Tiny 6x300W amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
When i looked at my old TAS5630 board i tought of putting all 3 amps on to a single board and it worked.

This is the result a 6x300W amp on only a 100x160 single layer board. For the power for all the amps will be distributed off a big busbar going on the bottom on the board (Shown in blue) As i find that the best way of feeding the power to all of the powerful amps as the total power is a ridiculous 1800W (over 35A from 50V rail). Even tho i never plan to run it even close to that (Its mainly headroom and to keep it cold) I like to make what i design run well at top specs even tho i dont need it to.

Mounting holes for the board and heat sink will come later.

So what do you guys think about the board? See space for improvements?
 

Attachments

  • 6ch amp.png
    6ch amp.png
    57.3 KB · Views: 1,140
I didn't know this chip and just had a short look at its datasheet; if you look at the THD vs. power plot you'll find that a more realistic rating is 100W for this amp. The huge power figures are quoted at 10% THD which is only a number and has no connection to reality (that clipping can fry your tweeters). More reasonable is 1% THD IMHO.

So you don't have to worry too much about your bus bar ;)

Certainly clean wiring is always a good thing.

Have fun, Hannes
 
INDuctors

hiii
sexy design i would say... hmmm i m interested in CLass d.. Got a few samples... not the same chip though... its tooooooooo high power for me .. ( i love it but >... might get kicked by the landlord)
anyways i have a small baby TPA3122 & TPA3123... i not sure as to how to make the Inductors... i Dont have the LCR mEtter in BOmbay - inida... it a rare comodity in this country...... it would be a great help if you could help me in making 22uH / 4amps at 1khz i do have some ring ferrite core.. all would need how many turns and what diameter of wier to use ....
thanks
 
inductor

hii
thanks for that tip. but i m concerend of the wire gauge... and secondly i read some where that the inductors affect the output as in its level... if its not right value you wont get its output to drve teh speaker to actuall power...
how true is it
 
Well id use something like 1-2 mm2 wire on it. Oh and the thing is if you make a too big inductor its going to attenuate the high freqencys from the audio signal and that is a bad thing, A too small inductor will make a high idle current and dissipate more idle power in to the speaker while producing a lot of EMI.
 
I've noticed this chip too and plan on building 2xPBTL to see how it performs. It'll be interesting to see how it distorts and how it sounds under reasonable limits.

I'm really drawn to how efficient it claims to be and of course how easy it appears to be to implement. An easy way for people to experiment with active loudspeakers? ;)
 
Just another Moderator
Joined 2003
Paid Member
Perhaps a drawing will assist. This is how I am visualising mounting a heatsink. for something capable of 300W I would think it should be external and not inside the case. in order to mount to it easily the chips need to be near the edge of the board.

The rectangles on the left represent how I interpret the orientation of the chips on your board. I can't for the life of me see how you are going to mount these to a single heatsink ;) Maybe I'm missing something but the drawing on the right is how I would do the mounting :)

edit: Note you could bend the chip legs 90 degrees and use a heatsink with a shelf which would mean that they didn't need to be quite as close to the edge (but still oriented as per my diagram on the right). in this case the chips themselves would be mounted horizontally rather than vertically with respect to the board.

Tony.
 

Attachments

  • heatsink.png
    heatsink.png
    2.9 KB · Views: 337
The distortion goes down because thats THD+N. This D-class has pretty much no THD distortion, what your seing on that graph is the noise of the amplifier. Because of the switching frequency they create inaudible noise at 300Khz that you see here it can be seen very well because the THD is near nothing.


Oh and wintermute you are probably thinking this is a LM3886 like package. Well if you take a look at the datasheet its simply a SMD chip with a thermal pad on top and needs to be mounted to the heatsink pretty much like a Pentium CPU so to cool it all i do is screw a heatsink against the board and its done as the heatsink is then pressing against the thermal pad on the top of the chip. Also i dont plan to stick the heatsink out the case because these chips just dont get that hot. I ran the test board in to a string of big ceramic power resistors with a bass test, while the resistors ware smoking and burning my table the amp was barely warm. (Oh and let me tell you burning the high temperature ceramic smells really bad)

Oh and as for a kit i guess it would have to be offered preambled then too because most audio hobbyists never soldered a SMD before let alone this thing that has a 0,65mm pin spacing. Not that its that hard to solder these SMDs but you need at least a little SMD experience you you dont mess it up. All you need is a soldering iron,solder sucker and some solder wick(or solder braid whatever you call it) Because the tiny 0,65mm pins are too small to go pin by pin i just drag the iron over the whole side of the chip and just suck up the exsesive solder with the pump and then the solder wick. (I desolder SMDs in a similar way)

EDIT:
For those that are interested here is the datasheet for the chip
Datasheet: http://focus.ti.com/lit/ds/symlink/tas5630.pdf
 
Just another Moderator
Joined 2003
Paid Member
Ahhh that explains it!! sorry for confusing the issue, a look at the datasheet cleared things up :) I was assuming it was in a SIL type package yes a bit like an LM3886, though I realised while I was at work this afternoon (brain was obviously thinking about non-work stuff) were 7 pads on the pcb for the chip, was not enough for two channels and I was thinking I needed to check the data sheet!

Tony.
 
7 pads?.....Oh so you confused the input connector with a chip. Yeah i agree if those ware chips it would be the worst position to put them in.

This chip being SMD is a nice advantage since the big capacitors and all that are on the other side of the board where they are not in the way of the heatsink.

I love that these chips are so simple yet so powerful. This D class chip is almost as simple to wire up as a usual AB-class chips but dish out heaps of power while heating much less.Without the B class distortion.(This amazing chip just came out recently)

But i can imagine the tiny SMD package puts off a lot of hobbyists.Since building DIY amps is mostly all about trough hole components.
 
Because of the switching frequency they create inaudible noise at 300Khz that you see here it can be seen very well because the THD is near nothing.

THD+N at 1kHz usually uses a bandpass from about 100Hz to 20kHz and as these guys love to show nice numbers (see output power ;) ) they certainly used it. So, no 300kHz hash messing up your THD+N figures.

Have fun, Hannes
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.