Sure, it keeps working nicely. It wasn't any original design, but based upon manufacturer datasheet recommendations.
Are you referring to RK168 series? As per datasheet, worst case condition is supplying 150mA at 6v, nearly 1 Watt. The regulator I used will work up to 1,5A and up to several watts if properly heatsinked. Keep in mind that it is a puntual consumption, thus changing volume is not a constant task (no quiescent current).
In fact, I didn't finally mounted it inside the amp enclosure, but with a DCB1 in a separated preamp case, taking supply down to 5v from DCB1 PSU. How much current can you borrow? It will depend on power TX size, and how much ripple current will you allow to have on the LM3886. Asuming you're using BOM capacitors. But low current applicances should pose no problem.
Aren't you guys talking about the same? (powering LDR volume controls)
I will read the whole thread, and closely follow the development, it seems an interesting approach. But by the moment I am finishing an stereo BPA-300 (two channels, 12x LM3886), and still rest some work to do. After Building two standard gainclones, another one gainclone with regulated PSU, the My-Ref with U-BOM, a Nelson Pass' F5, and the current BPA-300, I think it's time to give a breath to the omnipresent LM3886 so.... maybe it's a good time to start with valve kits!
Putting 12K in parallel with R7 proved to be very interesting. I am going to measure and see what I find.
R7 cut to half
An externally hosted image should be here but it was not working when we last tested it.
Was there a specific reason why gm=1 for the original MyRef C design?
Last edited:
regiregi22,
Yea, same thing - I was just curious to know if anyone had successfully taken some power (5V) from the MyRef transformer to eliminate the wall wart/toroid for the Lightspeed in an integrated build. I wouldn't even consider it for the 12V for the Lighter Note. But apparently you took a different path.
Yea, same thing - I was just curious to know if anyone had successfully taken some power (5V) from the MyRef transformer to eliminate the wall wart/toroid for the Lightspeed in an integrated build. I wouldn't even consider it for the 12V for the Lighter Note. But apparently you took a different path.
I took a different path, but same procedure applies (DCB1 instead of My_Ref). I went even further than taking supply from the torroid. I took supply directly from the PSU smoothing caps, so you don't need to add another diode bridge neither smoothing caps before the regulator. The regulator just take down those +35vdc (but depends on your trafo voltage) to +5vdc.
Try it first without heatsinking the regulator, you are going to drop 25-30 volts, but due to the low current it has to supply, heat may not be an issue.
Regards,
I know this doesn't really have much to do with the My_Ref, theory wise but for those building a integrated amp and not have the chassis filled with tiny transformers; what regi did really interested me. I took some notes from this thread. the number after a users name is the post number of the Ultimate BOM thread so I could go back and read what I failed to copy to notepad.
Attachments
Last edited:
Based on current experiment and simulation, I instinct is that R7=4K R5&R8=47K might be an optimum solution. The only thing left to try myself is to get the 47K resistors with the tolerance required, will take a couple of days.
Hi Soongsc,
Very interesting, it could be a easy upgrade to the amp. Please let us know how it sounds when you implement the change.
PJN
Very interesting, it could be a easy upgrade to the amp. Please let us know how it sounds when you implement the change.
PJN
That combination provides for the best frequency response and damping so far. Sonic coloration is much reduced. The gain is also reduced due to R7. Improvements also include better separation of instruments, and depth location difference. Background recorded room timbre is more evident. It's getting very close to my reference.
Some additional improvement might be possible with C4 change, I just need to confirm whether what I hear is related with this or not. But also bear in mind that I have a different power supply, so people interested still need to know that there is some uncertainty.
Some additional improvement might be possible with C4 change, I just need to confirm whether what I hear is related with this or not. But also bear in mind that I have a different power supply, so people interested still need to know that there is some uncertainty.
C4 changed to a 0.68uF X2 cap I had on hand, but the original TP C13 could also be tried here. The low frequency timbre is more revealed, which will more correctly portay the recording room size. This also helps enhance the presence of the performance in a room. Although I still have Dario's alternate compensation in one channel, the difference between the original one is not very audible. I slightly prefer the original compensation.
I dont intend to make this a Lightspeed thread but wanted to mention that I dont think I ever recommended a LM338, but I did recommend a LM334 current source.
