Re: Is this what you are talking about?
Konnichiwa,
Don't bother. You really need to add a few extra transformers (they need no be huge size) and build seperate supplies. Working of the original shared supplies is just by far too much compromise. Leave the original supplies o power only Display/Control uP & Remote control circuitry, all other supplies should be build new from scratch and be galvanically insulated, so you can break up several bad ground loops and feed each and every section optimally.
For Players of the CD-720/21/22/23 ilk (to which yous also belongs IIRC) I'd implement as follows:
1) 2 X 15V (seperate supplies only joined to for +/-15V at the DAC's Starground) with a further +15V tapped off the positive supply to null offset and +5V tapped off the positive supply to power the DAC.
All regs. in this supply should be at least LM317 better LT/LM1085, adj pin bypassed and good "audio grade" output capacitor, lower than common value "set" resistor chains to draw at least 50mA standing current through the regulator (more is better within reason).
Diodes Schottky, PSU Cap's small value and raw voltages fairly high, ideally with some added RC filtering, something like 3R3 470uF - 10R - 4,700uF - Reg with a 18V transformer and schottky bridge and the regulators (+15V Op-Amp, 15V Offset Null, +5V DAC each set around 50mA ballast current (27R resistor from Adj to output). Use a Transformer 18V 1A + 18V 1A.
2) 2 X 12V (seperate supplies only joined to for +/-12V at the Servo Motor Drivers Starground)
The regulators are a little less critical, 78XX/79XX types should suffice if on a budget. Due the potentially quite high burst currents when the laser focuses etc the supply should be fairly solid, so put a substantial value capacitor before and after the Regulator. Again, adding some "Ballast resistors" to draw extra current can be a good idea. A 12V + 12V 1.5A Transformer or bigger would be a good idea.
3) 1 X 5V for the Digital Player sections (Decoder, Digital filter etc.). If you use standard type transformers with two secondaries you could use the second secondary to power a clock module. The rest of the supplies I'd arrange similar to those for the +/-15V Analog supply.
I hope this helps.
If you need to due to insufficient space inside the player, you can install all the supplies detailed above in an external case and add a Sub-D25 connector on the back of the player which allows with a short circuit plug the player to run of the internal supply but which otherwise breaks the supply lines to Analogue stage, DAC, Servo and Digital circuitry and connects the external supply.
Sayonara
Konnichiwa,
Fin said:
Don't bother. You really need to add a few extra transformers (they need no be huge size) and build seperate supplies. Working of the original shared supplies is just by far too much compromise. Leave the original supplies o power only Display/Control uP & Remote control circuitry, all other supplies should be build new from scratch and be galvanically insulated, so you can break up several bad ground loops and feed each and every section optimally.
For Players of the CD-720/21/22/23 ilk (to which yous also belongs IIRC) I'd implement as follows:
1) 2 X 15V (seperate supplies only joined to for +/-15V at the DAC's Starground) with a further +15V tapped off the positive supply to null offset and +5V tapped off the positive supply to power the DAC.
All regs. in this supply should be at least LM317 better LT/LM1085, adj pin bypassed and good "audio grade" output capacitor, lower than common value "set" resistor chains to draw at least 50mA standing current through the regulator (more is better within reason).
Diodes Schottky, PSU Cap's small value and raw voltages fairly high, ideally with some added RC filtering, something like 3R3 470uF - 10R - 4,700uF - Reg with a 18V transformer and schottky bridge and the regulators (+15V Op-Amp, 15V Offset Null, +5V DAC each set around 50mA ballast current (27R resistor from Adj to output). Use a Transformer 18V 1A + 18V 1A.
2) 2 X 12V (seperate supplies only joined to for +/-12V at the Servo Motor Drivers Starground)
The regulators are a little less critical, 78XX/79XX types should suffice if on a budget. Due the potentially quite high burst currents when the laser focuses etc the supply should be fairly solid, so put a substantial value capacitor before and after the Regulator. Again, adding some "Ballast resistors" to draw extra current can be a good idea. A 12V + 12V 1.5A Transformer or bigger would be a good idea.
