You wrote "Aiyima A07"... did you mean that or did you mistyped Aiyima A04?
This thread is only about Aiyima A04, hence asking... I know some people consider some chinese amps to be close, but when you take into account that the 04 has already a non constant production with parts and schematic variations, I don't subscribe to that (not even taking into account these are built around a chip that is different) when it comes to tweaking / changing parts. So you might be more lucky in a dedicated A07 thread (or opening one?)
If you meant A04, then I would say that the units we had in the past so far where hum free and sounding reasonably well. Forget the power ratins, they mean nothing, but that it isn't the problem here. So if you have hum I would either return the unit or try another power supply. That applies also to "no clarity, very poor dynamics, feeble bass, terrible highs" as apart from the deppest bass (where the unit doesn't shine and because the damping factors of various amps plays a role) I wouldn't say a good working unit qualifies as such in untuned form, unless you like a very biased sound (easy to check your tastes vs any commercial amp).
In short: if it is a A04, test another PS and if not happy return the units (just fit the original OPAs back in to reverse your mod) as you either got defective ones or, if they are recent, the production has perhaps changed for the worst. If it is a A07, sorry, cant comment, not the right thread IMHO.
Good luck either way
Claude
This thread is only about Aiyima A04, hence asking... I know some people consider some chinese amps to be close, but when you take into account that the 04 has already a non constant production with parts and schematic variations, I don't subscribe to that (not even taking into account these are built around a chip that is different) when it comes to tweaking / changing parts. So you might be more lucky in a dedicated A07 thread (or opening one?)
If you meant A04, then I would say that the units we had in the past so far where hum free and sounding reasonably well. Forget the power ratins, they mean nothing, but that it isn't the problem here. So if you have hum I would either return the unit or try another power supply. That applies also to "no clarity, very poor dynamics, feeble bass, terrible highs" as apart from the deppest bass (where the unit doesn't shine and because the damping factors of various amps plays a role) I wouldn't say a good working unit qualifies as such in untuned form, unless you like a very biased sound (easy to check your tastes vs any commercial amp).
In short: if it is a A04, test another PS and if not happy return the units (just fit the original OPAs back in to reverse your mod) as you either got defective ones or, if they are recent, the production has perhaps changed for the worst. If it is a A07, sorry, cant comment, not the right thread IMHO.
Good luck either way
Claude
Last edited:
Oh, and op rolling or any conclusions on what OPA tio use is meaningless if your unit doesn't work the way it should: you might actually prefer a bad OPA because that particular one might "help mitigating or hiding" the defects of the unit or the PS.
Save your OPAs for working amps, and defo don't buy from Ebay regardless the seller's rating!
All IMHO
Claude
Save your OPAs for working amps, and defo don't buy from Ebay regardless the seller's rating!
All IMHO
Claude
Hey Claude,
Thanks for your reply! I appreciate it!
The amps that I got are Aiyima A07, not A04. I think that both amps are based on TPA325... chips, which are pretty much the same thing. Do you think that Aiyima 04 and 07 are so different?
Meanwhile I tried the amps with Meanwell LRS-350-45 power supply. The hum got reduced, but still there. The sound is still kind of flat, not dynamics and the most I don't like is the lack of clarity, especially at low levels.
I screwed up one of the original op amps when pulled it out. So I'm returning just one amp.
I got several op amps from the same guy on eBay (he is in the US, not China) a few months ago to build a balanced pre amp, and I'm telling you, that the sound is absolutely fantastic, brilliant, completely matching the performance of Kaltecs' ultra-low distortion kits. I think you're right that installing good op amps just revels the faults of Aiyima 07.
I'm reading that post-filter feedback can improve the sound. Do you have any knowledge how to do it on Aiyima 07?
Hi,
I am sorry, I have no clue on the A07. It might, or might not, be different from A04: I have never seen or opened one so can't tell, but if they have different names there must be a reason for it, and they are supposed to have a different "heart", albeit from the same lineage indeed.
One really needs to open and play with at least one unit to recommend tweaks on it IMHO, so I don't feel legitimate at all to comment.
Good luck to you
Claude
I am sorry, I have no clue on the A07. It might, or might not, be different from A04: I have never seen or opened one so can't tell, but if they have different names there must be a reason for it, and they are supposed to have a different "heart", albeit from the same lineage indeed.
One really needs to open and play with at least one unit to recommend tweaks on it IMHO, so I don't feel legitimate at all to comment.
