thing still being in development , let skip money talk
regarding concept talk , it'll be something as :
SE ,stereo:
-24pos. rotary switch for relay selection , mounted on pcb with relay PSU and flat cable header
-flat cable
-two AVCs with 24 relay pcb in between , as assembled and tested unit ; selectable gain positions ,say, 0,+2,+4,+6,+8db
-pair of Al rods as mounting hardware
Bal package ,stereo :
-24pos. rotary switch for relay selection , mounted on pcb with relay PSU and flat cable headers
-flat cable , 2(two) pcs
-AVC with 24 relay pcb ,as assembled and tested unit , selectable gain positions ,say, 0,+2,+4,+6,+8db , 2 (two) pcs (one per channel)
-two pairs of Al rods as mounting hardware
well , plan is to make it at least good as Dave's ; Sowter not being in my palette of reference
NB that I'm strong opponent of using any sort of inductive attenuator sans decent input buffer ; without that , any talk about comparisons is pointless
one package (core, bobbins,strip) being around 20E , Metglas ; Finemet being 4 times that
payment and delivery costs extra
will try both
thanks for the info and your great effort, looking forward to completion : )
Is there a straightforward/ recommended way to hook up the AVC as a balanced to unbalanced converter?
Would you think that it is worth the effort (buffers) to feed the AVC balanced from a balanced DAC output and have the AVC feed an F5 or M2?
last question: I think I saw your schematic for the diamond buffer but I can't find it anymore. can anyone please point me there?
Thx.
Is there a straightforward/ recommended way to hook up the AVC as a balanced to unbalanced converter?
Would you think that it is worth the effort (buffers) to feed the AVC balanced from a balanced DAC output and have the AVC feed an F5 or M2?
last question: I think I saw your schematic for the diamond buffer but I can't find it anymore. can anyone please point me there?
Thx.
just take a look at Iron Pre schematics - everything is shown
just imagine Sowter shown as multitapped thingie
I can't remember did I post it , but I will , soon enough
if you r source is having sub 50R Rout , and you can keep short interconnects , we can say that you're buffered enough
if you're lacking in any of these , you need buffering
sorry, didn't make myself clear.
to recap my understanding: if I want to feed a SE amp like F5 or M2 I simply use one phase for the hot wire and Gnd goes to Gnd, right?
to understand the resulting gain I probably need to go back and check how my DAC converts Bal to SE, I presume its a diffAmp.
Question was, is there any merit in going balanced from DAC to AVC or would I be better off to use the SE out of the DAC and run the AVC SE in SE out into M2.
I plan to use a buffer in any case. For the balanced mode I just need twice as many.if you r source is having sub 50R Rout , and you can keep short interconnects , we can say that you're buffered enough
if you're lacking in any of these , you need buffering
it's simple as : use what's of greater quality
if you have true differential DAC output stage , it is most likely superior to SE , sound wise
in that case is worth of all hassle , going balanced all the way
if DAC is inside true SE , with bal out just glued as conveniency , keep it simple and go for SE afterwards
regarding Bal to SE , proper way of converting is through xformer (when galvanic isolation is must or preferable) ...... and it's simple as using one end of secondary as hot , other one as gnd
if you have true differential DAC output stage , it is most likely superior to SE , sound wise
in that case is worth of all hassle , going balanced all the way
if DAC is inside true SE , with bal out just glued as conveniency , keep it simple and go for SE afterwards
regarding Bal to SE , proper way of converting is through xformer (when galvanic isolation is must or preferable) ...... and it's simple as using one end of secondary as hot , other one as gnd
Ok, looked up the specs. DAC is a dacmagic with one WM8740 for each channel inside. Each feeds two ne5532 as amps, one for + and one for - phase. From there both xlr out and a OP275 are fed, each phase with 1:1 feedback. Output of OP275 feeds RCA.
Spec says unbalanced 2.1V rms and balanced 4.2V rms (2.1V per phase).
