Cool. Good to know...
By the way, both mechas from the ebay link I posted had bent spindles... I just gave a more thorough try to the second one and it is slightly bent too... Damn... I wish I had better tried both of them before opening a case on ebay... The vendor already sent me a replacement, but still...
By the way, the second one did not have that weird motor sound, but it still does not seem to produce such a clear sound as Tibi's did. Perhaps I need to check the eye pattern as well. Different batches could be configured differently.
Tibi, any news on your extra mechas' availability? I think I m done with random sources...
By the way, both mechas from the ebay link I posted had bent spindles... I just gave a more thorough try to the second one and it is slightly bent too... Damn... I wish I had better tried both of them before opening a case on ebay... The vendor already sent me a replacement, but still...
By the way, the second one did not have that weird motor sound, but it still does not seem to produce such a clear sound as Tibi's did. Perhaps I need to check the eye pattern as well. Different batches could be configured differently.
Tibi, any news on your extra mechas' availability? I think I m done with random sources...
Ah, to be in the EU or nearby. I spoke with Tibi and the shipping cost to get new mechs from him to the U.S. was not cost effective. I just got two new ones from a dealer here in the U.S. I haven't tested them yet but they came very well packed, each in their own box and those carefully packed in a larger shipping box. I have my fingers crossed.
By the way, I tried to do some mods on Tibi's PSU.
First of all, I removed all the parts that were not specified in the initial build. So basically tried to leave only the diodes, the 1000uF and the 2200uF caps and the 7808 (Did no tests at this point).
I also tried to remove the caps and install some Nichicon FWs I have laying around, but it seems that the big copper plains are a big PITA... Trying to clean the holes (pump, wick, max temp... nothing seem to manage to get both pcb sides melting hot...) and I eventually managed to lift one of the pads...
I am so angry and frustrated now, I am a moving accident waiting to happen, so I left it and I will rebuild it tomorrow.
On the stupid trials row, I also just tried a salas psu I have laying around that is tuned to 7,9V and 450mA max...
It seems that the motors really really need the juice to get the disc spinning... They couldn't even get it to read toc and the salas leds where flashing like a Christmas tree...
So I basically gave up for today... Some days are just... bad....
First of all, I removed all the parts that were not specified in the initial build. So basically tried to leave only the diodes, the 1000uF and the 2200uF caps and the 7808 (Did no tests at this point).
I also tried to remove the caps and install some Nichicon FWs I have laying around, but it seems that the big copper plains are a big PITA... Trying to clean the holes (pump, wick, max temp... nothing seem to manage to get both pcb sides melting hot...) and I eventually managed to lift one of the pads...
I am so angry and frustrated now, I am a moving accident waiting to happen, so I left it and I will rebuild it tomorrow.
On the stupid trials row, I also just tried a salas psu I have laying around that is tuned to 7,9V and 450mA max...
It seems that the motors really really need the juice to get the disc spinning... They couldn't even get it to read toc and the salas leds where flashing like a Christmas tree...
So I basically gave up for today... Some days are just... bad....
You are not alone. I was removing the infamous Samxon caps and also managed to lifted a pad
"Infamous Samxon" measure quite well on my Blue-ESR meter. I suggest to leave them in place unless you got some oscon's or polymer caps at the same value.
Regards,
Tibi
After some delays caused by severe winter and power outages, here is the summary of my output resistor tests.
0. The worst case is having no resistors at all.
1. Installing virtually any resistors in an L-pad configuration improves the sound clarity (pretty much any combination of resistor values between 70 and 400R will be an improvement).
2. The optimal L-pad seems to be close to 300R series, 100R shunt.
3. Adding another 100R to form a Pi-pad: 100R-300R-100R (shunt-series-shunt) brings further (minor) improvement in clarity.
4. Moving on to a 75-ohm Pi-pad (75-ohm input impedance, 75-ohm output impedance, 12dB attenuation), which consists of 125R-139R-125R brings a further improvement in clarity - quite significant when compared with the L-pad.