3) 1 X 5V for the Digital Player sections (Decoder, Digital filter etc.). If you use standard type transformers with two secondaries you could use the second secondary to power a clock module. The rest of the supplies I'd arrange similar to those for the +/-15V Analog supply.
I hope this helps.
If you need to due to insufficient space inside the player, you can install all the supplies detailed above in an external case and add a Sub-D25 connector on the back of the player which allows with a short circuit plug the player to run of the internal supply but which otherwise breaks the supply lines to Analogue stage, DAC, Servo and Digital circuitry and connects the external supply.
Sayonara
carlosfm said:Fin, the PSU of the analog stage (+/- 15v) may not have enough current for the mechanism.
They are usually very weak, some 100ma or even less.
Rotellian said:Fin, Carlos is right - that would be my main concern. (however thats not to say that the now redundant +/- supply cant be used to supply other things
Thanks Carlos and Rotellian.
This shows how much I know/don't know!!!!
I thought that the taps off the transformer for the +/-15V supplies were on the same winding as the +/-10V supplies - with the +/-10V just tapped off closer to the centre tap. Therefore, the +/-15V could supply the same total VA as the +/-10V supply, but at a higher voltage. If this is the case, looking at the unregulated values, the higher voltage taps could supply less than half the current of the +/-10V taps - so you are probably right. The winding wire after the 10V taps might also be thinner - making things even worse.
So - I need to go further!
Whow! Thanks Thorsten. This is excellent!
Sorry that I am not familiar with all of the terminology and I need remedial lessons.
This is what I want to achieve - but now realise that it is worthwhile going further than originally planned.
I have a Tent Clock with its own supply and transformer - so that's one item taken care of.
I have two pcb mount 25VA torriodal transformers. One has 2*22V secondaries which I intended to use for the +/- 15V supply (allowing plenty of headroom for pre-regulation etc). The other has 2*15V secondaries, which I intended to use for two separate +5V supplies for the digital and analogue side of the dac. Maybe I need to rearrange my ideas and get better use out of these units - and get a few more as well.
The CD624 is of similar design to players like the Philips CD850, Marantz CD41, CD52 etc. Not sure if the CD72x is similar.
I tried to attach a picture of the inside of the 624 but it won't attach.
Actually, it is quite similar to Bricolo's CD62 shown in this thread:
Power supply and PCB of CD62
except for the extra audio boards on the left.
Does this mean that everything (including the windings and rectifiers) should be kept separate until connected to the main pcb?
The centre tap from the tf (or is it the ground pin from the regulator) is connected to the groundplane (to a junction/link that connects the groundplane to the solder side of the board) as close to the dac as possible?
What is null offset?
Is the +5V supply tapped off before or after the the +15V reg?
Good "audio grade" output capacitor:- Elna Starget? or do I need something like a Black Gate?
Please explain: "lower than common value "set" resistor chains."
OK, I've got Schotkky diodes: Microsemi MBR150, MBR1100, MBR350 & MBR3100.
What type of capacitors? Panasonic FC, Elna RJH, Nichicon PM, Rubycon ZA/ZL/YXF/YXA? I can get all the usuals from RS Components and Farnell.
Should I keep the same setup with the 22V 568mA + 22V 568mA tf?
Is there any advantage to separating things even further by using a separate tf for the +5V supply for the dac - like in my original plan?
"Substantial value capacitor": 4,700uf or 6,800uf or bigger?
Again, what type?
"The rest of the supplies": Does this mean everything that is powered by this tf - ie. including decoder, digital filter etc.
How much current is required here (what size tf)?
This is a great help and most appreciated.
Space is not a problem - the 624 is one of those "mainly filled with air" CDPs. There is a large empty space to the left side of the pcb (which was used for separate audio boards in the more expensive versions) that should easily house a few transformers.
So it looks like I should get at least one, and maybe two more transformers to really do things right.
Whow! Thanks Thorsten. This is excellent!
Sorry that I am not familiar with all of the terminology and I need remedial lessons.
