2013, 02:39 PM   #6
diyAudio Member
Join Date: Feb 2009
Answering one of my own questions.
What happens changing the R2 resistor to 560 from 680 Ohm: gain goes up from 27.4 dB to 28.5 dB.
  Reply With Quote
Old 10th November 2013, 12:59 AM   #7
diyAudio Member
Join Date: Dec 2003
Location: Sydney
Try point-to-point wiring. Don't use PCB. Keep wire length to a minimum, and use solid core wire. Sound-coupling caps C1 and C2 should be of very high quality.

  Reply With Quote
Old 11th November 2013, 12:01 AM   #8
gootee is offline gootee  United States
diyAudio Member
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Are your speakers 8 Ohms nominal?

You can use any amount of reservoir capacitance. But the reservoir capacitance determines the RATED max output power (at the onset of clipping), and the maximum ripple voltage.

You might be better off not using a center-tapped arrangement and using two bridge rectifiers, instead. That would use the transformer a little more efficiently, if I recall correctly.

You might want to use Schottky diodes for rectification, to conserve voltage drop.

With such a low rail voltage, a lot will depend on exactly what the worst-case diode drop is, and also on what the chipamp's clipping voltage is (i.e. the minimum voltage between PSU and output), at your rail voltage.

Looking at the datasheet's plot of max output power vs rail voltage, it looks like the "vclip" voltage of the TDA2040 is about 1.5 Volts, when the rail voltage is around 11 Volts.

So, your peak secondary voltage (12.73 V), minus the two diode drops, minus the vclip of the chip, minus the maximum ripple voltage, will be your peak output voltage at the onset of clipping. To guarantee that clipping can't occur, even at the rated maximum output level, you should assume that the output is constant DC at the peak output level. Solving the capacitor-discharge (differential) equation with those conditions can give an algebraic equation for capacitance as a function of the corresponding max rated output power level or ripple voltage amplitude. I did that, a while back, and made a spreadsheet that provides a range of output power levels and gives the corresponding minimum reservoir capacitance for each one. I'll attach it. You will have to figure out what the diode drops will be at the maximum output current, and enter the correct number for that, since it will depend on what kind of rectifier diodes are used.

Remove the ".txt" from the end of the filename, to use the attached Excel spreadsheet file.

If you want to see what rated max output power you would get, for a particular capacitance value, single-click in any column's capacitance cell and then select Excel's "Data"-->"What If Analysis" option, and tell it to make the capacitance value what you desire, by changing the value in the "Rated/Theor (%)" row (the blue row) for that column. I have the capacitance values set for various combinations of 2200, 4700, and 10000 uF. But they will change if you modify the mains frequency or cap voltage rating field. So then you might want to reset them to convenient values.

Remember, too, that the capacitance and ripple voltage values are all for worst-case at max rated output power levels. You can use a lower reservoir capacitance, if you don't care what happens near the maximum output level.


Attached Images
File Type: jpg giralfino_xls.jpg (430.1 KB, 116 views)
Attached Files
File Type: txt giralfino_capacitance_gootee.xls.txt (113.0 KB, 9 views)
The electrolytic capacitors ARE the signal path: http://www.fullnet.com/~tomg/zoom3a_33kuF.jpg

Last edited by gootee; 11th November 2013 at 12:21 AM.
  Reply With Quote
Old 14th November 2013, 04:12 AM   #9
diyAudio Member
Join Date: Jul 2003
Location: Dona paula, Goa
Simple amp chips are available that are easy to use e.g. TDA7377/9. These can be used for stereo bridged or you can make a 2.1 speaker amp from just one IC and very few components.

Gajanan Phadte
  Reply With Quote
Old 14th November 2013, 09:04 AM   #10
diyAudio Member
Join Date: Feb 2009
Are your speakers 8 Ohms nominal?

You can use any amount of reservoir capacitance. But the reservoir capacitance determines the RATED max output power (at the onset of clipping), and the maximum ripple voltage.

thanks for your work. I downloaded your spreadsheet, and played a bit. The only problem is that from the datasheet it is difficult to choose a max rated power because the only graphs are for 0.5 and 10% distortion, and I'd design for a lower distortion. Probably for 0.1% the rated power is about half of the max theoretical one, and from that given only 1 8Ohm channel, I can use without problems what I already have (4700u 50V caps).
Thanks again
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
PSU using TDA2040 ? poynton Power Supplies 7 10th February 2012 03:48 PM
LM3886 based amp- questions Atilla Chip Amps 12 6th August 2008 04:55 AM
TDA2040 Amp Fritzell Chip Amps 0 20th July 2004 03:31 PM
Did the TDA2040 pan out? KT Chip Amps 0 6th May 2004 04:47 PM
Questions for my LM3386 based power-amp. JoeBob Solid State 24 22nd December 2001 11:53 PM

New To Site? Need Help?

All times are GMT. The time now is 01:28 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2