5. Finally, a 50-ohm input, 75-ohm output, 12dB Pi-pad: 72-114-157...
..yep, I do believe it is the Holy Grail of Shiga output terminations. The effect when moving from the 125-139-125 Pi-pad is a bit different from the previous changes though. Whereas all the previous "upgrades" could be described with a single word: clarity, I had some trouble naming what it is exactly that I prefer in this one. I instantly knew I liked it better, but could not quite pinpoint why until I did a lot of swapping back and forth. Eventually I settled on a conclusion that the sense of depth is enhanced, transients are sharper and (as a result) timbres are more true to life... You could say dynamic resolution is improved. This sound is more moving, with greater impact - while the previous Pi-pad sounds very similar in most respects, it is just not as exhilarating to listen to.
For anyone who has recently ordered resistors for a Pi-pad from me - I have changed your orders to this new Pi-pad.
On another news, regarding the motor caps: I confirm that caps between the + and - terminals are the worst (loss of bass and treble, flat sound), and that caps between the motor case and its terminals have similar, albeit much more benign effect. In other words: leave those motors alone!
An externally hosted image should be here but it was not working when we last tested it.
0. The worst case is having no resistors at all.
1. Installing virtually any resistors in an L-pad configuration improves the sound clarity (pretty much any combination of resistor values between 70 and 400R will be an improvement).
2. The optimal L-pad seems to be close to 300R series, 100R shunt.
3. Adding another 100R to form a Pi-pad: 100R-300R-100R (shunt-series-shunt) brings further (minor) improvement in clarity.
4. Moving on to a 75-ohm Pi-pad (75-ohm input impedance, 75-ohm output impedance, 12dB attenuation), which consists of 125R-139R-125R brings a further improvement in clarity - quite significant when compared with the L-pad.
5. Finally, a 50-ohm input, 75-ohm output, 12dB Pi-pad: 72-114-157...
..yep, I do believe it is the Holy Grail of Shiga output terminations
. The effect when moving from the 125-139-125 Pi-pad is a bit different from the previous changes though. Whereas all the previous "upgrades" could be described with a single word: clarity, I had some trouble naming what it is exactly that I prefer in this one. I instantly knew I liked it better, but could not quite pinpoint why until I did a lot of swapping back and forth. Eventually I settled on a conclusion that the sense of depth is enhanced, transients are sharper and (as a result) timbres are more true to life... You could say dynamic resolution is improved. This sound is more moving, with greater impact - while the previous Pi-pad sounds very similar in most respects, it is just not as exhilarating to listen to.For anyone who has recently ordered resistors for a Pi-pad from me - I have changed your orders to this new Pi-pad.
On another news, regarding the motor caps: I confirm that caps between the + and - terminals are the worst (loss of bass and treble, flat sound), and that caps between the motor case and its terminals have similar, albeit much more benign effect. In other words: leave those motors alone!
Nice one
Good to know about the motor caps...
I ll try the 300-100 L-Pad (dales) in place of the 75-75 one installed by Tibi and get back to you (R48 and R49 if I m not mistaken...).
I have also tried changing some other stuff in the kit and results are really big. I ll get back to you later today...
Good to know about the motor caps...
I ll try the 300-100 L-Pad (dales) in place of the 75-75 one installed by Tibi and get back to you (R48 and R49 if I m not mistaken...).
I have also tried changing some other stuff in the kit and results are really big. I ll get back to you later today...
I doubted the Samxon would leak like inside the PC right next to the heat source. Even 2 of them are the offeding CF series, the old issue may be fixed in the newer production. It just bugged me psychologically
The Samxon caps happen to be on the basic mod cap list of the original Shiga. So, I took the excuse to make the mod. FWIW,
- C13, C12 (bypass) = C952, C963 => Removed completely.
- C43, C42 (bypass) = C916, C917 => Replaced. Uncle Leon's mod guide said this is the most influential. Suggest parts, BG N 47uF/50 which I don't have. I happened to have a few BG PK 47/25. So, that's what I used.
- C10 = C953 => Repalced. Suggest parts, GB N 47uF/50. I used Silmic II 47/uf/50.
The result was an instant improvement of the detail and the clarity. I think this basic mod is well worth the time.