For this case you would go for balanced AVC, right?
well , if implementation is "by the book" , proper differential gain stage as shown on Fig.26 of datasheet , I would certainly choose balanced operation of following gadgets
disclaimer - I would certainly ( also
) work on tossing out all those dead bugz , replacing them either with other dead bugz but made by Papa SUSY , or lineup made of 10K:10K xformer , followed by JFet buffers in each phase ........... or lineup made of JFet buffer in each phase followed by low impedance 1:2 xformers
more iron in signal route , the merrier
disclaimer - I would certainly ( also
) work on tossing out all those dead bugz , replacing them either with other dead bugz but made by Papa SUSY , or lineup made of 10K:10K xformer , followed by JFet buffers in each phase ........... or lineup made of JFet buffer in each phase followed by low impedance 1:2 xformersmore iron in signal route , the merrier
Attachments
Last edited:
aha .... one WM per channel , I just didn't grasped that part
for my untrained eyes - I see it as two bloody SE (starting as bal , but translated to SE)
gain stages forming so called balanced output, but not real differential ...... at least not for my liking and sense
take SE signal ditto from XLR , pin2 hot , pin 1 gnd - it'll sound better than dreky RCA you already have
for me - you don't have worthy balanced source , so stay in SE realm all the way
for my untrained eyes - I see it as two bloody SE (starting as bal , but translated to SE)
gain stages forming so called balanced output, but not real differential ...... at least not for my liking and sense
take SE signal ditto from XLR , pin2 hot , pin 1 gnd - it'll sound better than dreky RCA you already have
for me - you don't have worthy balanced source , so stay in SE realm all the way
Thx. Short pin 3 to pin 1? This is what I do currently.
Gesendet von meinem SM-T520 mit Tapatalk
Gesendet von meinem SM-T520 mit Tapatalk
Dear Mighty One.
Thinking that an active crossover would be better after attempts with coils and caps I am wondering if there is any chance you might offer a board of the buffer alone. It would have to be good for this.
How many boards would be required for this to be worth your while?
Thinking that an active crossover would be better after attempts with coils and caps I am wondering if there is any chance you might offer a board of the buffer alone. It would have to be good for this.
How many boards would be required for this to be worth your while?
sorry , I'm just Mighty ZM ......... One is another one , and he's smarter than your Nelson
- to be sure that you're going in right direction , you need to make quick and cheap mock-up of active xover - either with discrete buffers , or OP based (maybe someone around is having some sort of commercial active xover?)
when you test that , and possibly decide for following that route , then is time do decide are you going with JFets ( either N type stacked or N-P complementary type) , some other discrete type , or some other chip based
it would be expensive , both money and time and ( -that is going to be third) energy - wise , to make everything and only then hear is it better ...... vs. passive solution you have now
I believe , that's worthy of dedicated thread , solving that dilemma
btw. present passive xover - is it line level , or speaker level ?
- to be sure that you're going in right direction , you need to make quick and cheap mock-up of active xover - either with discrete buffers , or OP based (maybe someone around is having some sort of commercial active xover?)
when you test that , and possibly decide for following that route , then is time do decide are you going with JFets ( either N type stacked or N-P complementary type) , some other discrete type , or some other chip based
it would be expensive , both money and time and (
-that is going to be third) energy - wise , to make everything and only then hear is it better ...... vs. passive solution you have nowI believe , that's worthy of dedicated thread , solving that dilemma
btw. present passive xover - is it line level , or speaker level ?
@ZM
creating a new thread in Papa land makes me think whether it's worthed, for my xo build it's just like build log not giving something new. but maybe I'll start my thread when i get home
@rick
you can start reading at B1 active XO thread to get overall idea on what you are going to chase. is it 12dB or 24dB, do you need BSC or bass boost? then you can start counting on how many jfets that you will need. Again as ZM mention, you also need to decide which jfet 2SK170 only (like B5) or 2SK170+2SJ74 (like B4).
for my active xo, i need 16pair and last time i bought SK+SJ from alweit was $14/pair. Too much money for me so i choose SK170 only. but maybe your wallet is deeper than mine since you bought those SIT1
btw sneak peek on my build http://www.diyaudio.com/forums/grou...as-sslv1-1-bib-shunt-reg-163.html#post4951358
creating a new thread in Papa land makes me think whether it's worthed, for my xo build it's just like build log not giving something new. but maybe I'll start my thread when i get home
@rick
you can start reading at B1 active XO thread to get overall idea on what you are going to chase. is it 12dB or 24dB, do you need BSC or bass boost? then you can start counting on how many jfets that you will need. Again as ZM mention, you also need to decide which jfet 2SK170 only (like B5) or 2SK170+2SJ74 (like B4).
for my active xo, i need 16pair and last time i bought SK+SJ from alweit was $14/pair. Too much money for me so i choose SK170 only. but maybe your wallet is deeper than mine since you bought those SIT1
btw sneak peek on my build http://www.diyaudio.com/forums/grou...as-sslv1-1-bib-shunt-reg-163.html#post4951358
Last edited:
Dear ZM,
I am using speaker level crossovers which have done a decent job. Or so I thought. Or so I convinced myself!
THEN I started taking room measurements with REW and found many interesting things. From what I found I have been able to improve the crossovers but I see there are areas that defy improvement with speaker level components.
The driver running straight in has a much flatter response in the range that matters than it does with a choke or a choke/cap - I tried all kinds of combos - many wacky just to see what happens. No matter what I do placing those components in the path makes the response much spikier. With nothing it is reasonably flat where it matters which is why I figure line level would be superior.
I have to say I think much of the problem is the horn itself - using a 12 inches driver it will just get to 400 hz. This driver in other horns is not limited to this so I know the horn is wrong. I know a line level crossover is not going to solve those problems - I have to replace the horn - but I would like this to be as good as possible while I have to use these and be able to use it when I replace these horns.
I could go with the FETs but I thought one of the advantages of the diamond buffer was the greatly less expensive devices which led me to believe it would be much cheaper using these. So that is not the case?
I think I only need a second order low pass - haven't decided whether to make this a bandpass filter - but that could wait - I do not have to make all of this at one time. Due to the nature of the SIT1s requiring input and output caps I have good control over the above 500 hz horn.
I would be glad to start another thread and know I am pushing the limits of the topic here. AS gadut says it looks like crossover threads die a quick death which is unfortunate.
If I stick with the FETs I would use the DIYAudio Store service and continue with the complementary buffer. I have four of these buffers on hand and will give these a try first.
Always wanting the latest and the most interesting which is why I want to consider the ZEN MOD Diamond Buffers for the task. In addition a PCB would make packaging the thing much neater!
I am using speaker level crossovers which have done a decent job. Or so I thought. Or so I convinced myself!
THEN I started taking room measurements with REW and found many interesting things. From what I found I have been able to improve the crossovers but I see there are areas that defy improvement with speaker level components.
The driver running straight in has a much flatter response in the range that matters than it does with a choke or a choke/cap - I tried all kinds of combos - many wacky just to see what happens. No matter what I do placing those components in the path makes the response much spikier. With nothing it is reasonably flat where it matters which is why I figure line level would be superior.
I have to say I think much of the problem is the horn itself - using a 12 inches driver it will just get to 400 hz. This driver in other horns is not limited to this so I know the horn is wrong. I know a line level crossover is not going to solve those problems - I have to replace the horn - but I would like this to be as good as possible while I have to use these and be able to use it when I replace these horns.
I could go with the FETs but I thought one of the advantages of the diamond buffer was the greatly less expensive devices which led me to believe it would be much cheaper using these. So that is not the case?
I think I only need a second order low pass - haven't decided whether to make this a bandpass filter - but that could wait - I do not have to make all of this at one time. Due to the nature of the SIT1s requiring input and output caps I have good control over the above 500 hz horn.
I would be glad to start another thread and know I am pushing the limits of the topic here. AS gadut says it looks like crossover threads die a quick death which is unfortunate.
If I stick with the FETs I would use the DIYAudio Store service and continue with the complementary buffer. I have four of these buffers on hand and will give these a try first.
Always wanting the latest and the most interesting which is why I want to consider the ZEN MOD Diamond Buffers for the task. In addition a PCB would make packaging the thing much neater!
Last edited:
I just received a few new toys to experiment with mighty one's ideas and M2 front end.
If only I did not mess up the hole size for the Edcors... You really need the diameter suggested on the pdf to compensate for deviations from nominal pdf dimensions. Fortunately, a bit of pin bending did solve this.
Now to get some nice PSUs 🙂
If only I did not mess up the hole size for the Edcors... You really need the diameter suggested on the pdf to compensate for deviations from nominal pdf dimensions. Fortunately, a bit of pin bending did solve this.
Now to get some nice PSUs 🙂
