Qusp, Ok lets clear up a few things....
1) You absolutely can use Legato single ended without *any* BAL/SE stage. It just will have higher distortion. Personally I think this would be extremely foolish. Just like any common base/gate output stage it will be subject to some thump on turn on/off when used like this. You can eliminate this with a relay.
2) I already have an awesome low distortion discrete BAL//SE stage, but there is no way it would fit on the legato board. It is called Ventus. I expect people who want to go that route would likely want to use that board.
3) Balanced output dual mono with a single or two Legatos without the BAL/SE stage is absolutely possible, in fact easy. For a single legato just parallel the outputs from the Buffalo II (mind the phase difference). For two legatos you would simply parallel the outputs of the legatos(keeping the same rule in mind).
I didn't do anything halfway. I finished it to the last detail and then went the extra mile by adding something **EXTRA**. That extremely convenient BAL/SE stage is excellent for the purpose, as you state yourself, and the very best thing about it for the DIY community is... IT'S COMPLETELY OPTIONAL!!!! I hope I made that clear enough.
You can still do SE output without it. The whole idea is to open up options.
1) You absolutely can use Legato single ended without *any* BAL/SE stage. It just will have higher distortion. Personally I think this would be extremely foolish. Just like any common base/gate output stage it will be subject to some thump on turn on/off when used like this. You can eliminate this with a relay.
2) I already have an awesome low distortion discrete BAL//SE stage, but there is no way it would fit on the legato board. It is called Ventus. I expect people who want to go that route would likely want to use that board.
3) Balanced output dual mono with a single or two Legatos without the BAL/SE stage is absolutely possible, in fact easy. For a single legato just parallel the outputs from the Buffalo II (mind the phase difference). For two legatos you would simply parallel the outputs of the legatos(keeping the same rule in mind).
I didn't do anything halfway. I finished it to the last detail and then went the extra mile by adding something **EXTRA**. That extremely convenient BAL/SE stage is excellent for the purpose, as you state yourself, and the very best thing about it for the DIY community is... IT'S COMPLETELY OPTIONAL!!!! I hope I made that clear enough.
You can still do SE output without it. The whole idea is to open up options.
Last edited:
Hi
Yes, there hasn't been much feedback about Legato! I know Russ and Brian are reluctant to give subjective opinions themselves, but Russ did ask for feedback!
I think I agree with this- very provisionally! I'm now using Legato + Bal to SE stage, as supplied (no buffers). It sounds awesome, though I feel, at some stage recently, I've heard even better - but only at the top end. Trouble is, I've also used Buffalo II with a Legato with ZTX651 in place of the BC560, Buffalo + another I/V design entirely, and IVY II. What's more I also modified my power amp not long ago.
I need to sort this out. I only mention it to perhaps encorage other Legato users to express their own experiences!
I've ordered another Legato. I'll fit SIL sockets for the transistors, and do some comparative listening. I'll keep the BS250, and the BC560 which biasses the CFP, but try some other BJTs in place of the remaining BC560s- ZTX651, then 2SC2547E- for starters. I'll also revisit the other I/V stages.
IMHO, changes to the active devices are more likely to provide noticable sonic differences (hopefully improvements- but given the care Russ put in, this won't be easy I guess) than fiddling with the resistors and cap types
Best wishes
Paul
Yes, there hasn't been much feedback about Legato! I know Russ and Brian are reluctant to give subjective opinions themselves, but Russ did ask for feedback!
I think I agree with this- very provisionally! I'm now using Legato + Bal to SE stage, as supplied (no buffers). It sounds awesome, though I feel, at some stage recently, I've heard even better - but only at the top end. Trouble is, I've also used Buffalo II with a Legato with ZTX651 in place of the BC560, Buffalo + another I/V design entirely, and IVY II. What's more I also modified my power amp not long ago.
I need to sort this out. I only mention it to perhaps encorage other Legato users to express their own experiences!
I've ordered another Legato. I'll fit SIL sockets for the transistors, and do some comparative listening. I'll keep the BS250, and the BC560 which biasses the CFP, but try some other BJTs in place of the remaining BC560s- ZTX651, then 2SC2547E- for starters. I'll also revisit the other I/V stages.
IMHO, changes to the active devices are more likely to provide noticable sonic differences (hopefully improvements- but given the care Russ put in, this won't be easy I guess) than fiddling with the resistors and cap types
Best wishes
Paul
dont know why you would do that, thump, plus throwing away a good portion of DNR and all the CMRR
not doubting that, though it hasnt been available for a while now
none of these are dual differential really 2 dacs, 2 x legato and 2 x SE convertors is the only way, IMO paralleling the outputs of the dacs or IV stage kinda defeats the point.
I didn't do anything halfway. I finished it to the last detail and then went the extra mile by adding something **EXTRA**. That extremely convenient BAL/SE stage is excellent for the purpose, as you state yourself, and the very best thing about it for the DIY community is... IT'S COMPLETELY OPTIONAL!!!! I hope I made that clear enough.
Russ mate, settle
this isnt even my view, just saying that there will probably be an element that will avoid it based on the opamp and gave the reasons for that. I dont believe any of it, other than the difficulties of doing a proper discrete dual DIFF build wth the modules as they are, needing to use 3 separate boards for each channel isnt really ideal. but hey dual diff is always going to be involved, would just be nice if someone; would do it. well actually erno's all fet IV will do it using 2 of the diff input and SE output version; but $$$$cant please everyone
but I repeat, for my actual current purposes, its excellent
Qusp,
My only aim is to enlighten. Please describe why paralleling as I described would defeat the point?
The whole "point" of dual differential is to combine the output of both sides of a DAC in such a way that each half is at opposite phase. Paralleling the outputs (of either the DAC or the Legato) is no less "proper" than any other means of achieving this. In fact it is an excellent way to do it. The net effect is the same what you would achieve by any other means.
You only need one single Legato. That is all. If you want discrete SE output then use one legato and two Ventus(they are mono).
Ventus is soon to be re-released as it has undergone some redesign. It will be available again very soon.
You could use just two Ventus just like you would use the Borbelly stage. The difference when using either of those directly is that you are now using the DAC as a voltage source. Thats not bad, its just not the same animal.
Last edited:
The LME49600 is nothing more than a diamond buffer plain and simple. Exactly the same as I would use discretely. In fact, it is better, because of the thermal coupling of the buffer to PCB which makes the whole PCB a heatsink. Add to that the tight matching of the transistors which would otherwise have to be done by hand(ick). Exploding this simple diamond buffer out into discrete components would do absolutely nothing to improve the design. Quite the opposite, it would be impossible to match the results.
It probably could indeed be omitted, but I would never recommend that.
Cheers!
Russ
Last edited:
because without having 2 of them, they are sharing a power supply and if bothering to spend the dollars having 2 of everything else, it seems a bit stupid to not finish the job at this stage where you only have another 60 dollars or whatever to spend to go all the way. plus it would make for a bit of a layout issue, lack of symmetry
and no, you dont have to use the borbely stage as voltage source; needs modification, but doable
and no, you dont have to use the borbely stage as voltage source; needs modification, but doable
Last edited:
I guess i'm looking at this differently. i'm looking at it as using 2 dac boards to achieve dual diff the same as in dual mono; or is that option from the chip not supported with the buff Russ? as it is with the.... other sabre; as in
DIFF MONO. The left and right outputs carries the same signal (left or right channel) but
out-of-phase with each other.
if not, sorry for the confusion; I guess I wasnt communicating correctly
Last edited:
Yes, two Buffalo boards, one Left, one Right, both feeding a single Legato (+ G -), with separate supplies for the Left and Right I/V stages.
because without having 2 of them, they are sharing a power supply and if bothering to spend the dollars having 2 of everything else, it seems a bit stupid to not finish the job at this stage where you only have another 60 dollars or whatever to spend to go all the way. plus it would make for a bit of a layout issue, lack of symmetry
How so? You could easily use two supplies. Remember that the legato board allows you to use seperate supplies. The layout could still be completely symmetrical. That part is up to you.
Also if you willing to build two supplies also using two Legatos does not seem much of a stretch.
I would love to see how you do that with the borbely stage unless you used 4 channels. That however, would no longer be a simple case. You could run it like an inverting opamp I/V, but then dual differential would take 4 channels. You could also do that with Ventus.
Last edited:
As Brian said, that is precisely what I have described here several times. And that is the way the buffalo works in dual (yes differential) mono mode.
right, so IMO its not dual mono on the buff II regardless of the label. there is another setup for the sabre that allows what is IMO more correctly dual mono. ie both left and right channels are the same and in phase; thus need to be summed after. so it was a matter of differing language interpretation confusion (also on my part). the other dac allows this other setting and to me that is dual mono, where what you describe and what I ended up talking about is diff mono.
I didnt think I remembered the buff having a diff mono setting, just dual mono in the manual and since I only have one buff II I havent tried it (or looked into it) I assumed that is what you were talking about, so was wondering where the summing was happening. So you need a SE stage for each channel to produce a balanced output, rather than SE output which isnt possible discretely with the legato, or really the ventus due to the rather excellent IC buffer....phew
so yes of course you can use the legato and just one... haha and all that for a hypothetical quibble
combined with a brain defect
peace out
I didnt think I remembered the buff having a diff mono setting, just dual mono in the manual and since I only have one buff II I havent tried it (or looked into it) I assumed that is what you were talking about, so was wondering where the summing was happening. So you need a SE stage for each channel to produce a balanced output, rather than SE output which isnt possible discretely with the legato, or really the ventus due to the rather excellent IC buffer....phew
so yes of course you can use the legato and just one... haha and all that for a hypothetical quibble
combined with a brain defect peace out
Last edited:
Buffalo II allows either option. The included firmware reverses the phase of the non-selected channel in mono mode, but you do not have to *wire* it inverted at all! Thats is completely up to you and how you wire it.
The reasons for reversing the off side should be self explanatory. If you would like I will explain it.
Physically speaking there is absolutely no advantage to having them both in phase, if anything you would want them out of phase so that the DAC is loading the power supplies exactly symmetrically cancelling harmonics. Still I doubt there is any measurable difference between running the off-channel one way or the other.
If anyone really wanted firmware that would run each side in phase I would be glad to provide it, but you then would not be able to use IVIII or Legato in dual mono mode, because each side would sum to zero.
. In short I can't think of any logical reason to do it.But I do want to clarify that the Buffalo II can be used either way. Right out of the box. If you want both side in phase, wire it that way. If you don't don't.
Last edited:
fair call hehe, been puzzling over that very thing (ie the point). but russ, why in that case would you need a summing stage, when as you both mention its possible to just reverse the wiring. its interesting that both yourself and acko have mentioned that it results in needing a summing stage, which would simply do the 'rewiring' for you at great cost. the in phase option does seem to me to fit the dual mono description better though.
(rhetorical question above, why have the option thats all, creeping featurism at its best I guess)
BTW midnight or 8am these last few weeks have ruined me. selling the house; open days every few days, working 60+ hours and then trying to find time to work on my own stuff is frying me badly. I work primarily online to asia, EU and the USA, so whenever I have commitments in AU it ruins me as I live as a shift worker pretty much. dont know whether i'm coming or going. normally thats only for a couple of days, but lately its constantly due to the open house stuff. so yeah it seems jibberish results
I promise i'll come back here with a valid comparison of the units in the OP. but for now i'll buy out and leave you to your regularly scheduled program. i'll keep an eye on it though, reports so far are very encouraging, but I wont be buying for a couple of weeks at least; till I have the power supply sorted out in the portabuff properly. I also need to know that the layout will fit well and my battery runtime can handle the extra ~180-200ma consumption compared to my current OPA1632 based output
(rhetorical question above, why have the option thats all, creeping featurism at its best I guess)
BTW midnight or 8am these last few weeks have ruined me. selling the house; open days every few days, working 60+ hours and then trying to find time to work on my own stuff is frying me badly. I work primarily online to asia, EU and the USA, so whenever I have commitments in AU it ruins me as I live as a shift worker pretty much. dont know whether i'm coming or going. normally thats only for a couple of days, but lately its constantly due to the open house stuff. so yeah it seems jibberish results
I promise i'll come back here with a valid comparison of the units in the OP. but for now i'll buy out and leave you to your regularly scheduled program. i'll keep an eye on it though, reports so far are very encouraging, but I wont be buying for a couple of weeks at least; till I have the power supply sorted out in the portabuff properly. I also need to know that the layout will fit well and my battery runtime can handle the extra ~180-200ma consumption compared to my current OPA1632 based output
Last edited:
Qusp, the answer is very simple.
The reason for reversing the off side in mono mode is it allows you to easily use the DAC in dual mono mode with IVY-III or Legato, because the signal at what would normally be the SE outputs will be differential. The other reason is that (at least theoretically) loading the DAC this way is better. Other DACs do precisely the same thing for the same reason. WM8741 for example in dual mono mode. Also some of TI's higher end DACs too. There is solid engineering behind that choice.
This way you simply mount the DAC as you would in stereo mode using the mezzanine connectors. It makes everything very elegant and simple.
The bottom line is you can configure it any way you like. There is no barrier to anything you would like to try.
Cheers!
Russ
The reason for reversing the off side in mono mode is it allows you to easily use the DAC in dual mono mode with IVY-III or Legato, because the signal at what would normally be the SE outputs will be differential. The other reason is that (at least theoretically) loading the DAC this way is better. Other DACs do precisely the same thing for the same reason. WM8741 for example in dual mono mode. Also some of TI's higher end DACs too. There is solid engineering behind that choice.
This way you simply mount the DAC as you would in stereo mode using the mezzanine connectors. It makes everything very elegant and simple.
The bottom line is you can configure it any way you like. There is no barrier to anything you would like to try.
Cheers!
Russ
Last edited:
no, I meant why provide the IN phase option when it provides no clear benefit and to my thinking as well doesnt seem like it would work that well even if it seems more 'mono' wouldnt CMRR be reduced as well? since the off side is physically seperated from the on side tyheoretically being subjected to slightly different levels of EMI etc and wiring in differential pairs difficult or impossible. complimentary pairs not producing complimentarty output (dont know how much that matters) when using it as is rather than rewiring it requires another stage of processing for seemingly no reason (that I can think of) it just seems odd thats all. I wasnt directing that at you as any accusation. just wondering why ESS has the option; I guess there must be some reason. but I think i'll just drop it. thanks for your patience
Last edited:
Oh well that I can explain too.
ESS makes the phase of all 16 DACs (2 for each channel) accessible to the firmware designer. You could actually take all sixteen and run them with exactly the same signal and exactly the same phase... But that would be very silly indeed.
It makes no assumptions about how you will use them. Using them *well* is completely up to the implementor.
ESS makes the phase of all 16 DACs (2 for each channel) accessible to the firmware designer. You could actually take all sixteen and run them with exactly the same signal and exactly the same phase... But that would be very silly indeed.
It makes no assumptions about how you will use them. Using them *well* is completely up to the implementor.