Good luck to you
Claude
Hi all,
Recently I bought a TPA3251 amplifier from Aliexpress with a very good price.. Which is;
https://tr.aliexpress.com/item/1005...st_main.5.2c1d3d12y2hBQc&gatewayAdapt=glo2tur
After I received my amp, the first thing I did was to examine its interior and components of course.
I would like to say that the material and PCB quality was obviously better than AIYIMA. As far as I can see, the implementation scheme is similar to the TPA3251 Evaluation Module. So it is more or less the same as AIYIMA. The only difference is that the opamps are in SOIC 8 form. On the other hand, XL7015 is used as step down and 78M12 is used for linear supply. There are also 1117-3.3 for digital operation and a SOIC 8 IC I can't identify and a few smd resistors and capacitors around it. Since no power-on or power-off noise I think this IC works for this.
As a result of my first listening to the native version of the amp; I can say that it sounds relatively good for a class D amplifier. However, I did notice that the mids were somewhat prominent, the trebles were bright, and the basses were boomy.
Naturally, I embarked on a basic modification.
1- Replaced opamp with OPA1656.
2- Replaced all 10uF electrolytics around the opamp with 10uF Silmic 2s. In addition, point to point 10uF Silmic 2s were added between pins 8-4 of the opamps.
3- Since I was suspicious of the 78M12 and XL7015 on the card, I replaced them with original products. and I also replaced the 220uF low quality capacitor after 78M12 with 330uF Panasonic FC.
4- Finally, I added 2 x 2000uF/50v 13mohm Nippon capacitors in addition to the 2x3000uF Nichicon main bypasses.
I should especially point out that; solder quality of the board was good according to AIYIMA. Components were easily removed and soldered.
The sonic result was far better the native version.
The sound is more and more balanced, no hearable distortion. Noise is lower than the original version.
Strongly recommended.
Best regards,,
Recently I bought a TPA3251 amplifier from Aliexpress with a very good price.. Which is;
https://tr.aliexpress.com/item/1005...st_main.5.2c1d3d12y2hBQc&gatewayAdapt=glo2tur
After I received my amp, the first thing I did was to examine its interior and components of course.
I would like to say that the material and PCB quality was obviously better than AIYIMA. As far as I can see, the implementation scheme is similar to the TPA3251 Evaluation Module. So it is more or less the same as AIYIMA. The only difference is that the opamps are in SOIC 8 form. On the other hand, XL7015 is used as step down and 78M12 is used for linear supply. There are also 1117-3.3 for digital operation and a SOIC 8 IC I can't identify and a few smd resistors and capacitors around it. Since no power-on or power-off noise I think this IC works for this.
As a result of my first listening to the native version of the amp; I can say that it sounds relatively good for a class D amplifier. However, I did notice that the mids were somewhat prominent, the trebles were bright, and the basses were boomy.
Naturally, I embarked on a basic modification.
1- Replaced opamp with OPA1656.
2- Replaced all 10uF electrolytics around the opamp with 10uF Silmic 2s. In addition, point to point 10uF Silmic 2s were added between pins 8-4 of the opamps.
3- Since I was suspicious of the 78M12 and XL7015 on the card, I replaced them with original products. and I also replaced the 220uF low quality capacitor after 78M12 with 330uF Panasonic FC.
4- Finally, I added 2 x 2000uF/50v 13mohm Nippon capacitors in addition to the 2x3000uF Nichicon main bypasses.
I should especially point out that; solder quality of the board was good according to AIYIMA. Components were easily removed and soldered.
The sonic result was far better the native version.
The sound is more and more balanced, no hearable distortion. Noise is lower than the original version.
Strongly recommended.
Best regards,,
Attachments
Looks based on your pix indeed VERY similar to A04... copy?
No wonder then that similar tweaks work similarly well...
Well done and enjoy your music... and thanks for the addition, as it appears this board is easier to work with (better built?), apart perhaps from the OPAs which could be a pain for some DIYer.
Claude
No wonder then that similar tweaks work similarly well...
Well done and enjoy your music... and thanks for the addition, as it appears this board is easier to work with (better built?), apart perhaps from the OPAs which could be a pain for some DIYer.
Claude
Yes sir, as a A04 modifier (as you recommended), I can say; they're very similar. But the solder quality is far better than A04. So working is easier.
The sound is impressive for a $50 amplifier.. More like a ~$1000 mid market device. Exactly worth to work on it.
But I forgot to say that;
DO NOT USE OPA1642.
They exactly don't work with single supply, (at least on this or A04 board). I've lost 3 days and 4 opamps !
Thanks for the comment sir.
Thanks for sharing your (sadly bad) experience and letting other benefit from it.
OPA1646 suits the A04 quite well IMHO... even better with some PS bypasses (probably specific to A04)
Thanks also for the compliments... yep, as posted in the past after some comparisons, in untuned form it plays in the modern 600$ bracket IMHO, once (heavily) tuned indeed in the 1500$ bracket.
So in both cases worth the money... and defo worth the mods to achieve a quite well balanced and satisfying unit!
Enjoy it very much, perhaps even more (depending on you LS) in bi-amp form as Gilles does as for that price it really offers bi-amping at an affordable price IMHO. As already posted, simple SMPS filters are the icing on the cake, and so easy to try / loop in between SMPS and amp (hint ;-)
Claude
OPA1646 suits the A04 quite well IMHO... even better with some PS bypasses (probably specific to A04)
Thanks also for the compliments... yep, as posted in the past after some comparisons, in untuned form it plays in the modern 600$ bracket IMHO, once (heavily) tuned indeed in the 1500$ bracket.
So in both cases worth the money... and defo worth the mods to achieve a quite well balanced and satisfying unit!
Enjoy it very much, perhaps even more (depending on you LS) in bi-amp form as Gilles does as for that price it really offers bi-amping at an affordable price IMHO. As already posted, simple SMPS filters are the icing on the cake, and so easy to try / loop in between SMPS and amp (hint ;-)
Claude
Thanks Sir,
I have a Meanwell LRS-350/36 for it. And used a 3.3 mH common mode choke on the output. Guess its a good combination.
And I want to replace the output coils with my Würth flat wire coils, however I am not sure if they are suit to replace ????
Do you have any recommendation fot this operation?
Attachments
As already posted with parts details and pix I did similar and TBH in our appications the coil replacement yeld very little, while being the most expensive tweak. Wasn't flat coil but something that looked at least as promising on the paper (specs) and following application sheet, and from a reputable manufacturer, as discussed in this thread (was along time ago, don't remember all details or parts, search function is your friend).
The result wasn't nile though, but I would say in our case not really worth the effort / the priority.
BUT indeed in our application we use probably max 3-5 W per channel, so saturation might be less of a problem than when using... well, whatever this amp can deliver in real life until its board blows (or whatever else) given how tiny everything is dimensioned.
Short answer and in our experience: coils are not the priority and for low power application defo not the bottleneck, and we tried hard as they did cost quite a bit so we had rather high expectations. For higher power deliveries we can't say though as we didn't try.
Note that we also used the coild swap to adapt the amp to our LS impedance "while at it", and I suspect the main very small benefit might be just that... again in our low power application.
I hope this helps
Claude
The result wasn't nile though, but I would say in our case not really worth the effort / the priority.
BUT indeed in our application we use probably max 3-5 W per channel, so saturation might be less of a problem than when using... well, whatever this amp can deliver in real life until its board blows (or whatever else) given how tiny everything is dimensioned.
Short answer and in our experience: coils are not the priority and for low power application defo not the bottleneck, and we tried hard as they did cost quite a bit so we had rather high expectations. For higher power deliveries we can't say though as we didn't try.
Note that we also used the coild swap to adapt the amp to our LS impedance "while at it", and I suspect the main very small benefit might be just that... again in our low power application.
I hope this helps
Claude
Dedicated AOP simple testrig driving headphones (original Beyer DT990 for pros - 600Ohms easy drive)
More details in the approriate thread as all these were not tested with the A04 which benefited from the results... was ages ago, good luck with the such function
Hmm... in a last tweak try to squeeze more sound of that little unit I might though got for another (unreasonable) try in the future...
More details in the approriate thread as all these were not tested with the A04 which benefited from the results... was ages ago, good luck with the such function
Hmm... in a last tweak try to squeeze more sound of that little unit I might though got for another (unreasonable) try in the future...
Thank you Claude.
Is it the OPA1656 id or idr?
Is it possible to take your PS filter in parallel? I ask because I want to use a bigger PS for several Aiyimas.
This weekend I read the complete 40sides and I must say you are really a very kind and helpful man. Really! Chapeaux.
It was a pleasure to read.
Ralf
Is it the OPA1656 id or idr?
Is it possible to take your PS filter in parallel? I ask because I want to use a bigger PS for several Aiyimas.
This weekend I read the complete 40sides and I must say you are really a very kind and helpful man. Really! Chapeaux.
It was a pleasure to read.
Ralf
Danke Dir, Ralf!
Regarding the OPA1656, I am not sure it matters really once installed, but for whatever it is worth mine are / were "ID".
Regarding PS and filters, as the final installation of the first build was bi-amping type, in fact instead of using a bigger PS feeding two separate amps we decided to have it all separate / dedicated.. That meant 2 separate amps, 2 separate SMPS and 2 separate filters of my own.
Now let me understand what you mean by parallel. Perhaps you don't really want to "parallel just the filters" in electrical terms! That would make very little sense vs building one out of components that can stand more Amperes. And wouldn't help channel separation etc.
Isn't it rather that you want 1 big SMPS feeding X blocks...blocks consisting each of "1 FILTER + 1 Amp", these entire blocks being in fact in parallel as seen by the SMPS power feed?
Eg. if we forget the filters and take X = 2, "one SMPS feeding 2 amps". Right? In your case you would just add to this simplistic view a filter upstream of each amp module, and of course... that would be absolutely no problem! In fact I did this with other similar filters in my VFET, one single SMPS feeding all L and R channel VFET modules, but each channel having upstream its SMPS filter. So the filters were just inserted between the SMPS and each channel amp modules in that case. Again, as here, that is NO PROBLEM in that case, these are rather parallel amp branches if you think of it than parallel connections of the filters alone at their entry and exit...
I hope this helps
MFG
Claude
Regarding the OPA1656, I am not sure it matters really once installed, but for whatever it is worth mine are / were "ID".
Regarding PS and filters, as the final installation of the first build was bi-amping type, in fact instead of using a bigger PS feeding two separate amps we decided to have it all separate / dedicated.. That meant 2 separate amps, 2 separate SMPS and 2 separate filters of my own.
Now let me understand what you mean by parallel. Perhaps you don't really want to "parallel just the filters" in electrical terms! That would make very little sense vs building one out of components that can stand more Amperes. And wouldn't help channel separation etc.
Isn't it rather that you want 1 big SMPS feeding X blocks...blocks consisting each of "1 FILTER + 1 Amp", these entire blocks being in fact in parallel as seen by the SMPS power feed?
Eg. if we forget the filters and take X = 2, "one SMPS feeding 2 amps". Right? In your case you would just add to this simplistic view a filter upstream of each amp module, and of course... that would be absolutely no problem! In fact I did this with other similar filters in my VFET, one single SMPS feeding all L and R channel VFET modules, but each channel having upstream its SMPS filter. So the filters were just inserted between the SMPS and each channel amp modules in that case. Again, as here, that is NO PROBLEM in that case, these are rather parallel amp branches if you think of it than parallel connections of the filters alone at their entry and exit...
I hope this helps
MFG
Claude
Thank you Claude,
so the coil is the limiting part with its 5,5A.
It is no problem to take one filter for one amp. Costs are really not that much. Is it necessary for the function of the filter to take 3000uF or can it be more, because there aren t these caps at mouser at the moment. The amp also has the input caps so I think it can be a bit more without problem.
On your pics of the filter there is a diode. What is it for?
Thanks and have a nice day.
Ralf
so the coil is the limiting part with its 5,5A.
It is no problem to take one filter for one amp. Costs are really not that much. Is it necessary for the function of the filter to take 3000uF or can it be more, because there aren t these caps at mouser at the moment. The amp also has the input caps so I think it can be a bit more without problem.
On your pics of the filter there is a diode. What is it for?
Thanks and have a nice day.
Ralf
Hallo Ralf
There are other coils of course, then it is a matter of other specs such as resonnance frequency, DC resistance, form factor...
It is unlikely you need all that current, and if so only on short bursts, as I doubt very much the 04 board is dimensioned really for sustained high power, but well... enough options there anyway
Of course no problem packing more uF with larger caps etc.
In fact, the limiting factor might probably be the hi-cup mode of the SMPS at start up, loading all the caps. That depends really on your PS, ramp up etc.
Having said that, once the filter was in place, once the tweaks around the PS were done, as you probably already know we tried increasing the PS caps. We did it conviniently at the filter, as we had already 2 x 3000uF inside our initial 04s. And that brought absolutely NO sonic benefit in our case. That was even the case when we used only 1 amp, as of course in bi-amping mode.
These Class D amps need rather fast PS caps than lot of uF, IMHO. That is of course from a ceratin value on. It is though true that IME SMPS despite their speed benefit from some external PS caps. But all in all we felt we had reached the limit there, well acounting for the filter did a difference, be it additional PS "buffering" caps and or filtering function.
Your case could although be different, so you have to try. Indeed, our need to drive our LS is below only 2W per channel! Not the best for a Class D BTW, but well. That means we do'nt need a lot of power, on the contrary, and in Giles case it is even splitted between 2 amps! LS requiring more power from the amp may or may not benefit from larger or additional caps. As long as these are fast, eg low ESR at high frequency etc., who knows...
It was though interesting to note that despite the low power Gilles needed, the sound benefitted from bi amping mode, and that despite specs distorsions that are for sure no more that great on the paper at these low power levels if I recall the spec sheet of the chip (could be wrong, was ages ago). But perhaps it is just his LS that like bi-amping mode, who knows. We tried the very same LS with my VFET amp, worked perfectly, so much that I use now same VFET and a set of same LS in my living room, limited to 2W perchannel BTW...
Have fun
Claude
There are other coils of course, then it is a matter of other specs such as resonnance frequency, DC resistance, form factor...
It is unlikely you need all that current, and if so only on short bursts, as I doubt very much the 04 board is dimensioned really for sustained high power, but well... enough options there anyway
Of course no problem packing more uF with larger caps etc.
In fact, the limiting factor might probably be the hi-cup mode of the SMPS at start up, loading all the caps. That depends really on your PS, ramp up etc.
Having said that, once the filter was in place, once the tweaks around the PS were done, as you probably already know we tried increasing the PS caps. We did it conviniently at the filter, as we had already 2 x 3000uF inside our initial 04s. And that brought absolutely NO sonic benefit in our case. That was even the case when we used only 1 amp, as of course in bi-amping mode.
These Class D amps need rather fast PS caps than lot of uF, IMHO. That is of course from a ceratin value on. It is though true that IME SMPS despite their speed benefit from some external PS caps. But all in all we felt we had reached the limit there, well acounting for the filter did a difference, be it additional PS "buffering" caps and or filtering function.
Your case could although be different, so you have to try. Indeed, our need to drive our LS is below only 2W per channel! Not the best for a Class D BTW, but well. That means we do'nt need a lot of power, on the contrary, and in Giles case it is even splitted between 2 amps! LS requiring more power from the amp may or may not benefit from larger or additional caps. As long as these are fast, eg low ESR at high frequency etc., who knows...
It was though interesting to note that despite the low power Gilles needed, the sound benefitted from bi amping mode, and that despite specs distorsions that are for sure no more that great on the paper at these low power levels if I recall the spec sheet of the chip (could be wrong, was ages ago). But perhaps it is just his LS that like bi-amping mode, who knows. We tried the very same LS with my VFET amp, worked perfectly, so much that I use now same VFET and a set of same LS in my living room, limited to 2W perchannel BTW...
Have fun
Claude
Hello Claude,
I ordered everything for many amps because I have a fully active 4-way horn system in 5.4 with DBA (22 speakers front, 8 rear) for the bass and infra sub (9 speakers) in the couch in a copmpletely dampened room.
If you say by sound the A04 plays in the abaout 1500-2000€ class it should be great. Normally I don't need much power at 110-118dB/1W speakers but it really sounds different if you give power on them because there is no distortion.
For testing a filter I ordered 2 chokes with 3,3uH to take them in parallel. If it does not work and sounds the same as with 1,5uH I take your version. As PS I try several of the LRS 200 (one bigger is much more expensive and with fan) if the filter with 2 chokes in parallel will not work and LRS 350 if it works for a bit "more headroom".
Again thanks for your advice and help.
Ralf
I ordered everything for many amps because I have a fully active 4-way horn system in 5.4 with DBA (22 speakers front, 8 rear) for the bass and infra sub (9 speakers) in the couch in a copmpletely dampened room.
If you say by sound the A04 plays in the abaout 1500-2000€ class it should be great. Normally I don't need much power at 110-118dB/1W speakers but it really sounds different if you give power on them because there is no distortion.
For testing a filter I ordered 2 chokes with 3,3uH to take them in parallel. If it does not work and sounds the same as with 1,5uH I take your version. As PS I try several of the LRS 200 (one bigger is much more expensive and with fan) if the filter with 2 chokes in parallel will not work and LRS 350 if it works for a bit "more headroom".
Again thanks for your advice and help.
Ralf