Looks like fun - I've just ordered up some TDA1545 (original, not 'A' version, which have I2S interface) for my next DAC project.
Try single transistor I/V, you won't look back when against passive I/V.
Try single transistor I/V, you won't look back when against passive I/V.
Hello abraxalito,
i read your respons in other threads and that was a stimulation for me to try the TDA1387.
In the past i also had TDA1541, TDA1545, Several AK and CS Dac's but the BB pcm1792 is the one i used the longest. Until a view weeks ago, i use now no dac at all. Just listen DSD.
I have already a favourit output stage but in this thread i do not like to discus that. Just see how to split the data stream and using TDA1387.
i read your respons in other threads and that was a stimulation for me to try the TDA1387.
In the past i also had TDA1541, TDA1545, Several AK and CS Dac's but the BB pcm1792 is the one i used the longest. Until a view weeks ago, i use now no dac at all. Just listen DSD.
I have already a favourit output stage but in this thread i do not like to discus that. Just see how to split the data stream and using TDA1387.
Good to see more designs using the TDA1387 🙂 You'll find that ceramics on pin7 don't sound as good as largeish electrolytics, particularly in the bass.
Does that mean you're giving up on the tda1387 8x?
Koifarm, your board idea is interesting. I want to do the same, but I've never designed any PCB, so there's a big learning curve standing in my way. 🙂
But, if I ever get around to it, some of the ideas I had were:
- 8x dac chips that, with jumpers, could be configured all in parallel, a single 4x group, or two 4x groups with one inverted
- space for each dac chip to have a big electrolytic for pin7, and maybe also for pin5
- i2s input
- configurable output - simple passive I/V, or Abaxalito's design from the 8x thread, or jumpers to some off-board solution (e.g. tubes)
- a switch for enabling/disabling the NOS droop filter
It remains to be seen if my determination to actually design this is as great as my aspirations. 😉
I'm intending to use the existing PCB as the basis for the new design, just won't be using any of the TDA1387s in it. Using the existing digital (and power supply) sections is the reason for using the TDA1545 and not the much more common TDA1545A (which isn't I2S). I'd like to develop a robust version of the mods (the DACs will move to the prototyping board) which can more likely survive shipping and partly because I want to prototype a lower power version which can be turned into a portable DAC. The TDA1545s are about half the power of the 1387s and in theory at least can deliver twice the current. So that's a 4-fold potential advantage.
I will try that at a later point. On the other side from the pcb there is space enough.
I use a CS8416 board for de input, just €12 on Ebay.
Attachments
Last edited:
My PCB has almost all the possibillities you mentioned.
Below a picture of the dirty datasplitter with 74ahc00 nands.
2hz = WS, 10hz = Data input.
Attachments
Last edited:
You can't have a single supply DAC with a bipolar output around zero. Or did I misunderstand your question?
i mean the output from 1387.🙂
sorry english not my first language.
does it have dc offset, does it not?
sorry english not my first language.
does it have dc offset, does it not?
Its a current output DAC so the 'offset' is in current and yes it has one. The offset current is about 580uA from the VDD (positive) supply at digital zero.
Do you mean this one - http://www.diyaudio.com/forums/digital-source/227677-using-ad844-i-v-21.html#post4408571 ?
If so no, because that'll only give up to 20uA offset either side of zero, but for the TDA1387 the offset current's more than 20X bigger. If I were you I'd use a 580uA current source (2 transistor type that I show in the thread on the TDA1387*8 DAC).
If so no, because that'll only give up to 20uA offset either side of zero, but for the TDA1387 the offset current's more than 20X bigger. If I were you I'd use a 580uA current source (2 transistor type that I show in the thread on the TDA1387*8 DAC).
I do it the simple way. After the I/V resistor, just a tube stage without the coppling capacitor. Workingpoint tubestage compensated with the bias voltage.
Ronny
- Status
- Not open for further replies.