Twindac+ and Django (BD-Design)
by dr.cube 
, Tuesday, September 19, 2006, 17:22 (6435 days ago)
I read that the Twindac+ has a transformer output stage. Also it seems that many people use the Django as a passive Volume Control, which also uses an transormer. So do you just connect the output of one trafo (from the twindac) to the input of another (the django)? If so, wouldn't it be even better if the output trafo of the twindac was itself a volume control, i.e. a trafo with many selectable outputs? You would spare a trafo in the signal chain which should result in even better signal quality.
What do you think about this?
Wolfgang
Twindac+ and Django
by Bert , Wednesday, September 20, 2006, 09:22 (6435 days ago) @ dr.cube
If so, wouldn't it be even better if the output trafo of the twindac was itself a volume control, i.e. a trafo with many selectable outputs? You would spare a trafo in the signal chain which should result in even better signal quality. What do you think about this?
Hi Wolfgang,
I discussed this with the designer of the TwinDAC+ because it seems a good idea. Less is more!
In reality it seems in this situation that things will be worse. Here is what he wrote to me:
"I do not think that will work, as you have to load the DAC in a correct way. Furthermore, you will change the winding ratio with each tap on the DAC transformer, which will have other negative effects on other parameters.
It is actually not a good idea to attenuate the tiny signal at the source. The output transformer acts as a passive pre-amp and the attenuation should be done on the amplified signal."
Ciao,
Bert
--
BD-Design - Only the Best!
Twindac+ and Django
by dr.cube , Wednesday, September 20, 2006, 11:50 (6435 days ago) @ Bert
Hi Bert,
thanks for your answer.
So you are using an active stage between the output of the dac and the input of the django? If not, the explanation doesn't make much sense to me because then you would also attenuate the tiny signal (no amplification in between!).
I guess the django has to be dc-free, i.e. it caNot be used as an anode load in a tube stage. Are you aware of any stepped transformers that allow some dc on the input? (I wanna have as little parts as possible in my system)
Wolfgang
Twindac+ and Django
by PeterSt. 
, Netherlands, Wednesday, September 20, 2006, 11:59 (6435 days ago) @ dr.cube
Hi Bert,
thanks for your answer.
So you are using an active stage between the output of the dac and the
input of the django? If not, the explanation doesn't make much sense to me
because then you would also attenuate the tiny signal (no amplification in
between!).
I guess the django has to be dc-free, i.e. it caNot be used as an anode
load in a tube stage. Are you aware of any stepped transformers that allow
some dc on the input? (I wanna have as little parts as possible in my
system)Wolfgang
I read it 10 times, and I don't understand it either.
And yes, the tiny signal is attenuated by the Django. But the tiny signal (500 mV) is large enough for the need to attenuate it. 
Anyway, my trying to understand lead to the route of having attenuation in the speaker leads. Things like these are made for that, but it's not easy to judge how legit it is, and how *that* will molest the signal.
Peter (Googling now)
Twindac+ and Django
by Bert , Wednesday, September 20, 2006, 12:09 (6435 days ago) @ dr.cube
Hi Wolfgang,
The transformer inside the DAC already "amplifies" the signal due to the winding differences. Changing the parameters (by making taps for attenuation) will change/disturb the primarely load of the DAC.
Attenuating after the DAC's "amplification" there are much less problems that will change the primarely load of the transformer inside the DAC.
There is no active stage or buffer between the DAC and my Django. And DC should always be avoided. This means all around the transformers (input and output for both the Django and the TwinDAC+
Ciao,
Bert
--
BD-Design - Only the Best!
Twindac+ and Django
by dr.cube , Wednesday, September 20, 2006, 13:09 (6434 days ago) @ Bert
Hi Bert,
Well, technically a transformer caNot "amplify", it caNonly "trade" Voltage for current (Power, which is U*I, caNot increase and will in fact decrease due to some losses (iron/copper)).
If you switch the output tap of the django (selecting different output levels), the input impedance also changes (See http://www.stevens-billington.co.uk/page102.htm), and therefore the reflected primary impedance of the transformer inside the dac (the impedance that the current source at the input of the trafo "sees") is also different.
So for me the explanation doesn't hold, technically there seems to be a transformer which can be spared. I don't think that there in principle is anything achievable by 2 transformers in a row that caNot be achieved using a single one (which of course has to be proberly designed)
Twindac+ and Django
by Bert , Wednesday, September 20, 2006, 13:55 (6434 days ago) @ dr.cube
Hi Wolfgang,
No problem, gain versus current. The signal gets more Voltage and therefore sounds louder... 
So for me the explanation doesn't hold, technically there seems to be a transformer which can be spared. I don't think that there in principle is anything achievable by 2 transformers in a row that caNot be achieved using a single one (which of course has to be proberly designed)
Perhaps this is possible, my technical knowledge around this is too limited to explain why it can't or should work. Its up to the designer of the TwinDAC+ to see if he can have something made that involves only one transformer with the ability to attenuate the signal without loosing quality... 
Ciao,
Bert
--
BD-Design - Only the Best!
Twindac+ and Django
Correct, a transformer caNot amplify power, but we can increase the voltage due to the turning ratio of a transformer. There is little current draw from the input of the amp. So, the drive of this transformer is sufficient for the amp. If it would not be, you would notice an early roll off in the low end of frequencies.
As you mentioned, a transformer merely transforms a load impedance on its secondary. The turn ratio or the voltage ratio between primary and secondary governs the transformed and reflected load impedance.
By having a transformer as suggested, you caNot achieve a stable impedance at the DAC chip output, as the impedance changes with different turn ratios.
The output transformer is optimal loaded for the DAC. This is critical. The reflected impedance is low. By adding a transformer attentuator, with a high loading impedance (amp), the reflected impedance of the transformer attentuator can be neglected. It does not change the impedance that the DAC sees practically.
In addition, as you mentioned as well, a transformer is always a compromise. Designing, using the right materials, winding techniques etc. is not that simple at all. That is also the reason that a quality transformer costs quite a lot of money.
Mirko
Twindac+ and Django
by dr.cube , Wednesday, September 20, 2006, 18:42 (6434 days ago) @ unknown
By having a transformer as suggested, you caNot achieve a stable
impedance at the DAC chip output, as the impedance changes with different
turn ratios.
But as I said before: If you load the dac output trafo only with the django, then the reflected input impedance also changes with the volume (tap) selection of the django. Unless you have an additional load on the dac trafo which had be smaller than the reflected input impedance from the django (which is usually quite high if you have a valve grid at the amps input stage. But I think you could provide that low load also if you have just one trafo.
Mirko, are you the designer of the twindac?
Wolfgang