IMO I'd pass on it, here are my thoughts:
1) Cost is almost $100 dollars.
2) The NE5532 Op Amps are soldered directly to the board making Op Amp rolling more difficult.
3) It's not housed in a mental chassis, this makes it more prone to EMI/RFI.
4) No controls are included.
5) We assume that all the parts are real and not counterfeit, this of course is a bad assumption.
***************************
Using ~$100 dollars as our budget, a standard Aiyima TPA3251 can be bought for about ~$50 dollars, the remaining ~$50 dollars can be used a quality PSU & OP Amps.
Rob43
1) Cost is almost $100 dollars.
2) The NE5532 Op Amps are soldered directly to the board making Op Amp rolling more difficult.
3) It's not housed in a mental chassis, this makes it more prone to EMI/RFI.
4) No controls are included.
5) We assume that all the parts are real and not counterfeit, this of course is a bad assumption.
***************************
Using ~$100 dollars as our budget, a standard Aiyima TPA3251 can be bought for about ~$50 dollars, the remaining ~$50 dollars can be used a quality PSU & OP Amps.
Rob43
Not sure if it's d2 version. But should be a safe buy for the 3251 chip
ICEPOWER 100AS2 AMPLIFIER - AMPLIFICATORE - NEW VERSION! | eBay
ICEPOWER 100AS2 AMPLIFIER - AMPLIFICATORE - NEW VERSION! | eBay
Last edited:
I purchased a 3e audio amplifier board and power supply and am currently using it. I think 3e audio's products are very good. And now, like you, we are waiting for the release of the TPA3255 All in One Amplifier.
Hi amigos
Take a look at Tpa3255 thread and previous pages. We already discussed about 3E Audio finished amplifier... seems not so good
Really?
Unless I've missed it, I can't recall anyone saying that they'd bought one, received it and listened to it?
Hi Danny.
I think the only discussion to date was whether having a DSP on board was going to have any adverse affects on the sound quality. The purists probably prefer a direct connection to the TPA325x with minimal components between source and output.
There was an interesting discussion about whether rolling Op-Amps was a waste of time with this model as the DSP can be programmed to flatten out any frequency response !
In fact their are now multiple discussions that state that Op-Amp Rolling and input signal capacitor Rolling is a waste of time and only the output filter is worth spending any time upgrading. This comes from the leading contributors to this and other forums.
Improving distortion measurements seems to be the only thing that matters - hence upgrading the output Inductors as these appear to be the major source of distortion.
No one has yet received one to test so we don’t really know yet.
Last edited:
Hi James,
Yes I read this. I received my TPA32XX 360 Customs based amplifier and also got my Ncore double Mono amplifier. Now I have some really high ends amplifiers on my hands ) I am still testing it and I will share a last test after a few hours listening experience
It has not been tested yet but I got a few month ago some pictures from 3E Audio before they release the final version of the amplifier.
I shared the pictures and just got a few coments.
But you are right, maybe wait for a real review now =)
Hi guys,
I am likely to take the plunge with my summertime puzzle, tweaking an Aiyima TPA3251.
Reasons are I am intringued by the feedbacks of Rich and Danny and a very good friend of mine is thinking about a low cost temporarly amp, so even if that one doesn't deliver no drama, and if it does, who knows, it could even become permanent.
That VG friend has similar tastes as mine, likes my system a lot and used to be classical trained piano player from a family of reknown musicians. We have close tastes and similar ears when it comes to music, but for that system he wants it a bit more rock and roll and crashing than classic listening, hence me advising to purchase a pair of Klipsch RF-8000F.
These are quite good speakers, efficient and easy to drive, came delivered new at less than 800E and are the first of the non classical Klipsch I heard that are not too coloured to my ears (say not only fine for home theater or not too horn sounding). Of course they rock if needed but can play classic and jazz with good resolution, although they seem to be no match re finesse and 3D soundstage for panels / magnetostats I cherish so much but they seem (so says my friend who knows my system) to deliver quite a good resolution if no deep soundstage. I have yet to listen to them properly run in with a known amp to make my mind, but more importantly HE is very happy and these are keepers for him.. So that's the baseline to play with and they should be revealing enough to see where that amp ranks and how it reacts to tweaks. Oh, Rich of course was inspirational as running Klipsches aswell, although not in the same league, so if that worked OK for him...
