modify D1 stage for 0.1mA DAC?

[promitheus]

those would be very good speakers and very good hearing

 

[Alien8]

Hmm you're probably right about that, but still....

It is still high frequency noise, so everything behind it will have to cope with it....

 

[promitheus]

well about the noise you are right, it depends on what kind of DAC you are using.
If you have the normal 44,1KHz Output you should use a 20KHz filter of second order.
If you have an upsampler you only need a simple filter from 30-40 KHz for a normal cd

 

[Klaus]

Re: Filter order

Hi Klaus (and others)

Just a short question.... maybe a bit unrelated but anyhow.

You mentioned a simple first order filter (RC) at 55 kHz after the I/V.
Don't you have problems due to the fact that it's a oversampling dac, so 'just' a 1 st order filter isn't enough ?

Or am I just talking rubbish here anyway.....

thx

Hi,

I can just state: listen to it - you will no longer even think about any "problem" with it. The sound is so much better, no words can cover this...

Regards

Klaus

 

[Alien8]

Hear Hear.

Maybe we should stop thinking and just listen (and not use a scope... )
Oh, while I'm at it...

The TDA1547 has a symmetrical output, so that means TWO times your schematic (on the german forum) PER CHANNEL... ?

You seem to have a balanced system, I don't so how do I do things ?
Use a discrete 'opamp' to sum things together, or just use one side ?

I'm NOT an expert... please consider

thx

 

[Klaus]

Hear Hear.

Maybe we should stop thinking and just listen (and not use a scope... )
Oh, while I'm at it...

The TDA1547 has a symmetrical output, so that means TWO times your schematic (on the german forum) PER CHANNEL... ?

You seem to have a balanced system, I don't so how do I do things ?
Use a discrete 'opamp' to sum things together, or just use one side ?

I'm NOT an expert... please consider

thx

Hi again,

yes, be sure, I hear and decide by listening test. To frequently hearing and measuring do not correlate.

To be full symmetrical and full consequent of cause means to have four times the D1 clone output branch (+ and – for left and right). But of cause it could be reduced to just unbalanced = two branches. But with this also the genius sound neutral volume control function via resistors between the branches is lost. I have the Aleph P1.7 and Aleph4 mono blocks as well, so I am using full symmetrical. And by the way, just with symmetrical signal transport real fun begins. It is like having 20dB more of dynamic range without this overlaid ground noise.

Before changing to the D1 clone solution I was very happy unsymmetrical with just upgrading the OP buffer to OPA637 (IMHO with much distance the best OP!), the TDA1547 PSU and get rid of any output filter beneath this passive low pass first order as said. In listening tests it really shocked some guys. So dynamic, focused and live-like music can sound. So I would recommend you to go for such an easier solution as a first step. Do you need information? Do you understand German text?

Regards

Klaus

 

[Alien8]

*snip* so I am using full symmetrical. And by the way, just with symmetrical signal transport real fun begins. It is like having 20dB more of dynamic range without this overlaid ground noise.

Hmm, I've read a lot of interesting stuff about symmetrical, but for now I'm stuck with 'consumer' audio toys so no symmetrical connections YET

Before changing to the D1 clone solution I was very happy unsymmetrical with just upgrading the OP buffer to OPA637 (IMHO with much distance the best OP!), the TDA1547 PSU and get rid of any output filter beneath this passive low pass first order as said. In listening tests it really shocked some guys. So dynamic, focused and live-like music can sound. So I would recommend you to go for such an easier solution as a first step.

Well, that's just one intermediate step. Thanks for the input. I've just began undertaking this whole modding thing. BTW my 'toy' is a Philips CD950. I'm mostly in the process of gathering input from more experienced people. Have done some minor things already. Right now I'm comparing opamps in the output stage.
It's just that I saw some people complaining about the 1547 because it uses the on chip opamps for the I/V conversion, so one idea I had (well the idea isn't probably mine anyway) was to bypass that. Which is exactly what you have succeeded in, hence my interest.

Do you need information? Do you understand German text?

Well as I said before, I'm a complete 'learned it on my own' guy, with no education in electronics, so any help is highly appreciated. I'd like to understand what's going on with these things. Maybe some of the pro's could have a look at what I'm coming up with, just so that the whole thing doesn't burst into flames. Und jawohl, Ich spreche ein wenig Deutsches. (had this run through an on line translator so it wasn't completely rubbish )

 

[Klaus]

Hmm, I've read a lot of interesting stuff about symmetrical, but for now I'm stuck with 'consumer' audio toys so no symmetrical connections YET
)

