TDA1519B practice amp

[Brymus]

UPDATE the problems were solved and soon I will update with schematics and finished build pics,I hope this thread helps someone else.

UPDATE here is the latest schematic working great so far.
http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519B_schematic.jpg
NEW SCHEMATIC COMING MUCH BETTER
CAB COMING SOON!

Here are some sound clips they sound like they were compressed for the server,if anyone wants the original MP3s PM me
http://vibecat.com/brymus

Hi all,I need some help I recently got some tda1519b chips and decided to make a guitar amp with one.
I followed the schematic in the application notes for the BTL at 12watts and simply replaced the LM386 in my Ruby type amp with the tda1519b using the right pins and a small heat sink.
I am using a 12V supply with a switch between pins 7 and 8.
The pre amp is a TubeScreamer topology using a NE5532 and a few FETS it sounds great going into the lm386 but the tda1519b not so much.
Also this thing is getting HOT I mean REALLY HOT!
I have a 8 ohm load and after a minute or so I am shutting it off cause of the excessive heat.I am using a fairly good size heat sink but maybe not enough.

I am powering the amp with 8 AA nihm cells putting out just under 12V
And for testing I am using two 16ohm 5watt spkrs in parallel from an old TV.

After almost 5hrs of googling this http://munkey.hopto.org/projectsdir/audiodir/simpleamp.php  and 2 TV schematics are the only schematics I can find.
I have it like this but its not right can someone show me what I am doing wrong?
I have been trying all day to make this work without any luck.
Is my input V or my input Z wrong? If it works great with the 386 chip?

[polo16mi]

Hi there:

I had been looking around at the link that you posted, and checked the TDA1519B datasheet too. If you look the datasheet and compare it with the circuit of the link, you will note that the switch is between the 7 and 8 pins, and the wire goes from 7 to the C1 and C2 caps,  NOT from the 8 to caps. Maybe it is your issue. Note that in the site of your link posted above  says that " This is the schematic for mono... (don't go by this schematic here, its wrong. Click on it and use the bigger one)"  If you go with the bigger at the linked one its the same that the datasheet.


Here the link of the datasheet http://www.datasheetcatalog.org/datasheet/philips/TDA1519B_CNV_2.pdf
Check page 8.

Let us know how its goes.

Regards.

[J M Fahey]

Hi Brymus.
Excuse me but I just don´t like any of those "TV optimized" chips.
I see them as simplified to hell, where saving a 1 cent resistor means U$S 10.000 (20 years salary in China) in 1000000 TV sets made monthly. Their no parts count makes them attractive to beginners, but ultimately frustrating, because they lack important things like any sensible gain adjustment, and more important: stabilizing stuff such as Zobel networks and the like.
That´s why your humble LM386 is better.
I´d suggest using an old LM383/TDA2003 , or, if you go power crazy, a tried and true TDA2005 which, by the way, provides quite higher power and efficiency with your 12V supply.
PS: their using just *one* 1000uF to decouple *two* amplifiers/speakers almost made me sick. Disgusting!!
What´s next?: making undersized PCB holes so the component pins fit with some pressure, to avoid soldering ? Give me a break!!

[Brymus]

Hi thanks for the replys.
I have the caps on the wire to pin 7 I have it wired exactly like the data sheet.
The only difference between mine and Munkeys' diagram is I have only one signal + going to pins 1 and 9 and have pins 5 and 2 going to ground.
His "corrected" schematic shows the + signal going to both pins 1 and 9 and pins 2 and 5 as well,surely this is a mistake right?
The amp has to be grounded somewhere doesnt it?

As far as using a different chip ,I already bought these and would like to make this work.
If you look at the picture of Munkeys amp you will see the EXACT same heat sink I am using.
He stated that he added that after the fact,My chip would have already melted had I not had it.
I think it takes less than a minute for it to heat up so much its almost to hot to touch.
(the sink not the chip)
I have checked pin 3 and have exactly half V+ there so I assume thats correct in BTL it shows it not connected to anything.
As it is I get sound but the HEAT makes me shut it down.
Shouldnt this chip shut down by itself if it were to overheat?
Isnt that what they mean by thermal shutdown and overload protection?

I really appreciate any help with this as I am well  :loco after trying all day and night yesterday to no avail.

[Brymus]

OK I triple checked everything Googled for hours,and even found a German forum where they were having the same problem(but I dont read German)so I dont know how they fixed it.
Only the guys amp was heating up super fast (like mine)and the solution involved pins 2 and 5 and the ground being tied to the power? Isnt that just a short?
I was able to translate one page and He smoked several chips ordered 4 more still having the same problem,
He stated it was either bad chips or data sheet error?
Then(on the next page which I couldnt get to translate)he solved the problem very simply,
And said it roared to life and embellished the thread with beer drinking smily's

So I used a much larger heat sink with fins ,now it takes about two minutes to heat up  :loco
I have sound and at one volume its actually pretty good.About halfway its decent tone.
I am using a 50K pot instead of the 100k I had before.
With the volume turned down it makes a horrible hum sorta noise and with the volume up its a terrible screech.

Does this info help? I can tell this amp will be quite loud if I ever getting running right.

On the TV schematics I found they have a 6800pf going from the input to ground what does this do?
Also one schematic had both sides of the RL tied to ground with .1 caps again why?
And shouldnt the spkr have caps on both the +and- terminals since its floating?
Sorry if these seem like silly questions but I am trying to understand this chip and circuit,and really want to get this working.

[tonyharker]

From what you are saying it sounds to me that the chip is oscillating.  You may have the input and output leads or circuitry too close together especially if you are using matrix board.  The 6800pf going to ground on the TV amps may have been put there to stop this.
By the way Pins 2 & 5 are ground.  Ignore the + in the symbols munkey used he has probably just cut and pasted the same ones.

Tony

[Brymus]

OK thanks for that I believe it is oscillating "wildly"
Update I have tried so much its about 50% there,
I increased the heat sink even more- used one off a pentium3 That helped alot,let me realize the spkr had to be decoupled.
The data sheet is wrong I used a 220uf on each side of the spkr this got it going,NO HEAT :tu:
But still oscillating at low and high vloume and not much sound,sounds awful still :trouble.
Tried the 6800pf from input to ground this helped the oscillation at low volume.
Messed around with adding Bourchers cells changing the cap values going to the spkr.
Changing the value of the 6800pf cap,no improvement.
Added a 56K resistor from input to ground fried the first chip,WHY?

Finally just used one cap on the spkr- big volume increase so different values I try.
Anything smaller than 22uf it oscillates at low or no volume.
Anything larger and I start loosing volume but still oscillating at higher volume.
At this point with the 6800pf cap and the 22uf on one side of the spkr it sounds like a decent 3-5 watts it actually gets a nice distortion by uping the gain on my clean channel.
The dirty channel oscillates at all but the lowest gain needed to get the signal through the two clipping diodes in the feedback loop of the 5532 chip.
So where its normally louder its now the quieter of the two and its easier just to use the clean channel with a pedal.
So if I can just get rid of the oscillation at higher volumes this amp will be ready to put in an enclosure.
ANY SUGGESTIONS PLEASE?
I tried running it off an 18V 1.5amp power supply with no improvement over the 8 AA nihm cells.
Should signal ground be separated from power ground with a cap or resistor?
Do I need some kind of RC network between the SPKR leads?

[J M Fahey]

>>"ANY SUGGESTIONS PLEASE?"<<
Yes ! Emphatically !! = Trash that piece of sh*t/junk (your choice)
If even the factory suggestion is bad, what else can be said?
Go for a real power chip: TDA2002/3/4/5/6/30/40/50 and follow factory instructions.
If even the nerdy techno-obssesive Germans have trouble with them !!! what else ?? and no , it´s not a racist/despective comment , quite the contrary.
Even if you get them to work and even sound beautifully, but only if submerged in liquid nitrogen, well, that´s not good engineering practice.

[joecool85]

>>"ANY SUGGESTIONS PLEASE?"<<
Yes ! Emphatically !! = Trash that piece of sh*t/junk (your choice)
If even the factory suggestion is bad, what else can be said?
Go for a real power chip: TDA2002/3/4/5/6/30/40/50 and follow factory instructions.
If even the nerdy techno-obssesive Germans have trouble with them !!! what else ?? and no , it´s not a racist/despective comment , quite the contrary.
Even if you get them to work and even sound beautifully, but only if submerged in liquid nitrogen, well, that´s not good engineering practice.

In all seriousness, a TDA might be a good way to go.  I'd look into it.

[teemuk]

Without seeing the schematic of the exact circuit you used and some photos of wiring and board layout it's kind of difficult to guess what's wrong.
You have AC coupled the volume control potentiometer from the chip's inputs, right?

The BTL circuit doesn't need AC coupling for the speakers since the DC potential at both speaker terminals should be equal and hence there is no DC current flow from one terminal to another. If AC coupling makes this circuit work better it is a clear indication that you have a DC offset issue that is disrupting the proper operation of the chip/circuit.

---

I am hundred percent sure that the chip will work well in the circuit depicted in the datasheet, assuming board layout is done correctly.

[Brymus]

EDIT:Crap I feel so stupid right now :duh
I think Teemuk is right and I think I found the problem as I was getting ready to take pics for
him.
I will let you know as soon as I find the camera,I think the others with the same problems proly did the same thing...The simplest oversight ever...


Thanks again for the input and sticking with me through this guys.
And Teemuk while I hold your opinoin in the HIGHEST regard same as RG and the like.
I am 90% certain that this chip will NOT work with the application circuit from the data sheet.
Please, no disrespect intended I very much appreciate your time helping me with this.
After days googling I found a company called Conrad offers kits with these chips and forums from europe are full of people having the same issues as me.
One guy reported his IC exploded after a days use and was propelled through his enclosure on to the floor where it burnt a hole in his rug his heat sink was twice the size of my Pentium3 sink and was pure copper-not to say he had it wired right he went off the data sheet.BUT STILL EXPLODED?
Its kinda hard to mess up a layout that only involves 3caps-I have succesfully built TrainWreck -Marshall clones  which is much more difficult and layout sensitive than this chip amp.
BUT I am not an engineer just a monkey who can read schematics...And copy others.

YES I have a DC offset issue,BUT why IDK
I will post a schematic of my preamp so you can explain what I am doing wrong.
I fried a second chip sort of,I overloaded the input with my guitar and a single FET now the DC offset makes the chip unusable I have one more left.
Apparently this chip needs a very low input my guitar is enough to drive it.
My best luck so far was with a single 22uf cap on one side of the spkr and a 56n from input to ground this only had two issues,
The Volume was like a tone control (first order filter) so that 90% volume offered the best tone.
The other issue was it still oscillated at low-no volume.
I fixed this by moving the volume pot behind the buffer after my tone control.
This produced a lower output but no oscillation and no tone change with the use of the volume pot.
It was actually pretty decent at this point just not a 10-12 watt output,so I continued.

Every stage of my pre-amp is decoupled with 100-330n caps.And I used a .47uf input cap to the PA.
I also tried my Multi FX staight in through a 1m volume and a FET buffer it has both line-and low level output,even the low level output was overloading it at 3/4 volume.
I am going to go with another chip but would like to learn why I had so much difficulty with DC offset and oscillation.
I also learned its better to have a chip that uses external components for feedback and gain control ,decoupling and the like.

[Brymus]

YEP I FOUND IT :tu:
OK Teemuk you were absolutely right it works perfect from the data sheet circuit.
What you said got me looking for where the circuit wasnt decoupled...

Before I explain let me start by saying I mostly have built tube amps and not solid state and the test area I have set up with my breadboards has been used for lots of pedals,and a few 386 based designs.
That being said see if you can spot the problem from these pics- its as clear as the nose on my face.
I sorta feel like an idiot but then again I understand how this happened.
PICS  
(1) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild034.jpg

(2) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild032.jpg

(3) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild028.jpg

(4) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild002.jpg

(5) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild008.jpg

(6) http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild010.jpg



Finally here was the OH so simple solution its the reason I couldnt bridge my 386 chips either,
I thought it was because I was using the high output versions and no heat sink on the DIP8 chips.
Gosh the simplest thing can make a big difference

http://i419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/12WattTDA1519Bbuild021.jpg


The whole folder so far is here I will update it with schematics and pics of the finished cabinet when its finished
http://s419.photobucket.com/albums/pp276/Bry928/SolidState%20Amps/12Watt-TDA1519B%20build/
Really thanks for the effort guys at least its fixed with my 3rd and last chip.
This is my second chip amp build(I have built several 386 amps though)

Hey Teemuk do you think the signal ground should be tied to power ground?
Thats the way I have it now with the signal ground on one bus and the power ground on another and tied together at the pins of the IC.
Also now that its fixed do you think my 386 amp would be OK as the pre instead of taking it from the 386s preamp?
I will post a schematic for this in case anyone else using this chip wants to copy my build.
It sounds fantastic REALLY I am impressed- I got these chips at Apex Jr for .79 cents each.
I am glad I didnt give up it was worth the effort.
I just wish it had more components the and ability to adjust the circuit,like feedback and gain.

Now I have the confidence to start my lm3886 build and found what would have been a problem for that circuit as well before ruining one of those chips or worse.

[Brymus]

HINT the answer and the reason the ICs werent working right is in the fourth picture down

[teemuk]

Let me guess, the two jacks below are speaker jacks and their sleeves short to chassis/ground?

[teemuk]

Hey Teemuk do you think the signal ground should be tied to power ground?
Thats the way I have it now with the signal ground on one bus and the power ground on another and tied together at the pins of the IC.

Yes, they need to be tied together (after all, both pins need a ground reference) – but preferably at one point only. I would try tying them together at the main ground point in the power supply (likely the grounded pin of the main filter cap), but if the pins of the IC work just as well then it's fine. It's all about finding the best layout anyway and sometimes it happens by experimenting a little.

The thing is mostly a noise issue; preventing power supply ground return currents from inflicting the signal ground return currents. They all return to main ground point so joining the two separate ground returns together at that point might be the least-noisiest configuration (think star grounding principle).

Also now that its fixed do you think my 386 amp would be OK as the pre instead of taking it from the 386s preamp?

By adding the LM386 there you'll be increasing the overdrive originated in the preamp. The Ruby's preamp is just a clean buffer that prevents the low input impedance of the IC from loading down the guitar. Its gain is less than unity. If you already have a tubescreamer –type circuit preceding it, it introduces all the buffering you'll need and the FET buffer can be ditched.

You can try adding the LM386 there and see if you like the result. The LM386 will likely need some attenuation at its output to prevent overdriving the following TDA1519.