Interesting but you should be aware that you're using a different capacitor, not only a different value... so it could simply be that the X2 cap is better... 😉
In fact if the TP supplied cap is the usual Wima MKS4 a lot of caps are better... MKS4 tends to be too much 'romantic' with recessed voices.
You should repeat the test with the same cap in different values...
In the next days I'll do the test too.
In my setup the difference is pretty obvious... BTW in the last two days I'm using again the original compensation and in this moment I like it best.
I'll do a more critical listening test in the next days.
The process I went through was as follows:
1. Changed R7 to 6K, removed C13 bypass cap, noticed the gain was reduced. After balancing the channels, it seemed the stage depth was increased, sound was cleaner with more detail, but also seemed slightly suppressed. (note: before, the Dario compensation seemed cleaner with slightly less bass) There was lack of emotion in the performance in both channels. Measurements showed better phase and frequency performance, and also damping improvement. So it was encouraging.
2. Changed R7 to 4K, changed C4 to 0.68uF X2. Simulation showed that there should be better damping. Listening impression was sound becomes more live, and more detailed.
3. Changed R7 in the channel with Dario's alternate compensation, removed bypass on C13, now two channels had stage depth that were similar. But the channel with alternate compensation sounded a bit dead.
4. Changed C4 to 0.68uF X2, now both channels sound almost the same. A very slight difference in timbre.
I think I am going to reuse C13 bypass caps and see what changes. I still feel that there is some suppression, but can't make out what is causing it.
When I get my hands on 0.1% 47K resistors, those are going to R5 & R8.
1. Changed R7 to 6K, removed C13 bypass cap, noticed the gain was reduced. After balancing the channels, it seemed the stage depth was increased, sound was cleaner with more detail, but also seemed slightly suppressed. (note: before, the Dario compensation seemed cleaner with slightly less bass) There was lack of emotion in the performance in both channels. Measurements showed better phase and frequency performance, and also damping improvement. So it was encouraging.
2. Changed R7 to 4K, changed C4 to 0.68uF X2. Simulation showed that there should be better damping. Listening impression was sound becomes more live, and more detailed.
3. Changed R7 in the channel with Dario's alternate compensation, removed bypass on C13, now two channels had stage depth that were similar. But the channel with alternate compensation sounded a bit dead.
4. Changed C4 to 0.68uF X2, now both channels sound almost the same. A very slight difference in timbre.
I think I am going to reuse C13 bypass caps and see what changes. I still feel that there is some suppression, but can't make out what is causing it.
When I get my hands on 0.1% 47K resistors, those are going to R5 & R8.
Last edited:
Applying C13 bypass caps was interesting. The original MyRef C had improvements removing the slight suppression I was hearing; the channel with the alternate compensation improved as well, but added edgyness to the sound. I am not sure what to make of this result yet.
Probably I will stick the caps from the TP kit back in and see what happens.
Probably I will stick the caps from the TP kit back in and see what happens.
Last edited:
Continued listening, Dario's alternate compensation seems better at high volumes. The closer I get to maximum volume, the better it seems to perform. The background seems darker, the performers are more pronounced in detail. In contrast, the original MyRef seems to become gradually compressed dynamically and stage depth. There is a volume range where both nearly sound the same.
Last edited:
i have found that comparison by keeping one channel as a control is not the best way to determine the sonic impact of a mod. not only does this give no indication of soundstaging differences but you'd also have to imagine the actual difference to be double of what you perceive to arrive at a safe conclusion. i.e. the treble of the mod is only .5db higher for one channel- not a big deal. but when fully implemented it'll be 1db- that's pretty hot! ;D
not to mention the memory fade between a/b among other variables that need to be dealt with
+1 - Absolutely agree. You have to do identical mods on both channels to assess it accurately. Note that Rev C and Rev E (alt FE) are actually sonically very similar, and the differences are small but significant - which makes it difficult to evaluate if you audition it with heterogenous channels.
The only reason to do a mod on a single channel alone is to verify that it still works - i.e. that the magic smoke doesn't escape.
I have found alternatingly modifying channels to be effective when one has learned how to do it. I did not invent this, but rather learned it from another designer. After significant long learning process, it works for me. I'm sure it's not for everone.😉
Last edited: