|
|
Post by jcgss77 on Nov 25, 2010 22:43:40 GMT -5
Amp gurus, I have got a question. I recently picked up some TDA2050's and wanna make an amp. In my search for information in building one, I read a couple articles about bridging them, as I have 2 8 ohm 30 watt speakers. The only problem is I have found little on doing that. I did find this schematic: i744.photobucket.com/albums/xx81/jcgss77/tda2050Circuit.gif?t=1290743101Does this seem stable running at 4 ohms, or am I expecting too much out of those little guys? I don't plan on running it at the full recommended voltage per the datasheet, but I do desire around 40-50 watts into that load. If this schematic is stable, what kind of voltage do I want to power this amp with? Are there any other considerations I need to think about with this design if it works? Thanks for your answers! Sorry, but I couldn't get the picture to come up the normal way. Don't know if something is wrong with the technique or if something is just wrong with me. |
|
|
|
Post by sumgai on Nov 26, 2010 1:15:56 GMT -5
jc,
Start by getting your hands on the datasheet. What you want to make sure of is that you have enough heat-sink area to dissipate the heat your device is gonna throw off. Even at less-than-maximum voltage from the power supply, you'll still end up with a brick if you don't plan for shedding that heat.
Next, I'd skip on that particular schematic, it's full of holes, IMO. Find something a little more mundane, and work from there. I might suggest a search for common middle- or lower-end hi-fi stuff, or perhaps car audio.... this series of amps (TDA20xx) has been around for quite awhile, and lots of companies have used it at one time or another.
Good luck!
sumgai
|
|
|
|
Post by jcgss77 on Nov 26, 2010 10:31:40 GMT -5
Thanks sumgai! It is funny you mention car audio amps-I am going to be building this build into a recycled car amp, so I know there is a LOT of heat dissipation capability. I know that there are lots of holes in that schem, as pretty much all of the power supply leads are omitted, but I wanted a base to start. I will design something which will be kind of a hybrid of the datasheet layout and my Crate G-40C, which runs around tda2040's. I'll post that layout soon for your viewing pleasure!
|
|
|
|
Post by jcgss77 on Nov 26, 2010 18:29:27 GMT -5
Okay, after more research, the best I could come up with is the bridge schematic for TDA2040 chips. Since they are pretty much pin-for-pin, I figure with some component value tweaking this would work. However, the schematic calls for a split power supply, and I don't have one. I was planning on using a printer psu. Can I just tie the V- to the negative at the psu output jack to keep it separated?  I have also read that when bridging these chips, you don't get your proper output current. Would it be a good idea to add a couple of transistors with that function to the output, and if so, would you suggest a particular one? |
|
|
|
Post by sumgai on Nov 27, 2010 2:53:42 GMT -5
Hmmmm...... Long dissertation.....This circuit will work, within it's few limitations. Nice and simple, but the device itself has a pretty wide tolerance of circuit variables, so I think it'll perform as you intended. As to the issue of single-ended versus split power supplies, let me refer you to the datasheet. There are sample schematics for both kinds on there, I'll leave it up to you to make the appropriate adaptations. Of course I could do this for you, but where's adventure in that, eh? Not to mention that it would also knock a serious hole in your "Nutz-ness" rating, wouldn't it?   TDA2050 DatasheetHTH sumgai p.s. Pleased define what you mean by "proper output current", that one has me scratchin' my head.... |
|
|
|
Post by jcgss77 on Nov 27, 2010 11:20:42 GMT -5
sorry...by "proper output current" I meant milking it for all its worth in regards to driving the speakers. I heard that bridging these chips doesn't result in as much current driving them than you can get, and that adding some more power by the means of output transistors after the chips would increase that, getting a better response from the speakers. Or am I just showing how much I don't know about amps yet?  BTW, what are the limitations you notice in this circuit? |
|
|
|
Post by sumgai on Nov 28, 2010 5:52:54 GMT -5
sorry...by "proper output current" I meant milking it for all its worth in regards to driving the speakers. I would've thouht that, but I wanted to make sure. Oh? And the obvious implication here is, bridging some other amp chips does yield "full power" to the speakers? Or is the statement meant to imply that no matter what kind of device is used for bridging, one will never achieve "full power"? In either case, my BS detector has sprung a leak of massive proportions. Further elicidation may be forthcoming.... in the near future..... to a computer near you..... if I ever get off my lazy butt!Errr, excuse me, but wasn't an increase in power the reason to bridge the two chips in the first place? And you did choose that particular chip for some reason, no? Cheap, simple, sounds, good, I don't care what the reason, the fact is, you've already chosen them. But I suppose it's not too late to take off in another direction..... Suffice it to say, you're not gonna find entrance to HappyLand ® by trying to drive output transistors with a chip designed to drive a speaker. It can be done, but at great effort and expense, and in the end, you'd only end up poorer in the wallet, with no real gain to show for it. Bad ju-ju, I recommend against it. Better as in more power, or better as in Mojo, i.e. Tone? Adding more power won't give more tone, unless you're driving the speaker(s) to some condition of overload that sounds good to you (and hopefully to your audience). But the trade off is (usually) a shorter life-span for at least the speaker(s), if not for the output stages too. This is one of those grey areas where a lot of choices have to be balanced against each other, and in the end it'll be a compromise that only you can settle on. We can help, but that's all we can do, I'm sure you understand. Well, don't go off half-cocked, thinking that I know something about amps!! We're all learning here, and I'm no exception. (But admittedly, I started out on my learning curve a few years ahead of you.  Where to start? The only thing I don't particularly care for here are those two cap/resistor combos hanging off of each chip's output - R8/C8 and R9/C9. They are meant to prevent parasitic oscillations from destroying the chip, but in point of fact, they are set to values that I think will load down the chip's output. I mean, come on, 0.1µf, that'll pass frequencies down to approximately 1KHz, give or take a decade or two. I think that a cap valued at 1/10th that would be a more likely candidate to give you some marbles*, right where it counts. HTH sumgai * Reference the film "Major League Baseball II". |
|
|
|
Post by jcgss77 on Nov 28, 2010 10:20:20 GMT -5
Point well taken. I feel that this whole hubbub over bridging the chips will yield little extra power in comparison to running 2 chips in a stereo format. So, with all due respect, I am backing off of the bridging endeavor, and will save the dual amps for when I attain a suitable transformer. Until then, I have put together a simple (very simple) amp, consisting of an AMZ Mini Booster (I have 30 Jfets laying around that I need to do something with), a Baxandall tone stack, and a single TDA2050 poweramp. The schematics are all pretty much true to the original author's works, with a few additional pots.  Uploaded with ImageShack.usAre VR5 and VR6 overkill, will they have a noticeable effect on the sound that will warrant them being in the circuit? More importantly, are they wired correctly in this layout? |
|
|
|
Post by sumgai on Nov 29, 2010 3:30:52 GMT -5
jc, Yes, you do need those two pots in there. If you just replace them with fixed resistors, you'll have no gain at all from the circuit. IOW, you'll end up setting them to some value besides 'full over to the stop'. Besides, after diddling around a bit, you might find a setting that suddenly brings forth The Mojo ™, and that ain't a bad thing atall. HTH sumgai |
|
|
|
Post by jcgss77 on Nov 29, 2010 19:07:31 GMT -5
OOOHHHH yeah, I want some of The Mojo™. Once I find it, I will be sure to leave credit to the TradeMark holder on the schematic. The combination of C11 and R17-is this what is called a Zobel Network? As you stated about the above schematic, that the capacitor value should be about 1/10 of the value. So, if I bring the cap value down to about .047uF, would this work better? Another thing that comes to mind is the low-value resistor. Since the components go to ground, wouldn't I want a higher value resistor to allow less signal to go to ground, and more to the speaker? |
|
