I'm not sure about whether I'll use the LCD screen yet so I'll just say probably for now.
I won't be able to offer any comparison to the AK4490, not least because I no longer have one and memories aren't reliable.
I do plan to use tubes on the output in some shape or form but detail to be confirmed. I'm also looking at using a transformer on the DAC outputs. Would be interest in seeing some info on your ecc82 output.
Ray
sangram, I've made an enquiry about suitable Llundahl transformers for both the Diff/SE conversion but also possibilities for performing the summing too, as you suggested. Do you have any suggestions for suitable transformers yourself?
Having read more about the TVC and the potential loading it would place on the DAC outputs I have discounted using it.
Ray
Hi
So I must confess that I've never used a transformer output on a Vout DAC, but I think what you're looking for is a high-impedance 1+1:1+1 matching transformer.
I also have not been able to locate anything suitable after looking at Jensen and Oxford catalogs. There are similar products but most matching transformers tend to be suited to low source impedances. Still, some of the Jensen units can tolerate up to +20dB signal level so even the low impedance units should be possible. Since all of these guys (Sowter is another) are close to you, maybe give them a call?
The usual issue will be frequency response on lowering transformer load, the higher signal levels caused by lower loads, and the low impedance presented to the DAC output. None of these are issues in current output DACs, where one only need worry about the compliance of the DAC's current sources. The 4497 is equipped with a heavy load mode which may be used to drive 600 ohm loads, but I read a Japanese blog which commented on sound quality deterioration (no measurements were provided). If you do use the 600 ohm termination, a buffer will probably not be needed at all. However the performance of the DAC looking into such a load will need to be considered.
You will typically connect the output windings in parallel, and the DAC outputs will be connected to the transformer input. The phase is important, as if connected wrongly you will get no sound at all, or very poor output quality. The +L and -L will connect to one winding, and the -R and +R to the other (note reversed polarity).
It may be possible to eliminate coupling caps if you used this configuration, as the net DC across the transformer will be zero. I would still urge you to check with the manufacturer if this is possible.
So I must confess that I've never used a transformer output on a Vout DAC, but I think what you're looking for is a high-impedance 1+1:1+1 matching transformer.
I also have not been able to locate anything suitable after looking at Jensen and Oxford catalogs. There are similar products but most matching transformers tend to be suited to low source impedances. Still, some of the Jensen units can tolerate up to +20dB signal level so even the low impedance units should be possible. Since all of these guys (Sowter is another) are close to you, maybe give them a call?
The usual issue will be frequency response on lowering transformer load, the higher signal levels caused by lower loads, and the low impedance presented to the DAC output. None of these are issues in current output DACs, where one only need worry about the compliance of the DAC's current sources. The 4497 is equipped with a heavy load mode which may be used to drive 600 ohm loads, but I read a Japanese blog which commented on sound quality deterioration (no measurements were provided). If you do use the 600 ohm termination, a buffer will probably not be needed at all. However the performance of the DAC looking into such a load will need to be considered.
You will typically connect the output windings in parallel, and the DAC outputs will be connected to the transformer input. The phase is important, as if connected wrongly you will get no sound at all, or very poor output quality. The +L and -L will connect to one winding, and the -R and +R to the other (note reversed polarity).
It may be possible to eliminate coupling caps if you used this configuration, as the net DC across the transformer will be zero. I would still urge you to check with the manufacturer if this is possible.
Last edited:
Firstly, I found a couple of other threads where people have done something similar;
http://www.diyaudio.com/forums/digi...ntial-passive-transformer-output-stage-2.html
http://www.diyaudio.com/forums/digital-line-level/261711-0-5-0-5-1-line-transformer-question.html
so it looks possible to sum and perform diff to SE via a transformer.
I contacted Kevin at K and K Audio and asked him to recommend a suitable Llundahl transformer an he has suggested either the LL1674 or LL1676 depending on how much stepdown I need (I need stepdown because my power amps have a sensitivity of around 1V and with the DAC output at around 2.5V, which is doubled with diff to SE conversion I may have too much gain). Just asked some supplementary questions but I think I'll act on the suggestions.
http://www.lundahl.se/wp-content/uploads/datasheets/1674.pdf
http://www.lundahl.se/wp-content/uploads/datasheets/1676.pdf
http://www.diyaudio.com/forums/digi...ntial-passive-transformer-output-stage-2.html
http://www.diyaudio.com/forums/digital-line-level/261711-0-5-0-5-1-line-transformer-question.html
so it looks possible to sum and perform diff to SE via a transformer.
I contacted Kevin at K and K Audio and asked him to recommend a suitable Llundahl transformer an he has suggested either the LL1674 or LL1676 depending on how much stepdown I need (I need stepdown because my power amps have a sensitivity of around 1V and with the DAC output at around 2.5V, which is doubled with diff to SE conversion I may have too much gain). Just asked some supplementary questions but I think I'll act on the suggestions.
http://www.lundahl.se/wp-content/uploads/datasheets/1674.pdf
http://www.lundahl.se/wp-content/uploads/datasheets/1676.pdf
Yes, from what I saw via the link it looks like a nice solution for a single DAC implementation but looks as though it would need a summing section between the DAC outputs and the Super V/I.
http://www.diyaudio.com/forums/digi...al-mono-dac-ak4490-ak4497-16.html#post4917127
and following posts...
See the attachment for what I plan.
Last edited:
I am sorry but I don't agree with the "V DAC output LL1676f.pdf" schematic.
From what I know, at this setup the AK has Vref/2 dc offset at the outputs.
LL1676 is a transformer, so you won't measure any dc offset at the secondary connections. But with such dc offset that AK has, the core will be saturated, and bass frequencies are essentially not passed by the transformer!
From my opinion if you wanna use a transformer, you must find a nice way to cut the dc-offset before the primaries...
From what I know, at this setup the AK has Vref/2 dc offset at the outputs.
LL1676 is a transformer, so you won't measure any dc offset at the secondary connections. But with such dc offset that AK has, the core will be saturated, and bass frequencies are essentially not passed by the transformer!
From my opinion if you wanna use a transformer, you must find a nice way to cut the dc-offset before the primaries...
You may well be right. The information I received is that the DC offset is negligible so saturation may not be an issue, however, if I find this to be a problem I will experiment with caps to block the DC.
The offset for the DAC is between the outputs and earth, not between each output. Put another way, the 'P' and 'N' outputs are sitting at the same voltage, so there is no offset between them. This means there is no significant DC going into the transformers.
P and N aren't opposite of each other? And P and N don't each average to gnd?
DAC works on single supply voltage. The current sources within the DAC sit at half the supply voltage, which is represented at the outputs. In case of current output DACs the Iout should ideally sit at 0v (virtual ground). In case of Vout DACs the +inputs of the internal opamp are connected to the reference voltage and thus the output is at Vcc/2.
This is the same for both phases of each output and thus there is no relative offset between the two phases.
This is the same for both phases of each output and thus there is no relative offset between the two phases.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.