5. Finally, a 50-ohm input, 75-ohm output, 12dB Pi-pad: 72-114-157...
..yep, I do believe it is the Holy Grail of Shiga output terminations
Thanks so much for the great work and analyis, uncle_leon. Please look for a PM from me!
Busy day today... I managed to buy some wood, so I started planning on the chassis...
Now to the electronics part, I have tried a few stuff.
First of all, I changed the PSU of the kit with a p2p built one using 2x1000uF Panasonic FRs before and 1 after the 7808.
I have only MUR860 diodes so I used those for the rectification. (A friend suggested I gave these a try... D9E Russian Military Germanium Diode Genuine 20 | eBay)
The increase in clarity was huge even with freshly soldered and unused parts (not broken in).
At that time, I saw a piece of scrap wenge so I decided to add some damping to the entire PSU a la Eric and Jaroslaw.
The caps were stripped off their plastic wrap and fitted nicely and tightly in 10mm holes.
The diodes were fitted in 10,5mm holes and the 7808's face in a shallow 12,5mm hole, while its screw tightened it to the wood.
I know the layout is pretty crude, but it was more of a "proof of concept" trial.
The result was rather subtle and mostly textures were affected. I have a nicer layout in mind, that will allow tighter bonding of the diodes and the 7808 in the wood, but I do not expect it to have any further effect. It will just be nice to have in the final chassis.
Further to modding, I changed C13 with a Panasonic FR at 47uF (I just love these caps )
The result was much more controlled bass. Quite less boomy and better defined. But at the same time, grain appeared and mid highs felt quite distorted. At the same time, the mecha started reading stuff quite faster, but what was really most noteable is that it now stops/breaks really fast and smooth.
At that point I also removed C43's bypass (C42) and also removed C13 and its bypass C12.
This made most if not all the grain disappear. I guess the bypasses were working nicely with Tibi's caps, but were completely wrong for 47uF in C43.
I also wanted to replace C10, but it is in a rather tightly populated area, so I d rather do it when I am less tired. By the way, Jaroslaw's guide suggests removing it altogether. Anyone tried it?
By the way, moding Tibi's PCB is not for the faint of heart... It requires patience, proper tools and lots of flux to properly remove old parts, clean holes and insert the new ones.
To be honest my station (450C max) had a really rough time... In the end I ended up soldering C8 as an SMD to avoid further heating the pads and risking to lift them...
Pizza break now... and after that I will also try to separate the digital stuff psu from the motors.
Now to the electronics part, I have tried a few stuff.
First of all, I changed the PSU of the kit with a p2p built one using 2x1000uF Panasonic FRs before and 1 after the 7808.
I have only MUR860 diodes so I used those for the rectification. (A friend suggested I gave these a try... D9E Russian Military Germanium Diode Genuine 20 | eBay)
The increase in clarity was huge even with freshly soldered and unused parts (not broken in).
At that time, I saw a piece of scrap wenge so I decided to add some damping to the entire PSU a la Eric and Jaroslaw.
The caps were stripped off their plastic wrap and fitted nicely and tightly in 10mm holes.
The diodes were fitted in 10,5mm holes and the 7808's face in a shallow 12,5mm hole, while its screw tightened it to the wood.
I know the layout is pretty crude, but it was more of a "proof of concept" trial.
The result was rather subtle and mostly textures were affected. I have a nicer layout in mind, that will allow tighter bonding of the diodes and the 7808 in the wood, but I do not expect it to have any further effect. It will just be nice to have in the final chassis.
Further to modding, I changed C13 with a Panasonic FR at 47uF (I just love these caps
)The result was much more controlled bass. Quite less boomy and better defined. But at the same time, grain appeared and mid highs felt quite distorted. At the same time, the mecha started reading stuff quite faster, but what was really most noteable is that it now stops/breaks really fast and smooth.
At that point I also removed C43's bypass (C42) and also removed C13 and its bypass C12.
This made most if not all the grain disappear. I guess the bypasses were working nicely with Tibi's caps, but were completely wrong for 47uF in C43.
I also wanted to replace C10, but it is in a rather tightly populated area, so I d rather do it when I am less tired. By the way, Jaroslaw's guide suggests removing it altogether. Anyone tried it?
By the way, moding Tibi's PCB is not for the faint of heart... It requires patience, proper tools and lots of flux to properly remove old parts, clean holes and insert the new ones.
To be honest my station (450C max) had a really rough time... In the end I ended up soldering C8 as an SMD to avoid further heating the pads and risking to lift them...
Pizza break now... and after that I will also try to separate the digital stuff psu from the motors.
Attachments
This pretty much echos my experience from my cap mods except in a greater detail of the changes. I changed my caps all at once so it lacked the detail of the effect of each change. End result seems to be similar.Further to modding, I changed C13 with a Panasonic FR at 47uF (I just love these caps
The result was much more controlled bass. Quite less boomy and better defined. But at the same time, grain appeared and mid highs felt quite distorted. At the same time, the mecha started reading stuff quite faster, but what was really most noteable is that it now stops/breaks really fast and smooth.
At that point I also removed C43's bypass (C42) and also removed C13 and its bypass C12.
This made most if not all the grain disappear. I guess the bypasses were working nicely with Tibi's caps, but were completely wrong for 47uF in C43.
I also wanted to replace C10, but it is in a rather tightly populated area, so I d rather do it when I am less tired. By the way, Jaroslaw's guide suggests removing it altogether. Anyone tried it?
FWIW, I replaced C10 with 47uF Silmic II. I lifted the ground pad of this cap. So, there is no going back. Search the previous thread for C953. You will find some references to it.
My gut feeling is that, the C13 and the C43 (and the bypasses) mods produce the biggest impact which are the must mods documented by Peter. This seems to hold true with the Tibi board as well.
The enclosure sides are almost complete. I finished the openings this morning, did some sanding and then started putting in the reinforcement corner braces with wood epoxy. When fully assembled the sides will be screwed to the bottom moutning board. Once the current step is done, I have to cut out the top panel and then locate and cut the opening for the laser / platform.
I'm glad I'm doing the stock version first. One thing I've learned so far is that it's a lot easier to do things when you have the right tools!
I'm glad I'm doing the stock version first. One thing I've learned so far is that it's a lot easier to do things when you have the right tools!
Attachments
C952 (C13) and its bypass C963 (C12) are unnecessary (and indeed harmful) because the PSU output caps are already doing the same job. The only use they might have is when your PSU is located far from the main board.
I do NOT recommend removing C10. I only suggest getting rid of its bypass, C14. Likewise, I suggest removing C42, which is the bypass for C43.
The MSR860 are quite noteiceably better than MUR860, which are boomy and obviously coloured. Leave your MUR860 for Gainclones where it works best, and get some MSR860 for your Shiga
@pchw - If you have any more of those Black Gates, they are guaranteed to work better than the Silmic - Silmics may have their uses, but Shigaclone surely isn't one of them. If you ran out of Black Gates, try Starget or Os-Con.
Tibi, capacitor's ESR at 100kHz (or any other arbitrary frequency) has little correlation with the cap's performance in terms of sound. The noise profile, mechanical damping, interactions with other components and a multitude of other factors can affect the sound. I can bet big money that the Black Gate on your picture actually sounds a lot better than the Samxon, despite the poor measurements.
I do NOT recommend removing C10. I only suggest getting rid of its bypass, C14. Likewise, I suggest removing C42, which is the bypass for C43.
The MSR860 are quite noteiceably better than MUR860, which are boomy and obviously coloured. Leave your MUR860 for Gainclones where it works best, and get some MSR860 for your Shiga
@pchw - If you have any more of those Black Gates, they are guaranteed to work better than the Silmic - Silmics may have their uses, but Shigaclone surely isn't one of them. If you ran out of Black Gates, try Starget or Os-Con.
Tibi, capacitor's ESR at 100kHz (or any other arbitrary frequency) has little correlation with the cap's performance in terms of sound. The noise profile, mechanical damping, interactions with other components and a multitude of other factors can affect the sound. I can bet big money that the Black Gate on your picture actually sounds a lot better than the Samxon, despite the poor measurements.