Sorry... other than what I guessed regarding output filtering, difficult to say for sure based on this pix alone.
Hmm... that pix must come from somewhere, right? Perhaps you downloaded it from its original website? If so you could possibly click right on your original pix and look into "Information". That should give you the source and web address of the pix, so my guess is a link to whatever F forum you took it from?
Just thinking out loud...
Claude
Hmm... that pix must come from somewhere, right? Perhaps you downloaded it from its original website? If so you could possibly click right on your original pix and look into "Information". That should give you the source and web address of the pix, so my guess is a link to whatever F forum you took it from?
Just thinking out loud...
Claude
With pleasure and yes post the link so we can follow your findings
It is true that apart from Rob's posts to come, this Aiyima topic is closed for me. 'Been there, done it" and don't even own such a unit...
Class D is not exactly my cup of tea, but I discovered that it could be a very good solution for a second system, enabling to save money for the loudspeakers, so always worth following this development.
Stay safe
Claude
It is true that apart from Rob's posts to come, this Aiyima topic is closed for me. 'Been there, done it" and don't even own such a unit...
Class D is not exactly my cup of tea, but I discovered that it could be a very good solution for a second system, enabling to save money for the loudspeakers, so always worth following this development.
Stay safe
Claude
daniboun: that’s the older version of the 3e board, correct? It had two different main power supply connections, to either side of the main output caps. The person just didn’t connect one PS per connection? In my electrical naïveté I would have thought that to cause shorting issues. The extra supply caps have to be connected (somewhere) into that arrangement as well..wish they should have used different color wiring ;>)
Whats your "go to" DIY amp? Im assuming something class A/B, which I also really like.
Very personal question, LOL! But here we go, all IMHO... but after all, we are speaking about me 
After some long thinking, I settled my choice on FSSA-2, in a derivative to be hopefully published soon and in a high bias version in order to have quite a few Watts in Class A under 4R while beneftting from a good damping factor and enough reserve (Watts / Amperes) to drive my or any LS.
It bears of course no comparison whatsoever with the project in this thread, being easily 8 times the price (of a tuned unit) and hopefully as much better re sonic quality. Not even to mention DIY complexity and design involvement on the builder's side, starting from completely scratch or nearly so.
To remain on more affordable DIY projects, I must confess I was quite impressed by the sonic results, footprint, easy for beginners, VFM of this little Class D project. In bi-amping form it starts to be a really good amp that is likely to please the vast majority of quite demanding modern music lovers.
In the same spirit, if going Class A/B, I would follow with interest the last Xmas chip amp... seems an interesting project with potential, as was this project last year.
Much more budget and personal involvement, and going Class A on OK to drive LS, while relying on a debugged / documented path (so not too difficult and a community to help), I would consider any USSA variant... or if really high efficiency speakers perhaps Papa's intriguing next power amp around those old big cup transistors
Just my 2p's, and subject to modifications as some amps are known quantities... whereas some are not even released yet and just (very) promising!
Not to mention some are not even known yet by me and yet to come, whatever the Class
Claude
After some long thinking, I settled my choice on FSSA-2, in a derivative to be hopefully published soon and in a high bias version in order to have quite a few Watts in Class A under 4R while beneftting from a good damping factor and enough reserve (Watts / Amperes) to drive my or any LS.
It bears of course no comparison whatsoever with the project in this thread, being easily 8 times the price (of a tuned unit) and hopefully as much better re sonic quality. Not even to mention DIY complexity and design involvement on the builder's side, starting from completely scratch or nearly so.
To remain on more affordable DIY projects, I must confess I was quite impressed by the sonic results, footprint, easy for beginners, VFM of this little Class D project. In bi-amping form it starts to be a really good amp that is likely to please the vast majority of quite demanding modern music lovers.
In the same spirit, if going Class A/B, I would follow with interest the last Xmas chip amp... seems an interesting project with potential, as was this project last year.
Much more budget and personal involvement, and going Class A on OK to drive LS, while relying on a debugged / documented path (so not too difficult and a community to help), I would consider any USSA variant... or if really high efficiency speakers perhaps Papa's intriguing next power amp around those old big cup transistors
Just my 2p's, and subject to modifications as some amps are known quantities... whereas some are not even released yet and just (very) promising!
Not to mention some are not even known yet by me and yet to come, whatever the Class
Claude
Last edited:
Wow, yes that's quite a different animal entirely. I'm definitely interested in the x-mas amp but it looks like the pcb isn't good sale anywhere? I admittedly didn't talk the entire that but it looks like you have to have the pcb made somewhere. I've never done anything like that.
It would be nice if a kit was sold.
It would be nice if a kit was sold.
Just follow the X-mas thread, there was (is?) a group buy for the PCBs, but they had to open another thread in an other section for that.
Very easy to built - it is a simple chip amp, a bit like most Class D amps- and for sure some tweaking will be possible (PS etc.). I guess quite a few will go for it.
I could also see the PCBs ending one day in XRKaudio's shop. And of course you could still have them made for you directly by whatever Chinese non expensive quality shop probably.
That's another topic, see you there
Claude
Very easy to built - it is a simple chip amp, a bit like most Class D amps- and for sure some tweaking will be possible (PS etc.). I guess quite a few will go for it.
I could also see the PCBs ending one day in XRKaudio's shop. And of course you could still have them made for you directly by whatever Chinese non expensive quality shop probably.
That's another topic, see you there
Claude
Ok, I just got and changed out the 2200uf Nichicons with the 4500uf United Chemi-Con ones. Definitely a better option even without breaking these caps in yet. These are very low ESR (22 ohm) on top of double the capacity.
Claude, youre probably right about it not being as good a comparison to both increase capacity and lower ESR but one or both of these things definitely improved the sound vs both the 3000uf (fake?) Elna's and the 2200uf Nichicons.
Sound is again more open, but not as "thin" as it was with the 2200uf ones. Highs are smoother, bass is tighter, and the entire sound is more open and dynamic.
I wanted to make sure I was happy with the power cap change before starting the bypass caps but now that I am Im going to do all those as well. Anyone looking to replace the main caps should definitely consider these. These are the largest capacity, lowest ESR caps I could find that actually fit in this small case.
Digikey link
Mouser link
Claude, youre probably right about it not being as good a comparison to both increase capacity and lower ESR but one or both of these things definitely improved the sound vs both the 3000uf (fake?) Elna's and the 2200uf Nichicons.
Sound is again more open, but not as "thin" as it was with the 2200uf ones. Highs are smoother, bass is tighter, and the entire sound is more open and dynamic.
I wanted to make sure I was happy with the power cap change before starting the bypass caps but now that I am Im going to do all those as well. Anyone looking to replace the main caps should definitely consider these. These are the largest capacity, lowest ESR caps I could find that actually fit in this small case.
Digikey link
Mouser link
Last edited:
Glad you are delighted
Pity for science though you didn't first bypass the existing caps as I did, and then replaced the main caps to see if there is still a real sonic improvement. Indeed, it is easier and less costly to bypass rather than to replace.
Do you believe you could bypass your existing cap as we did (not sure it will yeld much further benefits as in our case now you improved already) AND then swap back to the original PS caps, still with the bypasses? That interesting experience would enable to evaluate how much benefit there really is replacing the OEM caps provided you already did what was recommended.
What do you think?
I still have our super PS caps lying around but didn't bother replacing them given when we fitted an external SMPS filter plus extra 3000uF low ESR caps (admittely remote) on top of the existing ones the improvement (filter +extra capacity) wasn't huge by no mean.
Wouls be interesting to follow your developments, even more so having the same starting point (future tweaker can then evaluate as far they want to go)...
Claude
Pity for science though you didn't first bypass the existing caps as I did, and then replaced the main caps to see if there is still a real sonic improvement. Indeed, it is easier and less costly to bypass rather than to replace.
Do you believe you could bypass your existing cap as we did (not sure it will yeld much further benefits as in our case now you improved already) AND then swap back to the original PS caps, still with the bypasses? That interesting experience would enable to evaluate how much benefit there really is replacing the OEM caps provided you already did what was recommended.