Kuei Yang Wang said:
This is what I want to achieve - but now realise that it is worthwhile going further than originally planned.
I have a Tent Clock with its own supply and transformer - so that's one item taken care of.
I have two pcb mount 25VA torriodal transformers. One has 2*22V secondaries which I intended to use for the +/- 15V supply (allowing plenty of headroom for pre-regulation etc). The other has 2*15V secondaries, which I intended to use for two separate +5V supplies for the digital and analogue side of the dac. Maybe I need to rearrange my ideas and get better use out of these units - and get a few more as well.
Kuei Yang Wang said:
The CD624 is of similar design to players like the Philips CD850, Marantz CD41, CD52 etc. Not sure if the CD72x is similar.
I tried to attach a picture of the inside of the 624 but it won't attach.
Actually, it is quite similar to Bricolo's CD62 shown in this thread:
Power supply and PCB of CD62
except for the extra audio boards on the left.
Kuei Yang Wang said:
Does this mean that everything (including the windings and rectifiers) should be kept separate until connected to the main pcb?
The centre tap from the tf (or is it the ground pin from the regulator) is connected to the groundplane (to a junction/link that connects the groundplane to the solder side of the board) as close to the dac as possible?
What is null offset?
Is the +5V supply tapped off before or after the the +15V reg?
Kuei Yang Wang said:
Good "audio grade" output capacitor:- Elna Starget? or do I need something like a Black Gate?
Please explain: "lower than common value "set" resistor chains."
Kuei Yang Wang said:
OK, I've got Schotkky diodes: Microsemi MBR150, MBR1100, MBR350 & MBR3100.
What type of capacitors? Panasonic FC, Elna RJH, Nichicon PM, Rubycon ZA/ZL/YXF/YXA? I can get all the usuals from RS Components and Farnell.
Should I keep the same setup with the 22V 568mA + 22V 568mA tf?
Is there any advantage to separating things even further by using a separate tf for the +5V supply for the dac - like in my original plan?
Kuei Yang Wang said:
"Substantial value capacitor": 4,700uf or 6,800uf or bigger?
Again, what type?
Kuei Yang Wang said:
"The rest of the supplies": Does this mean everything that is powered by this tf - ie. including decoder, digital filter etc.
How much current is required here (what size tf)?
Kuei Yang Wang said:
This is a great help and most appreciated.
Space is not a problem - the 624 is one of those "mainly filled with air" CDPs. There is a large empty space to the left side of the pcb (which was used for separate audio boards in the more expensive versions) that should easily house a few transformers.
So it looks like I should get at least one, and maybe two more transformers to really do things right.
Konnichiwa,
You would have a hard time finding such an item off the shelf. Now if you have a transformer specially made, then make it ONE really good ones, implementing ALL the various features known to make "good sound" (in short, distributed airgap in the core, low flux operation [more turns than usual] fully symmetrical primaries and secondaries for high CMRR, electrostatic screens all over an finally primaries wound over the secondaries [secondaries inside the transformer]). But that would cost a lot more than a bunch of small transformers off shelf and would likely be a substantial hunk of Iron. Few people may wish to be as commited as that to something like a CD-723.... ;-)
Local 431 Shunts with "tripplet" C decoupling for digtal circuits remains my favourite. For Audio Stages with Op-Amp's using a 3-dimensional circuit arrangement with further Op-Amp's of the same type providing the regulation is my preferred solution, together with battery options everywhere. There are many options and approaches. What I have suggested here uses just generic, off the shelf parts and generic circuitry that is well documented and nearly "fool proof" to implement. It is a starting point.
My personal take on a CD-723 in this day and age would include my Valve Analogue stage, a massive outboard supply with huge torroids in "field cancelling" arrangements with low flux hookup or custom made transformers and so on. Except, in this day and age I have more fun messing about with Shanling players, which look infinitly better and have a very solid build too. ;-)
Sayonara
Bricolo said:That makes a lot of transformers
Would a single one with many secondaries be much worse?
You would have a hard time finding such an item off the shelf. Now if you have a transformer specially made, then make it ONE really good ones, implementing ALL the various features known to make "good sound" (in short, distributed airgap in the core, low flux operation [more turns than usual] fully symmetrical primaries and secondaries for high CMRR, electrostatic screens all over an finally primaries wound over the secondaries [secondaries inside the transformer]). But that would cost a lot more than a bunch of small transformers off shelf and would likely be a substantial hunk of Iron. Few people may wish to be as commited as that to something like a CD-723.... ;-)
Bricolo said:
Local 431 Shunts with "tripplet" C decoupling for digtal circuits remains my favourite. For Audio Stages with Op-Amp's using a 3-dimensional circuit arrangement with further Op-Amp's of the same type providing the regulation is my preferred solution, together with battery options everywhere. There are many options and approaches. What I have suggested here uses just generic, off the shelf parts and generic circuitry that is well documented and nearly "fool proof" to implement. It is a starting point.
My personal take on a CD-723 in this day and age would include my Valve Analogue stage, a massive outboard supply with huge torroids in "field cancelling" arrangements with low flux hookup or custom made transformers and so on. Except, in this day and age I have more fun messing about with Shanling players, which look infinitly better and have a very solid build too. ;-)
Sayonara
Re: So - I need to go further!
Konnichiwa,
Torroids usually have high leakage due to the close coupling of secondary and primary.
Sounds fine to me.
Based on what you send me, the 624 is an earlier unit. I am not sure which DAC is used. My comments are very specific to the group of Machines including the Philips CD710/11/13/20/21/23 (probably also 12/22 and the Marantz CD36/46 et al. These have a TDA1545 DAC and similar chipsets and circuits.
That is how I prefer to do it. It keeps all current loops clearly defined.
You really need to tailor what you are doing to the specific Player and circuits used.
The 723 and related machine s(see above) all use a DAC with only one +5V Supply. This leads to an output offset of around 3V on the Op-Amp used in the analog stage. By feeding a certain current (around 1mA for 2mA full scale current) this offset can be eliminated, allowing the output coupling capacitor to be dispensed with.
I used before.
Elna Silmic preferred, Cerafine then Starget.
Look at the LM317 Datasheet. By reducing the two resistors that set the Voltage you draw additional current.
I personally would likely use the Rubycon ZA/ZL at this point in time.
Can't see why not.
If you look at the current loops you can see that the DAC and Analogue stage can use a combined supply, however separating the DAC out further will reduce any ambiguity of the ground routing, as only signal and not supply currents then flow through the ground. So yes, separating out more supplies will be beneficial. If you can having sepertae transformers andsupplies/regulators for each supply node is ideal, however it can become a little complex.
Please read the datasheets relevant to the regulators you will be using. They are usually quite clear as to what load capacitors are usable and what the stability limits are in terms of ESR etc....
No, it means the actual circuit for these supplies. As they are not subject to high burst currents using a RCRC filter priot to the regulators is recommended.
That depends upon the players chipset, please peruse the Datasheets for your players chipset to find out.
Sayonara
Konnichiwa,
Fin said:
Torroids usually have high leakage due to the close coupling of secondary and primary.
Fin said:
Sounds fine to me.
Fin said:
Based on what you send me, the 624 is an earlier unit. I am not sure which DAC is used. My comments are very specific to the group of Machines including the Philips CD710/11/13/20/21/23 (probably also 12/22 and the Marantz CD36/46 et al. These have a TDA1545 DAC and similar chipsets and circuits.
Fin said:
That is how I prefer to do it. It keeps all current loops clearly defined.
Fin said:
You really need to tailor what you are doing to the specific Player and circuits used.
Fin said:
The 723 and related machine s(see above) all use a DAC with only one +5V Supply. This leads to an output offset of around 3V on the Op-Amp used in the analog stage. By feeding a certain current (around 1mA for 2mA full scale current) this offset can be eliminated, allowing the output coupling capacitor to be dispensed with.
Fin said:
I used before.
Fin said:
Elna Silmic preferred, Cerafine then Starget.
Fin said:
Look at the LM317 Datasheet. By reducing the two resistors that set the Voltage you draw additional current.
Fin said:
I personally would likely use the Rubycon ZA/ZL at this point in time.
Fin said:
Can't see why not.
Fin said:
If you look at the current loops you can see that the DAC and Analogue stage can use a combined supply, however separating the DAC out further will reduce any ambiguity of the ground routing, as only signal and not supply currents then flow through the ground. So yes, separating out more supplies will be beneficial. If you can having sepertae transformers andsupplies/regulators for each supply node is ideal, however it can become a little complex.
Fin said:
Please read the datasheets relevant to the regulators you will be using. They are usually quite clear as to what load capacitors are usable and what the stability limits are in terms of ESR etc....
Fin said:
No, it means the actual circuit for these supplies. As they are not subject to high burst currents using a RCRC filter priot to the regulators is recommended.
Fin said:
That depends upon the players chipset, please peruse the Datasheets for your players chipset to find out.
Sayonara
Hi KYW,
Just wanted to check a few things before ordering the bits:
1) I/V + DAC supply:
I have a 0-15, 0-15, 20VA Avel Lindberg toroidal transformer - the potted, black type with interwinding screen, similar to Farnel part 178-155. Is this any good for the analogue supplies or would I be better off with Farnell: 899-707 which is 0-18, 0-18 30VA, EI type with claimed low interwinding capacitance (hopefully indicating dual bobbin construction - though the data sheet does not say).
2) VAM1201 supply:
Needs +12V @ > 500mA and +5V @ 150mA.
Use the triple o/p version - Farnell: 992-215, 0-14.5 @ 0.45A, 0-14.5 @ 0.45A, 0-9 @1.5A , 30VA. I would parallel the 14.5V windings for the +12V supply and use the third winding to provide the +5V supply.
3) Display and uP:
Use the existing power supply
4) Clock PSU (assuming Tent XO is fitted)
Need +5V @ 100mA
Is it OK to use the same 0-9 winding used for the VAM1201 (since the ground ref is the same) and an LC filter then a TL431 EF supply (as per G Tent's website instructions)?
One more question - I have a Trichord clock 4 @ 16.9344MHz - is it viable to use this with a 74AC109 to divide by 2 for the clock?
Thanks for all the excellent help so far.
Dave
Just wanted to check a few things before ordering the bits:
1) I/V + DAC supply:
I have a 0-15, 0-15, 20VA Avel Lindberg toroidal transformer - the potted, black type with interwinding screen, similar to Farnel part 178-155. Is this any good for the analogue supplies or would I be better off with Farnell: 899-707 which is 0-18, 0-18 30VA, EI type with claimed low interwinding capacitance (hopefully indicating dual bobbin construction - though the data sheet does not say).
2) VAM1201 supply:
Needs +12V @ > 500mA and +5V @ 150mA.
Use the triple o/p version - Farnell: 992-215, 0-14.5 @ 0.45A, 0-14.5 @ 0.45A, 0-9 @1.5A , 30VA. I would parallel the 14.5V windings for the +12V supply and use the third winding to provide the +5V supply.
3) Display and uP:
Use the existing power supply
4) Clock PSU (assuming Tent XO is fitted)
Need +5V @ 100mA
Is it OK to use the same 0-9 winding used for the VAM1201 (since the ground ref is the same) and an LC filter then a TL431 EF supply (as per G Tent's website instructions)?
One more question - I have a Trichord clock 4 @ 16.9344MHz - is it viable to use this with a 74AC109 to divide by 2 for the clock?
Thanks for all the excellent help so far.
Dave
OK. Just to be sure I got it:
What you advice, is to use dual (or more) secondaries for each transformer. No center tap.
Separate rectifiers for each chip/function, but for those sharing the same transformer (the same secondary would be more correct) also share the same bridge and RC filtering. Is that it?
What you advice, is to use dual (or more) secondaries for each transformer. No center tap.
Separate rectifiers for each chip/function, but for those sharing the same transformer (the same secondary would be more correct) also share the same bridge and RC filtering. Is that it?
Konnichiwa,
Not sure. If there truely is an electrstatic screen, great, use the torroid.
I think this is a bit iffy, curent wise, on the 12V side.
Preferably not. Better get a 6V+6V torroid to get two seperate clean 5V supplies (one Decoder etc Digital, one clock) and one 6V+6V to get you a jucy +12V for the drive, each transformer can then be a similar 20VA Avel Lindberg thingeys....
Do you need an 8MHz clock? I thought the chipset in the 723 et al worked on 16.9... or 11.2... MHz.
Sayonara
Dave S said:
Not sure. If there truely is an electrstatic screen, great, use the torroid.
Dave S said:
I think this is a bit iffy, curent wise, on the 12V side.
Dave S said:
Preferably not. Better get a 6V+6V torroid to get two seperate clean 5V supplies (one Decoder etc Digital, one clock) and one 6V+6V to get you a jucy +12V for the drive, each transformer can then be a similar 20VA Avel Lindberg thingeys....
Dave S said:
Do you need an 8MHz clock? I thought the chipset in the 723 et al worked on 16.9... or 11.2... MHz.
Sayonara
Konnichiwa,
Yup. And the use only "N-Channel" regulators (high preformance are cheap and easy to get and have more sources, good negative regs on a chip are rare).
Yes, basically use as many seperate supplies as you care/dare to do. If you must share consider the ground routing and have ONE solid heavy ground wire and several seperate (regulated) supply lines, plus handle the final AC bypass locally at the consuming and, not at the regulator chip.
All this is much easier if you design all this in. Adding serious supplies later because the accountants did the player design always involves compromises.
Sayonara
Bricolo said:
Yup. And the use only "N-Channel" regulators (high preformance are cheap and easy to get and have more sources, good negative regs on a chip are rare).
Bricolo said:
Yes, basically use as many seperate supplies as you care/dare to do. If you must share consider the ground routing and have ONE solid heavy ground wire and several seperate (regulated) supply lines, plus handle the final AC bypass locally at the consuming and, not at the regulator chip.
All this is much easier if you design all this in. Adding serious supplies later because the accountants did the player design always involves compromises.
Sayonara
N-channel? LDOs
In fact, I'm still a student. So "as many as you care/dare" isn't a problem. "as many as you can afford" will be the limiting factor
Oh, something just came to my mind: what should we do with the player's groundplane?
BTW, thank you a lot for your explanations! That's great stuff
Konnichiwa,
Well, whatever takes your fancy, me getting old, being naturally conservative, better the devil you know and all would probably just use Nat Semi LM1085....
Go shopping in surplus places. Often nice and servicable transformers can be obtained from 12V AC or DC "plugtop" supplies you have bought along the lines of 10-Pack for 20 Euro or so.... They include the rectifier PCB and revervoir capacitor and can be dumped and potted in an external box, if you are in need of a "PhrugalPhile" solution.
Whatever is happening with todays students, in my days we came up with all sorts of interesting way to do neat stuff on the cheap.
Leave it as digital groundplne. Hopefully it will not carry much nose ciurrent, so just leave it and tell it to think "I'm ground" very hard.... ;-)
Sayonara
Bricolo said:N-channel? LDOs
Well, whatever takes your fancy, me getting old, being naturally conservative, better the devil you know and all would probably just use Nat Semi LM1085....
Bricolo said:In fact, I'm still a student. So "as many as you care/dare" isn't a problem. "as many as you can afford" will be the limiting factor
Go shopping in surplus places. Often nice and servicable transformers can be obtained from 12V AC or DC "plugtop" supplies you have bought along the lines of 10-Pack for 20 Euro or so.... They include the rectifier PCB and revervoir capacitor and can be dumped and potted in an external box, if you are in need of a "PhrugalPhile" solution.
Whatever is happening with todays students, in my days we came up with all sorts of interesting way to do neat stuff on the cheap.
Bricolo said:
Leave it as digital groundplne. Hopefully it will not carry much nose ciurrent, so just leave it and tell it to think "I'm ground" very hard.... ;-)
Sayonara
but I'm not sure how to connect the groundplane
All the ICs will have separate regs. Some will share the same secondary/bridge/rc filtering and so, the same "local ground". I think I should connect the bridge's ground near the most critical IC (the same principe as for the clock you explained some pages ago). Am I right?
Now... where do I connect all those grounds together (to keep them at the same voltage)? Will this be done with the groundplane?
So I connect the the "local grounds" to the most critical IC related to it, and this IC will still be also soldered to the groundplane?
This would make 2 things:
-connect all the "local grounds" together (placing them at the same potential)
-connect the ground pin of the non criticals ICs to the "local ground"
but I'm not sure the 2nd point will be correctly done this way
All the ICs will have separate regs. Some will share the same secondary/bridge/rc filtering and so, the same "local ground". I think I should connect the bridge's ground near the most critical IC (the same principe as for the clock you explained some pages ago). Am I right?
Now... where do I connect all those grounds together (to keep them at the same voltage)? Will this be done with the groundplane?
So I connect the the "local grounds" to the most critical IC related to it, and this IC will still be also soldered to the groundplane?
This would make 2 things:
-connect all the "local grounds" together (placing them at the same potential)
-connect the ground pin of the non criticals ICs to the "local ground"
but I'm not sure the 2nd point will be correctly done this way
Is this slightly different approach close enough?
Thanks again Thorsten.
This is obviously not good - but still better than using the original transformer, I assume (still a step in the right direction)?
Does this high leakage outweigh the benefits of the lower radiated magnetic field?
The CD624 is from around 1990/91, and the DAC - this is where you cringe - is an early Bitstream SAA7321. It is similar to that used in the Philips CD840, CD850, QED Digit, Rotel RCD965, Micromega Leader, Meridian 203...........(trying to put it in good company).
I know this is not your favourite, with its switched capacitor "sonic poison" and all - but it would be interesting to see what is possible - before thinking of installing a multibit.
Even though your comments are specific to a particular group of machines - do you think most of them can be applied in a generic way?
Your approach to separating transformers/supplies, etc. and the size of transformers and types of capacitors to use for various sections should still be similar, regardless of the CDP to which they are applied??
The SAA7321 uses six +5V supplies:- Digital Section, Crystal Oscillator, Analogue Left, Analogue Right, Analogue Logic and Reference voltage generator. So I assume I don't need to worry about "null offset"?
I was thinking of using the 2*15V tf to power the DAC chip only. One winding for the digital side and the other winding for the analogue side - with separate regulators for each supply pin. But maybe this is going a bit too far?
What do you think of using one winding for the analogue side of the dac and the other winding for the digital side of the dac and the decoder?
This would mean I only need one more tf to power the servos - or have I missed something?
Maybe there is a better way to to group these supplies - remembering that the filter is included in the dac?
So - very low impedance caps before the reg is desireable?
I'm still a little unclear about the following:-
You are suggesting a separate 5V supply for the Decoder and Digital Filter. Sould anything else be supplied from this?
"The rest of the supplies" and "the actual circuit for these supplies".
Still not sure what I should be looking at here?
Thanks again Thorsten.
Kuei Yang Wang said:
This is obviously not good - but still better than using the original transformer, I assume (still a step in the right direction)?
Does this high leakage outweigh the benefits of the lower radiated magnetic field?
The CD624 is from around 1990/91, and the DAC - this is where you cringe - is an early Bitstream SAA7321. It is similar to that used in the Philips CD840, CD850, QED Digit, Rotel RCD965, Micromega Leader, Meridian 203...........(trying to put it in good company).
I know this is not your favourite, with its switched capacitor "sonic poison" and all - but it would be interesting to see what is possible - before thinking of installing a multibit.
Even though your comments are specific to a particular group of machines - do you think most of them can be applied in a generic way?
Your approach to separating transformers/supplies, etc. and the size of transformers and types of capacitors to use for various sections should still be similar, regardless of the CDP to which they are applied??
The SAA7321 uses six +5V supplies:- Digital Section, Crystal Oscillator, Analogue Left, Analogue Right, Analogue Logic and Reference voltage generator. So I assume I don't need to worry about "null offset"?
I was thinking of using the 2*15V tf to power the DAC chip only. One winding for the digital side and the other winding for the analogue side - with separate regulators for each supply pin. But maybe this is going a bit too far?
What do you think of using one winding for the analogue side of the dac and the other winding for the digital side of the dac and the decoder?
This would mean I only need one more tf to power the servos - or have I missed something?
Maybe there is a better way to to group these supplies - remembering that the filter is included in the dac?
So - very low impedance caps before the reg is desireable?
I'm still a little unclear about the following:-
You are suggesting a separate 5V supply for the Decoder and Digital Filter. Sould anything else be supplied from this?
"The rest of the supplies" and "the actual circuit for these supplies".
Still not sure what I should be looking at here?
Konnichiwa,
Yup. If you share a supply then connect the "ground" part of this supply to the most critical/logical point. Often this is not the most sensitive IC, but the one producing the most noise....
Groundplane and PCB Ground. Do not forget, all the grounds are already connected, we ony add supplies to this to give better insulation from ground noise and noise crosstalk through the supplies.
This applies if you design a new piece of gear. In that case I would recommend having all the various supplies mounted on board, with reg's as close as possible to the IC Supplied and rectifiers/Filtering along one edge of the board.
Sayonara
Bricolo said:
Yup. If you share a supply then connect the "ground" part of this supply to the most critical/logical point. Often this is not the most sensitive IC, but the one producing the most noise....
Bricolo said:
Groundplane and PCB Ground. Do not forget, all the grounds are already connected, we ony add supplies to this to give better insulation from ground noise and noise crosstalk through the supplies.
Bricolo said:
This applies if you design a new piece of gear. In that case I would recommend having all the various supplies mounted on board, with reg's as close as possible to the IC Supplied and rectifiers/Filtering along one edge of the board.
Sayonara
Mechanical Mods
Thinking about a spruce plywood base for the CD723 - I notice that the VAM2101 mech attaches to and rests on 4 bits of bent chassis.
Is it best to remove these and attach the mech to the new base with 4 rigid pillars or some sort of flexible arrangement? I normally use elastomer discs (can't remember the brand but they are bright green - lovely!) under the player for a bit of decoupling and I'm not sure whether I should be going for rigidity and mass above this, or whether the mech needs decoupling from the rest of the CD player.
As usual any advice is much appreciated.
Dave
Thinking about a spruce plywood base for the CD723 - I notice that the VAM2101 mech attaches to and rests on 4 bits of bent chassis.
Is it best to remove these and attach the mech to the new base with 4 rigid pillars or some sort of flexible arrangement? I normally use elastomer discs (can't remember the brand but they are bright green - lovely!) under the player for a bit of decoupling and I'm not sure whether I should be going for rigidity and mass above this, or whether the mech needs decoupling from the rest of the CD player.
As usual any advice is much appreciated.
Dave
Nuuk - Yes, thanks, I had already seen your mods. I would prefer to keep it as a front loader - fits in my rack better.
General comment:
I'm really looking for some scientific explanation of why it should or should not be mass loaded / damped / rigid / sprung etc.
As an example, the Audionet CD player uses 2 layers of granite with damping in between - I would like to know exactly how this is done and why this works and if the same principles apply to the CD723.
KYWs electrical mods follow good engineering practice and scientific reasoning, with plenty of experience thrown in. I would like to see the same for the mechanical mods.
General comment:
I'm really looking for some scientific explanation of why it should or should not be mass loaded / damped / rigid / sprung etc.
As an example, the Audionet CD player uses 2 layers of granite with damping in between - I would like to know exactly how this is done and why this works and if the same principles apply to the CD723.
KYWs electrical mods follow good engineering practice and scientific reasoning, with plenty of experience thrown in. I would like to see the same for the mechanical mods.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.