As mentioned before I have got the -31 working and sounding lovely...however if I try to ditch the mainboard by taking the orange/grey connection loose from the transport
the sound suddenly becomes very distorted and a digital hiss en other unlistenably noise comes immediately....anyone got a clue what is happening here...
I have not powered the led ....so only 8 Volts to the transport..
BTW The -51 is beyond repair probably a chip.....
.....so I have got my first spare parts
the sound suddenly becomes very distorted and a digital hiss en other unlistenably noise comes immediately....anyone got a clue what is happening here...
I have not powered the led ....so only 8 Volts to the transport..
BTW The -51 is beyond repair probably a chip.....
.....so I have got my first spare parts
Re: shigaclone
I've tried 6 or 7 various types, best one (not only to my taste, several blind tests decided) in my setup is 10n MultiCap RTX; pls note other caps influence very much optimum value and type of this critical cap - almost no caps changed in my clone #5, only PS, all smd in place not removed except 2 pcs at PS input
quan said:
I've tried 6 or 7 various types, best one (not only to my taste, several blind tests decided) in my setup is 10n MultiCap RTX; pls note other caps influence very much optimum value and type of this critical cap - almost no caps changed in my clone #5, only PS, all smd in place not removed except 2 pcs at PS input
I've seen these panel mount receptacles and cordsets where I work and was wondering if there's similar stuff available for use for external power connectors for our shigaclone. Has anyone see similar stuff through Digikey, etc.?
Brad Harrison
Regards,
Dan
Brad Harrison
Regards,
Dan
Just got my RCEZ31. My first impressions:
RCEZ32 (the one kevink talks about) has a better DSP chip (and one much closer to original Shig) than the RCEZ31 (which benefits/suffers from Sanyo's integration program which is meant to reduce chip count for low-end designers).
The fact that LC78601RE has to devote significant die space to an integrated u-controller w/ display block, lcd driver, and control processing means corners were cut in other areas and RFI interference between the blocks will be more common. I would rather see the influence of the display and control processing in the signal-processing block minimized by separate physical location, supplies, and decoupling of two separate packages. U-controllers are some of the nastiest sources of noise in a digital system...
On the other hand, what is nice about the RCEZ31 (and which improves on Shig even) is the connection between laser pickup and ASP is made as short as possible, with the ASP located directly under the laser sled. Note however that reducing the distance here increases the influence of spurious radiation and cares should be taken to shield the two blocks with shield lines or anti-EFM material (in early Flatfish it was a ground shield around the eight-lead harness carrying the pickup photodiode and track error signals). Note also that the direct connection between main pcb and laser sled mechanism means that there will be mechanical disturbances created by the servo motors on a pcb stuffed with microphonic ceramic chip caps and digital ICs. I would strongly advice detaching the pcb from the laser sled assembly, and making the connection to the motors via short wires, rather than letting the motors have their ways with that pcb. I would also recommend removing that plastic housing guard over the laser pickup--less is better here--and (unlike 47 Lab Shig) mounting adjustable spikes for the CD player directly on the four stand-offs that support the laser assembly unit, with some added mass, rather than spiking a traditional chassis at some point located a distance from the stand-offs or placing the whole thing on some wood/metal/ceramic plinth. The resonances from the CD mechanism need to leave it as directly as possible, and there is no more direct route than the stand-offs. Also, the manufacture of a precise delrin platter/clamp which fits tightly to the motor shaft without need of any bonding substance is also very desireable.