What do you think?
I still have our super PS caps lying around but didn't bother replacing them given when we fitted an external SMPS filter plus extra 3000uF low ESR caps (admittely remote) on top of the existing ones the improvement (filter +extra capacity) wasn't huge by no mean.
Wouls be interesting to follow your developments, even more so having the same starting point (future tweaker can then evaluate as far they want to go)...
Claude
I probably could, but to be honest I dont really like the way those original caps sounded. Theyre much more colored than either of the new ones I tried. Bass is more bloated and artificial and the mids&highs are less resolving. Depending on how motivated I am I may try that though just to give others a full comparison.
My logical mind tells me that even if bypassing improves the overall sound (Im sure it does) that if one cap starts out sounding better than another Im not sure how it would end up sounding worse after bypassing.
I would submit changing opamps and to these capacitors without doing anything else gives you at least a +6-7 vs the stock version. These capacitors alone are worth a +3 (at least) by themselves. Depending on what opa's and how you feel about them will determine the actual total. Im guessing the Elna caps that came with this arent real, as Im sure most of the components in here arent. I would be curious to see someone replace all the components in here with verified genuine ones to see what that does but obviously thats not cost effective.
My logical mind tells me that even if bypassing improves the overall sound (Im sure it does) that if one cap starts out sounding better than another Im not sure how it would end up sounding worse after bypassing.
I would submit changing opamps and to these capacitors without doing anything else gives you at least a +6-7 vs the stock version. These capacitors alone are worth a +3 (at least) by themselves. Depending on what opa's and how you feel about them will determine the actual total. Im guessing the Elna caps that came with this arent real, as Im sure most of the components in here arent. I would be curious to see someone replace all the components in here with verified genuine ones to see what that does but obviously thats not cost effective.
Last edited:
To be complete we used these quality PS caps
https://www.mouser.fr/ProductDetail/647-UBY1V302MHL/
While they are only 3000uF, they are 4.26A ripple and 26mR ESR.
Whereas conventional amps seem to benefit from a PS caps capacity increase (that's what I do for Class A and A/B amps), ripple and ESR seem to be the key factors for Class D PS caps (provided you have a minimum of capacity, in our case application sheet says 2x2200uF, even more true if you don't drive them to the limit).
Your caps look good indeed with ESR = 22mR, but not so good re ripple with only 2.8A. Whereas that is relevant or less so is difficult to tell. How they figure vs what Aiyima fitted is impossible to tell and I am glad they are an improvement vs these.
They are rated quite highish 30% tolerance (ours 20%), probably not critical, but also 105° (ours 135°). The temperature means that whereas the advertised lifetime seemed better in your case (5000h... but at 105°), in fact ours have a longer lifetime (3000h@135° translates into 24000h@105°). That is nearly 5 times more lifetime in ours, something I like especialy as that unit is for a friend. On the other hand, both are likely to live much longer anyway due to temp being probably under 55°, so who knows what will be the bottleneck re life expectency of this amp.
Sad you are so far away, you could have compared both options to you ears, I still haven't bothered fitting the remaining ones that are just lying around...
My bet is your cap can though probabaly benefit aswell from our bypasses and the values we went for (2x2.2uF X7R + nearly 0.1uF C0G) could probably be still adequate in your case
Have fun
Claude
https://www.mouser.fr/ProductDetail/647-UBY1V302MHL/
While they are only 3000uF, they are 4.26A ripple and 26mR ESR.
Whereas conventional amps seem to benefit from a PS caps capacity increase (that's what I do for Class A and A/B amps), ripple and ESR seem to be the key factors for Class D PS caps (provided you have a minimum of capacity, in our case application sheet says 2x2200uF, even more true if you don't drive them to the limit).
Your caps look good indeed with ESR = 22mR, but not so good re ripple with only 2.8A. Whereas that is relevant or less so is difficult to tell. How they figure vs what Aiyima fitted is impossible to tell and I am glad they are an improvement vs these.
They are rated quite highish 30% tolerance (ours 20%), probably not critical, but also 105° (ours 135°). The temperature means that whereas the advertised lifetime seemed better in your case (5000h... but at 105°), in fact ours have a longer lifetime (3000h@135° translates into 24000h@105°). That is nearly 5 times more lifetime in ours, something I like especialy as that unit is for a friend. On the other hand, both are likely to live much longer anyway due to temp being probably under 55°, so who knows what will be the bottleneck re life expectency of this amp.
Sad you are so far away, you could have compared both options to you ears, I still haven't bothered fitting the remaining ones that are just lying around...
My bet is your cap can though probabaly benefit aswell from our bypasses and the values we went for (2x2.2uF X7R + nearly 0.1uF C0G) could probably be still adequate in your case
Have fun
Claude
Quote: My logical mind tells me that even if bypassing improves the overall sound (Im sure it does) that if one cap starts out sounding better than another Im not sure how it would end up sounding worse after bypassing.
I am absolutely with you on that
It was just for future builders to know if once properly bypassed in fact the starting point / PS caps might not be that relevant in this amp. That would mean once bypassed perhaps it is not really worth the hassle to replace the PS caps (means less money, less work, less risk for most here).
It helps living with indeed all the fakes or non adequate components in this amp - keeping them for what they are and just replacing the absolutely critical ones and helping (bypass & Co) the critical one to make the most of it. For the PS caps we wanted initialy to increase + replace + bypass... we left it at increase + bypass in Gilles amps and at bypass only in the 3rd amp I did.
For you it is indeed just a matter of finding out / for science, no real added value for you. I was just curious about the outcome, indeed no real gain to expect in your case, just a matter or confirming or not whereas the end result is close. If you find the time, always nice to share experiences...
Thanks for all that feedback
Claude
I am absolutely with you on that
It was just for future builders to know if once properly bypassed in fact the starting point / PS caps might not be that relevant in this amp. That would mean once bypassed perhaps it is not really worth the hassle to replace the PS caps (means less money, less work, less risk for most here).
It helps living with indeed all the fakes or non adequate components in this amp - keeping them for what they are and just replacing the absolutely critical ones and helping (bypass & Co) the critical one to make the most of it. For the PS caps we wanted initialy to increase + replace + bypass... we left it at increase + bypass in Gilles amps and at bypass only in the 3rd amp I did.
For you it is indeed just a matter of finding out / for science, no real added value for you. I was just curious about the outcome, indeed no real gain to expect in your case, just a matter or confirming or not whereas the end result is close. If you find the time, always nice to share experiences...
Thanks for all that feedback
Claude
Last edited:
To be complete we used these quality PS caps
UBY1V302MHL Nichicon | Mouser France
Your caps look good indeed with ESR = 22mR, but not so good re ripple with only 2.8A. Whereas that is relevant or less so is difficult to tell. How they figure vs what Aiyima fitted is impossible to tell and I am glad they are an improvement vs these.
My bet is your cap can though probabaly benefit aswell from our bypasses and the values we went for (2x2.2uF X7R + nearly 0.1uF C0G) could probably be still adequate in your case
Have fun
Claude
Ripple voltage (as I know you know) only determines how much ripple voltage they can tolerate for a set period of time. ESR is (for sound quality purposes) the more important spec by far. Im not intending to keep this amp for 20 years and push it to its maximum. If you look at the Evaluation board TI produced that has Panasonic caps with much lower ripple voltage than the ones Im using. I highly doubt thats a concern.
I just completed the main cap bypass with 2 x 2.2uf + 1 x .1uf (low ESL) caps....the difference definitely isnt the same +2.8 you saw. I would say more like +1-1.5....slightly more clarity and tighter bass, but not that noticeable.
9 caps around the opa's bypassed. This actually seemed to make a bigger difference than the previous. I would say a genuine +2 like you saw. Each subsequent change is making for clearer, cleaner, and more dynamic sound.
Im actually using different 0.1uf caps too. Kemet low ESL ones as they seemed better than the Muratas.
All bypasses done now except the bigger ones on the opa's. I dont really have a way to add on the top of the board but I did use a single .1uf on the bottom.