--- with Aleph DIY you get real symmetry…

Well, that's just one intermediate step. Thanks for the input. I've just began undertaking this whole modding thing. BTW my 'toy' is a Philips CD950. I'm mostly in the process of gathering input from more experienced people. Have done some minor things already. Right now I'm comparing opamps in the output stage.
It's just that I saw some people complaining about the 1547 because it uses the on chip opamps for the I/V conversion, so one idea I had (well the idea isn't probably mine anyway) was to bypass that. Which is exactly what you have succeeded in, hence my interest.
)

--- it seams that you are on the perfect right way! You gather without any prejudices. So many engineers should also act so… I think the OPAs in the TDA1547 are good enough to reach an already high performance level – if the TDA PSU is tweaked well. Bypassing means to deactivate these internal OPAs and go external. But here really hardcore tweaks begin. And if bypassing, than I propose the ultimate solution like Pass D1 etc. But extra PSU with higher, symmetrical voltages requested.
So my final statement: if the rest of the hifi set is really on a very high tweaking level, than start to care about the bad internal OPAs. If not, there are other things to start. Then pick up the output signal after the first external summing OPa and give it to the output just with a first order low pass. Test it, you will be astonished…

I just asked about you German language capabilities because I have a detailed TDA1547 tweaking instruction – but in German…

Regards

Klaus

 

[Blitz]

Could you email me this instruction ?

Thanks a lot

 

[Alien8]

--- it seams that you are on the perfect right way! You gather without any prejudices. So many engineers should also act so…

thank you for the vote of confidence

So my final statement: if the rest of the hifi set is really on a very high tweaking level, than start to care about the bad internal OPAs. If not, there are other things to start. Then pick up the output signal after the first external summing OPa and give it to the output just with a first order low pass. Test it, you will be astonished…

Once again thanks for the tip(s). All changes I did so far did yield an audible improvement over the previous state, although blind testing is not really an option.

I just asked about you German language capabilities because I have a detailed TDA1547 tweaking instruction – but in German…

I should be able to understand this.... so if you could spend a moment.... (my email should be in my profile)

Thank you very much for your help, I'll keep you posted

 

[Klaus]

I should be able to understand this.... so if you could spend a moment.... (my email should be in my profile)

Thank you very much for your help, I'll keep you posted

Hi again,

I just brought together something for Blitz. I send this mail also to you. I will try to find the time to translate it first to English... What is the value of a tip if you would get it wrong, so.

Regards

Klaus

 

[Paulus1981]

Hello Klaus,

I'm the owner of a Marantz CD17, (which, as I'm right, is equipped with a single TDA1547) which I want to upgrade with the D1 board.

The only problem is that I'm an almost complete n00b in this material. I want to use these pcb's :

http://www.diyaudio.com/forums/showthread.php?s=&threadid=30208&perpage=10&highlight=&pagenumber=10

And if possible the parts as proposed by promitheus.

Is there anything I should change?

Thanks in advance!!

 

[Skorpio]

Hi again,

I just brought together something for Blitz. I send this mail also to you. I will try to find the time to translate it first to English... What is the value of a tip if you would get it wrong, so.

Regards

Klaus

If possible, could you email me the material on upgrading the TDA1547?

 

[tkraai]

TDA1547

First of all a Happy Newyear to all of you ;-)

And now my question. I would like to use the DAC R (pins 11 and 10) and DAC L (pins 22 and 23) signals so I can bypass the internal OPA. I read somewere that these current outputs have a maximum of 0,1mA output. Is that correct?
Does anyone know what is the optimal load (resistor) value to create a passive I/V?

In my case the TDA1547 is used in my TEAC VRDS10 and used in dual differential mode. So one TDA for LEFT and one for the RIGHT channel. That gives already 0,2mA per channel. And I plan (TDA's already ordered) to piggy-back the TDA, so I wil have 0,4mA per channel Which still is very low.

I also have a ZAPfilter2 (http://www.lcaudio.com/index.php?page=6) which is capable of I/V conversion, but I don't know if 0,4mA is enough. Can't find any specs.

Another option is to use a step-up transformer. I prepared a little schema using a Lundahl LL1636 (http://www.lundahl.se/pdfs/datash/1636.pdf) to achieve a 2V ouput voltage. See attached drawing. Any comments on this?

regards, Theo ;-)

 

[Alien8]

I've done mods to a TDA1547 in a CD950, just one dac chip there.

Though I don't have the D1 stage. I'm listening to an AD844 I/V stage with a 6.8 k Resistor. (times four)

This is followed by a discrete differential to single ended output stage. The differential stage also has some voltage gain and the output stage is just a class A emitter follower. Simple first order filtering in between.

Sounds pretty decent, and definitely better then the on chip I/V opamps, even with better opamps in the filter chain. Just more relaxed, better bass and much less listening fatigue.
Clock mods too, but that's another story...

Power for this comes through a dedicated analog supply with shunts close to the stages, 4 in total.

Your drawing isn't shown...

HTH
Diego