Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Class D Switching Power Amplifiers and Power D/A conversion

[design log] Neat 2x170W I2S in, I2C controlled, integrated DSP amp (TAS3251)
[design log] Neat 2x170W I2S in, I2C controlled, integrated DSP amp (TAS3251)
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
Reply
 
Thread Tools Search this Thread
Old 15th January 2019, 11:50 AM   #111
doctormord is offline doctormord  Germany
diyAudio Member
 
doctormord's Avatar
 
Join Date: Oct 2007
Quote:
So I understand that, as a basic rule of thumb, the higher the swithching frequency, the better. Provided that the PCB layout is not wrong.

Not in general, it depends on voltage differential, load current and efficiency goal.


The point here is to push the switching frequency of the voltage converter away from the switching frequency (and their harmonics) of the amplifier to avoid "beat-tones" which may occur as a subharmonic oscillation product between these two frequencies.


I.e. When amp is running at 600kHz and your SMPS is at 605kHz, beat-tones at 5kHz (and their harmonics) may occur. As frequency of both always shifts a bit due to temperature and load current, it's not only a single tone but something like a "weeping noise". This is an example but effects are not limited to just that.


Quote:
Another issue is with light loads when the " intelligent" device may adjust its frequency. Hence the interest in our applications of Constant on Time devices.

Right, avoid frequency foldback and burst-mode. COT is prefered/needed if flybuck is going to be used to generate a second output voltage. The other controllers are fine if "possible issues" keeped in mind and the converter is calculated/layed out right.


Quote:
Last , I'm not clear if there are drawbacks to go directly to 12V with the MP2459 or LMR16006 and skipping the LM2940 (compared to the EVM design that goes to 15V and then use a MP2940 to build the 12V)?

When using a relatively high converter switchinh frequency, the following LDO can be ommited as their is most likely no benefit on filtering. (Just have a look into the datasheet of the LDO to see their PSRR/CMRR at switching frequency. Pretty much all don't do much at frequencies above 100-200kHz) Filtering with simple RC or LC filter is more effective.


The EVM converter is running at rougly 66kHz. (Calculated from LM5010 datasheet equation 7, page 11 with known Ron=182kOhm from TAS3251EVM schematic)


LM5010: http://www.ti.com/lit/ds/symlink/lm5010a.pdf
TAS3251EVM:http://www.ti.com/lit/ug/slau757b/slau757b.pdf
  Reply With Quote
Old 15th January 2019, 06:48 PM   #112
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
Oups, busy time. I'll take time to read all this carefully.


Maybe I'm too cautious about hand soldering the LM5010 ? Could be easier than I think?


Best regards,
JMF
  Reply With Quote
Old 15th January 2019, 07:20 PM   #113
doctormord is offline doctormord  Germany
diyAudio Member
 
doctormord's Avatar
 
Join Date: Oct 2007
I'd recommend getting one of the "cheap" reflow stations for those kinds of ICs, makes repair easier as well.

If you're on a budget, get a 858D:
Neu SMD Heißluftlötstation Digital Rework Station Lötstation + 3 Nozzles 858D | eBay

If you want an "okay" soldering iron combined into this, get a 898D:
SMD 898D+ Digital Rework Station Heißluftlötstation Heißluft Entlötkolben W.E.P | eBay

I have the latter, they are okay for the money and handy for repairs and all kind of heat shrinking. The soldering iron is at about 40W, with simple bang-bang-control. Not to bad. Basically it's the same as the older 938 Aoyue types of soldering iron skillet. I use two 938 since about 15years now.

Just make sure their earth-connection is working and low resistance.

Last edited by doctormord; 15th January 2019 at 07:35 PM.
  Reply With Quote
Old 16th January 2019, 07:37 PM   #114
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
Quote:
Originally Posted by doctormord View Post
I'd recommend getting one of the "cheap" reflow stations for those kinds of ICs, makes repair easier as well.

If you're on a budget, get a 858D:
Neu SMD Heißluftlötstation Digital Rework Station Lötstation + 3 Nozzles 858D | eBay

If you want an "okay" soldering iron combined into this, get a 898D:
SMD 898D+ Digital Rework Station Heißluftlötstation Heißluft Entlötkolben W.E.P | eBay

I have the latter, they are okay for the money and handy for repairs and all kind of heat shrinking. The soldering iron is at about 40W, with simple bang-bang-control. Not to bad. Basically it's the same as the older 938 Aoyue types of soldering iron skillet. I use two 938 since about 15years now.

Just make sure their earth-connection is working and low resistance.

Thanks for the links. We have such things at the Fablab. But is is not clear to me how to use those to solder the LM5010 bottom pad. I'm sorry for this basic question that shows my ignorance. But I'm willing to learn.



Is it something like this YouTube


Or something like YouTube ?


Are you using solder paste / stencil ?