I think Ill stop here for now. Working with those .1uf caps takes surgical steadiness which I dont really have.....lol. And bypassing the opa's is such a tight space.
Im actually using different 0.1uf caps too. Kemet low ESL ones as they seemed better than the Muratas.
All bypasses done now except the bigger ones on the opa's. I dont really have a way to add on the top of the board but I did use a single .1uf on the bottom.
I think Ill stop here for now. Working with those .1uf caps takes surgical steadiness which I dont really have.....lol. And bypassing the opa's is such a tight space.
Last edited:
Great work
Yep the 9 caps bypass made little difference... that is, one of those, the one for the PS, made alone 50% of the end result LOL
So sounds good, similar findings but on the PS cap bypass... seems to indicate that your caps are a better starting point (as expected), but that bypasses can still enhence them somewhat. Seems also to indicate (based on your points) that at the end keeping the existing caps and bypassing them would give very similar results for less money, give or take 1 to 2 points perhaps. You probably have the max combo now
That ripple thing could be important for Class D but indeed no brainer in your case especialy given they are 2 of these caps anyway. As of capacity and ESR, our case is quite different at the end as we added 2x3000uF in parallel to the existing ones (roughly divinding the ESR by 2 despite some connections) and anyway that ESR is highly reduced depending on the frequency band by all the bypasses - also your case now. And having 2 amps in our case make all that double, sure overkill for the couple of Watts the Klipsch requires LOL! But of course when giving points we evaluated mods in 1 amp configuration.
It is very nice to see there is consistency and repetability in the tweak's results.
Isn't it great to see how far that little amp can be tweaked... quite easily?
Again congrats for all your efforts, obviously not just reading the thread but DOING and even better, SHARING your findings. Thanks for all that, very appreciated, especialy given the time it took me to type all that some called "novels"
Claude
PS : link to the Kemet you use? I couldn't find 0.1uF small size C0G, had to make do with 0.082uF as none was available when ordering and seems still on BO. Found C0G caps to be very good, when space and cost are a concern defo an option if I can't go PPP
Yep the 9 caps bypass made little difference... that is, one of those, the one for the PS, made alone 50% of the end result LOL
So sounds good, similar findings but on the PS cap bypass... seems to indicate that your caps are a better starting point (as expected), but that bypasses can still enhence them somewhat. Seems also to indicate (based on your points) that at the end keeping the existing caps and bypassing them would give very similar results for less money, give or take 1 to 2 points perhaps. You probably have the max combo now
That ripple thing could be important for Class D but indeed no brainer in your case especialy given they are 2 of these caps anyway. As of capacity and ESR, our case is quite different at the end as we added 2x3000uF in parallel to the existing ones (roughly divinding the ESR by 2 despite some connections) and anyway that ESR is highly reduced depending on the frequency band by all the bypasses - also your case now. And having 2 amps in our case make all that double, sure overkill for the couple of Watts the Klipsch requires LOL! But of course when giving points we evaluated mods in 1 amp configuration.
It is very nice to see there is consistency and repetability in the tweak's results.
Isn't it great to see how far that little amp can be tweaked... quite easily?
Again congrats for all your efforts, obviously not just reading the thread but DOING and even better, SHARING your findings. Thanks for all that, very appreciated, especialy given the time it took me to type all that some called "novels"
Claude
PS : link to the Kemet you use? I couldn't find 0.1uF small size C0G, had to make do with 0.082uF as none was available when ordering and seems still on BO. Found C0G caps to be very good, when space and cost are a concern defo an option if I can't go PPP
Well, in a search for the link for you I discovered what I bought arent actually 0.1uf, but 0.01uf.....Im an idiot. Certainly explains why these are so small. Somehow I thought 10,000pf = .1 which of course I now see it doesnt. This explains why they made such a minor difference. Guess Ill have to order some and redo all those connections. I need a couple other parts anyway.
I guess this is part of learning something new.....lol
The low ESL KEMET's are over $4/ea which Im not sure is worth it vs $0.82/ea for TDK's. Ill just get the TDK's as Im sure theyre more than fine.
Last edited: