What can I make with these?

[teemuk]

I wonder if Marshall is the only mfg that can't get those chips to work reliably. The TDA729x has plenty of usage in dozens of guitar amps and active monitor speakers (though Mode Four is the only one I've seen so far that uses the chips in a bridge and slaving mode). Anyway, I haven't heard similar horror stories about the great deal of other products with power amps based on that chip.

[docz]

I stumbled onto this circuit, that uses a TDA2030 and TIP31/TIP32 for a total of 30W, you guys think it will work good together with that preamp circuit I found?

http://www.elecfree.com/electronic/wp-content/uploads/2008/06/circuit-30w-amp-otl-by-tda2030-transistor.jpg

And how do I know what sort of load impedance the power amp can handle?

DocZ

[Minion]

The schematic it"s self shows a 4 ohm load so you can run at least 4 ohm and higher loads and maybe lower as the output transistors are Paralell to the TDA2030 which should allow you run lower loads but don"t take my word for it.....

[docz]

That electrolytic cap that is on the positive lead to the speaker, what is it's purpose? Is it just to block DC? Can I get away with different values?

DocZ

[Minion]

That amp uses a single Supply PSU so half of the DC Voltage will be on the speaker output.... Yes , It blocks DC , You could use a Different value but a Lower Value will Give you worse Frequency responce and a Bigger one will give you a better frequency responce so if you do use a different Value I suggest you use a Larger value .....

[docz]

Ok I get it, what about C1, that goes between the two input resistors? Is that some kind of filter? Again how will different values effect it?


DocZ

[J M Fahey]

Hi Teemu.
I guess "using them within their limits" is the key.
And bridging plus slaving is *not* a failsafe architecture.
*One* that fails, overstresses the parallel one, which has to bear full current,twice what was expected.
Seconds or minutes later, it burns/shorts, leaving the opposite pair trying to carry the full load on their own.
Couple that to the 32 flat wire connections needed working flawlessly ... and it becomes a ticking time bomb.
Besides, it isn't even cost-effective.
4 x IRFP250 can supply the same power in a more reliable way.
Besides, you can easily use 6 devices to better share the load, which is not possible with the TDA's.
I'm drawing a board to make 60W mini-cube Guitar and Bass amps with one TDA7294 each, and a latching (SCR) protection that mutes them whenever something strange happens, and in that use I have faith on them.
Anyway, only field use will tell.

[teemuk]

Ok I get it, what about C1, that goes between the two input resistors? Is that some kind of filter? Again how will different values effect it?

I suppose you mean C2?

Here's how it works:

To get optimal signal swing at poth postive and negative half waves, an opamp circuit must have its both inputs and the output biased to a DC voltage potential that is exactly half of the (power supply) rail-to-rail voltage. Think of that voltage potential as the zero crossing point of the output signal. The output will bias itself automatically because it is driven by the inputs.

If the opamp is powered from a "dual supply" that uses positive and negative supply rail voltage the required bias voltage potential is their difference, that is (usually) zero volts. This is a useful configuration because you don't have to AC couple the outputs and the inputs to block the DC bias voltage (since that voltage is practically zero).

In this case, however, the opamp driver is powered from a single ended supply. The two supply rails in this case are the positive one and the ground. Since the same rule about biasing still apples the opamp needs to have a bias DC voltage that is exactly half of the positive rail voltage.

Thus we come to the voltage divider consiting of R1 and R2. This, as you can see, will produce a required DC voltage potential for biasing. Exactly half of the supply rail voltage. The input of the opamp is then referenced to this DC potential with resistor R3. C2 is indeed a filter to keep the potential at steady DC.

The bigger value cap you use there the steadier the bias voltage will be and the less distortion there will be at low frequencies. The drawback is that the resulting RC filter has a higher time constant and it will take a longer time for the operating voltages to settle.

Due to prevailing DC potentials in inputs and output we also need to AC couple this sort of design. C1 is the coupling cap for the input and C8 is the one for output. As discussed, they also form RC filters with related resistors / impedances.

C4 is another coupling cap. It's for the other input (which is used for feedback loop). If that cap was omitted this opamp input would also need a reference to the biasing DC voltage. However, we can AC couple one of the inputs because biasing the other one is enough. One reference will drive the other input as well as the output to proper bias potential (assuming those terminals are properly AC coupled that is). Since C4 will also limit low frequency gain it's usually found even from designs that wouldn't "need" it.

---

If the TDA2030 can work from a dual rail you can simplify the circuit a lot because you can remove many of the coupling caps and the biasing network and simply bias the opamp with ground reference. I belive the datasheet of that chip has some examples about this stuff.

[docz]

Thank you very much for a great explanation! I get it C2 and the coupling caps now. I thought that was a voltage divider, but I was not sure. This is all pretty new to me. From what I read from the datasheet, this chip get's it's negative voltage on pin 3 and positive voltage on pin 5. But I can't seem to find a voltage divider there, is that what Q1 and Q2 do invert the power? The dual rail circuit in the datasheet is much simpler - and easier for me to understand. Could you guys point me in the direction I need to include those two transistors for the double boost in power?

DocZ

[teemuk]

I think these diagrams should clarify things quite a bit.

Note that I omitted some parts to show only the essentials and make the circuit diagrams simple. These parts are still required in the actual amp.



What are those other parts, then:

- C5: a main filtering cap for the supply rail. If you use dual rail, each rail will require its own but you can use half the capacitance (per rail) that is required by a single rail supply.
- C3: another filtering for the supply rail. This one is used for a fast response to momentary current draw. Due to layout issues the main filter cap C5 is usually quite a bit further from the transistors and the chip drawing current. Impedance and inductance of the wiring will thus become an issue and this can make the circuit unstable. C3 is placed physically as close as possible to the components drawing current to minimise the effect of wire inductance and resistance. Again, if dual supply is used you will need two of these. One per rail.
- R8 + C7: These are a so-called "Zobel network". They will compensate the speaker's inductance that icreases at higher frequencies. The network makes the amplifier operate more stable at difficult speaker loads.
- D1 & D2: These diodes protect from inductive voltage peaks by clipping the output voltage if it exceeds the supply rail potential. This will prevent voltage peaks from reverse-biasing and damaging the transistors or the chip. Note that these diodes actually go to supply rails. If dual supply is used the other one is connected to negative rail instead of ground.
- C6: this capacitor ensures that both transistors have an equal AC voltage driving them.

[docz]

Wov thank you very much teemuk for a very insightful explanation!

I will start to prototype this thing as soon as I get my desktop cleaned out. It is absolutely flooded with half-way-done projects

I think this will be great!

DocZ

[docz]

Just to clearify things, if you don't mind me asking that is. How does the values of the caps affect the different functions?

DocZ