JMF

Last edited by JMF11; 16th January 2019 at 07:47 PM.
  Reply With Quote
Old 17th January 2019, 05:22 AM   #115
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
I see that some of the designs developped here have commercial intentions. I don't have. Moreover, due to my lack of knowledge in some areas, I need collaboration to succeed having something working. But I don't see a lot of open designs here (gerbers at the most)



Question: do you think that going "fully open source" and sharing KiCAD source files would help ?


Is there a risk that experienced contributors may step back fearing to see their advices and contributions injected in easily accessible files ?



I don't think that what I do is rocket science, but I would like to go in a direction that maximizes contributions.


JMF
  Reply With Quote
Old 17th January 2019, 06:00 AM   #116
lutkeveld is offline lutkeveld
diyAudio Member
 
Join Date: May 2013
I would certainly contribute. The problems I see:
-How to make a final design decision in a group of contributers?
-How to work in a shared folder with the possibility of multiple people editing the same file? Or how would people contribute?
-Who is taking the risk for final production?
  Reply With Quote
Old 17th January 2019, 06:17 AM   #117
doctormord is offline doctormord  Germany
diyAudio Member
 
doctormord's Avatar
 
Join Date: Oct 2007
JMF, it's like shown in the videos, that way you may solder the LM5010.

For the point about sharing your project files, another option is, that you see your efforts done all the way finally in a product from China. Been there, done that.

Btw. people with commercial interest mostly give the least contribution in detail at design phase. (personal experience)
  Reply With Quote
Old 17th January 2019, 07:10 AM   #118
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
Quote:
Originally Posted by lutkeveld View Post
I would certainly contribute. The problems I see:
-How to make a final design decision in a group of contributers?
-How to work in a shared folder with the possibility of multiple people editing the same file? Or how would people contribute?
-Who is taking the risk for final production?
My idea was more basic. Like the TI EVM that people can have details about and start from, build upon.

It seems to me that we may all have different constraints, habits, understandings... So converging to a unique design that suits all has high probabilities to fail.

So it was more about allowing project "forks" based on shared work.


JMF
  Reply With Quote
Old 17th January 2019, 07:18 AM   #119
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
Quote:
Originally Posted by doctormord View Post
JMF, it's like shown in the videos, that way you may solder the LM5010.

For the point about sharing your project files, another option is, that you see your efforts done all the way finally in a product from China. Been there, done that.

Btw. people with commercial interest mostly give the least contribution in detail at design phase. (personal experience)
Yes, I understand this. As I don't have commercial interest for this project, my preferred option would be for it to be open to the community. Even OK for me for Chinese cloning if it derives in good product that I can buy for cheaper than I can make it by hand.

But it is important that going open won't prevent people like all those providing advices here, to contribute. This is my fear.

An alternative would be to PM files between interested contributors (or having a restricted access folder). And only publish PDF on the forum.

Don't know what is the best,

JMF
  Reply With Quote
Old 23rd March 2019, 07:23 AM   #120
JMF11 is offline JMF11  France
diyAudio Member
 
Join Date: Dec 2013
Hello,


The project is not completely stalled :-). I decided to start routing the project, event if it was not completly finalized. Sort of curiosity bias... Lot of professional activities aside made progressing slow.


For the routing, I took some inspiration from the TAS3251EVM (unfortunatly 4 layers) and from the TPA3255EVM (2 layers).


For hand soldering, I went to 0805 everywhere, replacing the 0402 and 0603 by bigger ones.


I still have to route the powersupply modules.


This is my first PCB of all my life... So I would really need advice about:
- the PCB, and avoiding the mysterious "ground loops",
- avoiding PCB mistakes
- tuning the values and part number of some caps and the coils.



10cmx10cm seems quite dense for the design. This was my target because of the super low PCB manufacturing prices, but 15x15 would be more reasonable.



However it moves forward ! I arrive to something like the schema below.


Other interesting news: they published more detailed information on the chip programming on TI site. I think that we have now the registry adresses.
Attached Images
File Type: jpg TAS3251.JPG (119.3 KB, 105 views)
  Reply With Quote

Reply


[design log] Neat 2x170W I2S in, I2C controlled, integrated DSP amp (TAS3251)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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Channelmaster 6653 (Rebadged NEAT P68H Idler Drive Table & Neat GA-12 Tonearm spaceistheplace Swap Meet 0 2nd September 2017 07:19 PM
New Design: remote controlled DAC MatM Digital Source 1 18th December 2012 04:41 PM
I ned a 6ch digitally-controlled attenuator. Anyone know of a design? Spasticteapot Digital Line Level 2 27th February 2009 08:17 PM
A rotary encoder controlled integrated chip amp LM3875 justblair Chip Amps 19 31st July 2007 06:16 PM
Wanted : design PCB for relay controlled volume ElectroNick Pass Labs 135 6th July 2004 07:50 PM


New To Site? Need Help?

All times are GMT. The time now is 11:38 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki