TDA1543 setup (alternate I/V output)

[Rocky]

Folks, I've got a large number of TDA1543 chips I'll use in a power DAC for my tweeter (400Hz+). I've calculated them to be able to deliver around 100mW into a 16ohm load, wich is ample, but I can't really figure out how I should make the output stage.

I want all the mA the DACs can deliver to be sent through my speaker to drive it, and if I use an I/V resistor plus cap coupling, I'll leak all the power over the I/V leaving nothing to drive the speaker with won't I?

Should I make an impedance correction network on the speaker ensuring 16ohm across it at all frequencies (including DC), and use this complete network as most people would use an I/V resistor (see the attached pic)? Measure and tweak to get it linear? Any other suggestions?

My suggested I/V network:

[dddac]

nice idea....

just quick calculation, 100mW in 16 Ohm is resulting in a Voltage (effective) of 1.26 Volt. With the TDA1543 running on 8 Volt this is easy to do. My dac design gives 1.56 Volt RMS at the output

The only question I may have is the current.... at "full" power you need roughly 80mA for the 100mW in 16 Ohm

I know that the maximum effective signal current 0dB signal coming from a TDA1543 is 720 uA ...

You will need therefore 80mA/720uA = 110 DAC chips

These 110 DAC Chips will consume like 6 Ampere 50 Watt power disapation...
Next question, how are you planning to drive 110 parallel TTL inputs ?

If you all do this, great stuff, not easy, I know... Just built a prototype with 60 DAC's in parallel.... So basically you need 4 times this ...

The idea of imepdance correction is kind of the right things if you don not want to go for 1000 DAC chips , never the less I am not sure how the dac reacts when the speaker unit is unloading current when it swings back after a pulse signal. the output impedance of the dac is pretty high...

If 1.2 Volt is enough for you, why not make a very simple MOSFET source follower ? than you just work with a few dacs, improving the sound and drive the tweeter strongly.

Any way, it sounds very interesting and it would be fun to try !!!!


regards
doede

[lohk]

Rocky,

if you have so many - please send me one!
It seems difficult to get one only.

Klaus

[John Swenson]

I just did the calculations and came up with 93 chips needed for 100mW into 16 ohms, I'm not quite sure why I got a different number than Doede, but even so thats a LOT of DAC chips! Driving all these and keeping the jitter low between them is going to be tough as is the PS and getting rid of all that heat.

I second Doede's suggestion, I use a cascoded mosfet follower to drive a tweeter and it works very well (great for headphones too).

John S.

[Rocky]

Thanks Doede, John

I'll probably go the MOSFET route then, you are suggesting something like the Headwize project at This HeadWize article ? Any recommendation on wich transistor that'll provide the best sound in this simple circuit?

I'd also like some recommendations on how to do the DAC and it's O/P stage. I can use as many TDA1543s as I like, I have a wide range of quality capacitors, I've got microphone transformers too (1+1:1+1 Lundahl) if that'll improve anything...

Regarding a chip for you, Klaus, I'll probable be able to spot you one. My email is down at the moment, try it in a day or so.

[Rocky]

Something along these lines?

[lohk]

Rocky,

what happened to your email?

[Rocky]

My mail server had a change of ISP. It became a mess, but it's all sorted out now.

[lohk]

Your mail adress is not showing up here - pls send me your mail adress to contact you directly.

thx
Klaus