One thing: do not waste your time. The ASP and DSP will present a bottleneck in terms of spec and number, but have little to do with sound quality: I spent hours playing with different RF amp configs, and putting high-quality passive components into the PLL clock generation circuit (two years of tinkering with Caddocks and BGs on an old Flatfish, where the board was through-hole and parts were easy to swap) to no avail. Besides mechanical construction (which is very important), the biggest bottleneck for a system like this will be the supply regulation, grounding and decoupling. Indeed, some regulators are better tha others, and the 5V regulator sitting inside the MM1469 has at best 75dB rejection at 1kHz during steady state operation (they do anything to optimize their specs); during heavy servo operation, it is far worse, and there is quite alot of noise of all frequencies modulated into the 5V reg block, esp. during TOC and track skip operations. Being a linear reg means that it cannot filter out HF and RF noise, but passes it through: result is a white, bodyless presentation, without fundamental tones (my main criticism of 47 Lab digital players). I won't criticize JVC: budget is budget. It cost me Eu39! Their player is fast, and rhythmic. But the PSRR at 10MHz or at 100MHz is next to nil. The noise will get into your chips, get stored in parasitic capacitances, released a little later, generate offsets, modulate or be modulated by your RF digital signals, generating all kinds of intermod products, noise modulations, etc. etc. My advice: Do not use MM1469 as your final reg. Using the LM7808 as a pre-reg for the MM reg is only half a solution. Using a low-noise, low-output-impedance 5V reg with good dynamic performance (like the ALWSR super-reg or other) for your ASP/DSP combo (and another separate one for the internal 1-bit DAC if you are using the analogue out option) while adding some series resistance before good local decoupling will suprise you even further how good this chip combination can sound.
RCEZ32 (the one kevink talks about) has a better DSP chip (and one much closer to original Shig) than the RCEZ31 (which benefits/suffers from Sanyo's integration program which is meant to reduce chip count for low-end designers).
The fact that LC78601RE has to devote significant die space to an integrated u-controller w/ display block, lcd driver, and control processing means corners were cut in other areas and RFI interference between the blocks will be more common. I would rather see the influence of the display and control processing in the signal-processing block minimized by separate physical location, supplies, and decoupling of two separate packages. U-controllers are some of the nastiest sources of noise in a digital system...
On the other hand, what is nice about the RCEZ31 (and which improves on Shig even) is the connection between laser pickup and ASP is made as short as possible, with the ASP located directly under the laser sled. Note however that reducing the distance here increases the influence of spurious radiation and cares should be taken to shield the two blocks with shield lines or anti-EFM material (in early Flatfish it was a ground shield around the eight-lead harness carrying the pickup photodiode and track error signals). Note also that the direct connection between main pcb and laser sled mechanism means that there will be mechanical disturbances created by the servo motors on a pcb stuffed with microphonic ceramic chip caps and digital ICs. I would strongly advice detaching the pcb from the laser sled assembly, and making the connection to the motors via short wires, rather than letting the motors have their ways with that pcb. I would also recommend removing that plastic housing guard over the laser pickup--less is better here--and (unlike 47 Lab Shig) mounting adjustable spikes for the CD player directly on the four stand-offs that support the laser assembly unit, with some added mass, rather than spiking a traditional chassis at some point located a distance from the stand-offs or placing the whole thing on some wood/metal/ceramic plinth. The resonances from the CD mechanism need to leave it as directly as possible, and there is no more direct route than the stand-offs. Also, the manufacture of a precise delrin platter/clamp which fits tightly to the motor shaft without need of any bonding substance is also very desireable.
One thing: do not waste your time. The ASP and DSP will present a bottleneck in terms of spec and number, but have little to do with sound quality: I spent hours playing with different RF amp configs, and putting high-quality passive components into the PLL clock generation circuit (two years of tinkering with Caddocks and BGs on an old Flatfish, where the board was through-hole and parts were easy to swap) to no avail. Besides mechanical construction (which is very important), the biggest bottleneck for a system like this will be the supply regulation, grounding and decoupling. Indeed, some regulators are better tha others, and the 5V regulator sitting inside the MM1469 has at best 75dB rejection at 1kHz during steady state operation (they do anything to optimize their specs); during heavy servo operation, it is far worse, and there is quite alot of noise of all frequencies modulated into the 5V reg block, esp. during TOC and track skip operations. Being a linear reg means that it cannot filter out HF and RF noise, but passes it through: result is a white, bodyless presentation, without fundamental tones (my main criticism of 47 Lab digital players). I won't criticize JVC: budget is budget. It cost me Eu39! Their player is fast, and rhythmic. But the PSRR at 10MHz or at 100MHz is next to nil. The noise will get into your chips, get stored in parasitic capacitances, released a little later, generate offsets, modulate or be modulated by your RF digital signals, generating all kinds of intermod products, noise modulations, etc. etc. My advice: Do not use MM1469 as your final reg. Using the LM7808 as a pre-reg for the MM reg is only half a solution. Using a low-noise, low-output-impedance 5V reg with good dynamic performance (like the ALWSR super-reg or other) for your ASP/DSP combo (and another separate one for the internal 1-bit DAC if you are using the analogue out option) while adding some series resistance before good local decoupling will suprise you even further how good this chip combination can sound.
Hi James please to meet you
I definately will try all the advice myself and see if I come to the same conclusion as Peter, and I will mail the results/ effects it has in my audiosystem for others...
The way the music sounds as you described before "your" mods is for me very recognizable...
I definately will try all the advice myself and see if I come to the same conclusion as Peter, and I will mail the results/ effects it has in my audiosystem for others...
The way the music sounds as you described before "your" mods is for me very recognizable...
Hi Peter,
Your addition of a LM7808 pre-reg brings performance enhancements, adds safety features and benefits the circuit with its internal protection circuitry. National makes some good sounding devices. A single LM7808 was present on the stock Flatfish/Shig bolted to a large heatsink. It outperformed some other reg combinations in areas. I got greedy for more detail and control.
On Shig I got highest resolution (my ultimate goal) with 5 separate regs:
1. LM7812 on the motor drive chip (it had voltage tolerance of 13VDC I believe). I disabled the onboard 5V reg on the motor drive chip, and cut the supply traces to the ASP/DSP/uController, and added a larger heatsink.
2. and 3. I used two ALWSR 5V super-regs with MOSFET-based Vbe pre-reg (described in pink fish media diy forum) for (a) the ASP and DSP and (b) the u-controller. (The stock ALWSR is excellent, but being a linear reg, it does not reach high enough into MHz area, and modding it with a gyrator is recommended.)
4. A separate 5V jfet-based TeddyReg (again, see pink fish media diy) located close to the DSP chip for powering the internal DAC (yes, I used the analog out option).
5. A separate supply module called the "flea" (pink fish media diy) for a Tent clock.
What I wanted in cases 2,3,4, 5 was a wide-band RF filter with high line rejection to attenuate noise present on the supply lines from other digital components and rectification... Unfortunately I have alot of nasty digital audio equipment working with high clock frequencies and line rejection is a big issue for squeezing more performance out of a CD player for me. For others this approach might not bring as much performance enhancement... nor even be to their taste. The resulting sound is somewhat analytic and precise, but not bright or etched. Bass is very powerful, and fundamental tones are much more present than with stock set-up. One word of caution to those pursuing a multiple reg approach however: the timing of the start up of different regs must be properly adjusted to avoid latch up because of an absent clock signal or such. Symptom will be no TOC or a servo spinning out of control upon start up.
If the manufacture of a CD platter/clamp for the motor spindle can be arranged through this forum, this would be a worthwhile group-buy. Count me in.
Will post more impressions as I chug along with my EZ31. If anyone knows where to get EZ32s in Europe, please PM me.
Your addition of a LM7808 pre-reg brings performance enhancements, adds safety features and benefits the circuit with its internal protection circuitry. National makes some good sounding devices. A single LM7808 was present on the stock Flatfish/Shig bolted to a large heatsink. It outperformed some other reg combinations in areas. I got greedy for more detail and control.
On Shig I got highest resolution (my ultimate goal) with 5 separate regs:
1. LM7812 on the motor drive chip (it had voltage tolerance of 13VDC I believe). I disabled the onboard 5V reg on the motor drive chip, and cut the supply traces to the ASP/DSP/uController, and added a larger heatsink.
2. and 3. I used two ALWSR 5V super-regs with MOSFET-based Vbe pre-reg (described in pink fish media diy forum) for (a) the ASP and DSP and (b) the u-controller. (The stock ALWSR is excellent, but being a linear reg, it does not reach high enough into MHz area, and modding it with a gyrator is recommended.)
4. A separate 5V jfet-based TeddyReg (again, see pink fish media diy) located close to the DSP chip for powering the internal DAC (yes, I used the analog out option).
5. A separate supply module called the "flea" (pink fish media diy) for a Tent clock.
What I wanted in cases 2,3,4, 5 was a wide-band RF filter with high line rejection to attenuate noise present on the supply lines from other digital components and rectification... Unfortunately I have alot of nasty digital audio equipment working with high clock frequencies and line rejection is a big issue for squeezing more performance out of a CD player for me. For others this approach might not bring as much performance enhancement... nor even be to their taste. The resulting sound is somewhat analytic and precise, but not bright or etched. Bass is very powerful, and fundamental tones are much more present than with stock set-up. One word of caution to those pursuing a multiple reg approach however: the timing of the start up of different regs must be properly adjusted to avoid latch up because of an absent clock signal or such. Symptom will be no TOC or a servo spinning out of control upon start up.
If the manufacture of a CD platter/clamp for the motor spindle can be arranged through this forum, this would be a worthwhile group-buy. Count me in.
Will post more impressions as I chug along with my EZ31. If anyone knows where to get EZ32s in Europe, please PM me.
http://www.at-view.co.uk/alwsr.htm
Hello Stan I do not know if this is okay to mention but please moderator remove it if it is uncalled for...
Andrew L. Weekes is the guys name..hence the company.....until James mentioned him I honestly never heard of him...
I am willing to try everything out because my present JVC/Sanyo set-up is still not in the league of the modded Wadia....
I must admit the blackgates are not mounted on de motorboard yet even as the mechanical construction...but they must be close to a miracle...
I will also try the 392/92 resistor combination....but as Peter advised I take my time after each change...
On the other hand there are a few tracks in which it most definitely already outperforms my Wadia...
So for me it is actually a thrilling process....
Hello Stan I do not know if this is okay to mention but please moderator remove it if it is uncalled for...
Andrew L. Weekes is the guys name..hence the company.....until James mentioned him I honestly never heard of him...
I am willing to try everything out because my present JVC/Sanyo set-up is still not in the league of the modded Wadia....
I must admit the blackgates are not mounted on de motorboard yet even as the mechanical construction...but they must be close to a miracle...
I will also try the 392/92 resistor combination....but as Peter advised I take my time after each change...
On the other hand there are a few tracks in which it most definitely already outperforms my Wadia...
So for me it is actually a thrilling process....
Attachments
Erik van Voorst said:http://www.at-view.co.uk/alwsr.htm
Hello Stan I do not know if this is okay to mention but please moderator remove it if it is uncalled for...
Andrew L. Weekes is the guys name..hence the company.....until James mentioned him I honestly never heard of him...
I am willing to try everything out because my present JVC/Sanyo set-up is still not in the league of the modded Wadia....
I must admit the blackgates are not mounted on de motorboard yet even as the mechanical construction...but they must be close to a miracle...
I will also try the 392/92 resistor combination....but as Peter advised I take my time after each change...
On the other hand there are a few tracks in which it most definitely already outperforms my Wadia...
So for me it is actually a thrilling process....
I think you will find changing the stock caps out for BGs and doing the other recommended mods is a miracle. The difference the modifications I made to the RCEZ32 based version I use were just astounding - nothing at all subtle about those changes.
The mechanical end of things is very important as well. I followed Peter's recommendations pretty closely, barring some differences due to the additional mpu board - and I would have to say he was right on all counts.
Did you find out what you did wrong that resulted in the noise issue you mentioned a few posts back? (Ground issue?)
Erik van Voorst said:
Thanks Eric, I also noticed that peranders makes a small super reg board 1 1/4" X 1 3/4" if space is a concern, the pre regulator supply is off board or one could tap into the existing 5V supply.
http://sjostromaudio.com/joomla/index.php?option=com_content&task=view&id=55&Itemid=27
kevinkr said:
No not yet...I have got an enormous help from an audiofriend so he will look into it whenever he can make time ...so we can ditch the board and have more space for things to come...hahahahaha
At the moment I am listening LOADS of cd's to form an opinion about the upgrade the 8V power supply change and that is
downright very good...