I have a few ideas re tweaking in mind, one being to feed the amp through a low ripple SMPS that could accept 1 to 2 of these amps ( we are considering possibly a biamping option, vertical or horizontal, TBD, so we may perhaps buy 2 amps which would be handy also for direct tuning evaluation/monitoring).
I don't know what is inside the amp, but if following TI's papers my guess is that a Mean Well LRS-350-24 could well do the trick. It can be tuned up to 28.8V (I don't mind limiting from 120 to 120W under 8R as unlikely the chip will deliver and if not loud enough /distording than either way another amp would be the fix), it is lower ripple than the usual SMPS (150mVpp) and should with 14.6A provide enough current to drive even 2 of these amps at lower loads should it be required (given the board I don't trust anyway really 6A into these, but well). Oh, and the built in fan is very unlikely to kick in in our case.
We could also go for 2 SMPS units if doubling amps but that would take some space (wish to fit it inside an old amp) and TBH I am not sure one can hear the difference between 1 and 2 SMPS once enough current is delivered AND possibly some additional intermediate large filtering caps are fitted between PS and the amps. It is low cost and easy to implement (vs linear PS) so money can be spent elsewhere as I have more tweaks in mind. It is my first Class D DIY project, so if I missed anything of course just shout
First thing would be to test that single amp with that PS on the Klipsch against a recent Onkyo integrated amp in the mid-fi bracket. That one has a direct line out, so we could compare directly the Aiyima vs the Onkyo just using the latter's "digital receiver" section.
If it performs, it stays and we consider op amp rolling, ouput coils and many other bits. If it doesn't, then well that would have been a non expensive amp for the summer that can probably still find its place in the garage when working on the car, so no real harm at that price.
All this might take some time as I am quite busy and don't have the space to do that at home - my very own HIFI progresses are miserable :-(( - so requires me working at my friend's, but might be fun for the summer.
That's the intention, let's try to purchase the unit(s) first
Claude
I am likely to take the plunge with my summertime puzzle, tweaking an Aiyima TPA3251.
Reasons are I am intringued by the feedbacks of Rich and Danny and a very good friend of mine is thinking about a low cost temporarly amp, so even if that one doesn't deliver no drama, and if it does, who knows, it could even become permanent.
That VG friend has similar tastes as mine, likes my system a lot and used to be classical trained piano player from a family of reknown musicians. We have close tastes and similar ears when it comes to music, but for that system he wants it a bit more rock and roll and crashing than classic listening, hence me advising to purchase a pair of Klipsch RF-8000F.
These are quite good speakers, efficient and easy to drive, came delivered new at less than 800E and are the first of the non classical Klipsch I heard that are not too coloured to my ears (say not only fine for home theater or not too horn sounding). Of course they rock if needed but can play classic and jazz with good resolution, although they seem to be no match re finesse and 3D soundstage for panels / magnetostats I cherish so much but they seem (so says my friend who knows my system) to deliver quite a good resolution if no deep soundstage. I have yet to listen to them properly run in with a known amp to make my mind, but more importantly HE is very happy and these are keepers for him.. So that's the baseline to play with and they should be revealing enough to see where that amp ranks and how it reacts to tweaks. Oh, Rich of course was inspirational as running Klipsches aswell, although not in the same league, so if that worked OK for him...
I have a few ideas re tweaking in mind, one being to feed the amp through a low ripple SMPS that could accept 1 to 2 of these amps ( we are considering possibly a biamping option, vertical or horizontal, TBD, so we may perhaps buy 2 amps which would be handy also for direct tuning evaluation/monitoring).
I don't know what is inside the amp, but if following TI's papers my guess is that a Mean Well LRS-350-24 could well do the trick. It can be tuned up to 28.8V (I don't mind limiting from 120 to 120W under 8R as unlikely the chip will deliver and if not loud enough /distording than either way another amp would be the fix), it is lower ripple than the usual SMPS (150mVpp) and should with 14.6A provide enough current to drive even 2 of these amps at lower loads should it be required (given the board I don't trust anyway really 6A into these, but well). Oh, and the built in fan is very unlikely to kick in in our case.
We could also go for 2 SMPS units if doubling amps but that would take some space (wish to fit it inside an old amp) and TBH I am not sure one can hear the difference between 1 and 2 SMPS once enough current is delivered AND possibly some additional intermediate large filtering caps are fitted between PS and the amps. It is low cost and easy to implement (vs linear PS) so money can be spent elsewhere as I have more tweaks in mind. It is my first Class D DIY project, so if I missed anything of course just shout
First thing would be to test that single amp with that PS on the Klipsch against a recent Onkyo integrated amp in the mid-fi bracket. That one has a direct line out, so we could compare directly the Aiyima vs the Onkyo just using the latter's "digital receiver" section.
If it performs, it stays and we consider op amp rolling, ouput coils and many other bits. If it doesn't, then well that would have been a non expensive amp for the summer that can probably still find its place in the garage when working on the car, so no real harm at that price.
All this might take some time as I am quite busy and don't have the space to do that at home - my very own HIFI progresses are miserable :-(( - so requires me working at my friend's, but might be fun for the summer.
That's the intention, let's try to purchase the unit(s) first
Claude
Klipsch RP-8000F
Hello,
I use TPA3255EVM (op-amps bypassed) also with Klipsch RP-8000F with very good result
Concerning the Klipsches soundstage Consider extreme toe-in meaning axis crossing way in front of listening position It works very well
Succes with modding the Aiyima amplifier
Enjoy
Hello,
I use TPA3255EVM (op-amps bypassed) also with Klipsch RP-8000F with very good result
Concerning the Klipsches soundstage Consider extreme toe-in meaning axis crossing way in front of listening position It works very well
Succes with modding the Aiyima amplifier
Enjoy
Thanks for the advice!
I told my friend to experiment but not too that extreme, but he seems to prefer them parallel... but then he waits for us to turn up during the WE to fine tune his existing system, LOL
Will consider that option asweel and good to read that it works in your case!
Claude
I told my friend to experiment but not too that extreme, but he seems to prefer them parallel... but then he waits for us to turn up during the WE to fine tune his existing system, LOL
Will consider that option asweel and good to read that it works in your case!
Claude
When I was using a Sure 500W Tamp module I tested different power supplies with different Voltages and Amperage. Under real usage (a 15" PE sub driver with an miniDSP .9V as crossover), no matter what I send to it , I could not get the amp to suck more than 2.5-3 Amps. The recommended Amperage by the manufacturer was 10 Amps, but that would be a real case scenario if you feed the amp with test tones at 1.7 Volt pp or higher with 3 Ohms static load with resulting distortion over 10% . IMHO hardly a practical and real work scenario.
Sizing the power supplies to your needs by probing the current load seems to me a more effective approach than buying a very oversize power supply just to comply with an unrealistic manufacturer specs, unless (such as in my Sure module), your goal is to drive 3 Ohms speakers with 10% distortion at close to 2Vpp test tones.
That said, buying two power supplies of smaller power for two modules can cost (for the power supplies) less than buying a single larger one. Performance not use is just overpriced heat. I personally like the LED waterproof bricks as they need not cooling and they are designed to operate 24/7/365 at full power.
If a power supply requires a cooling fan to operate, some can only provide full power for a short time. After that the cooler kicks is. Even if at that time the cooling is not enough, the power supply will derate itself, reducing output to be able to comply with thermal specs.
Derating is many times ignored, and we like to just look at the cover page of the specs, but if the part has a good manual, on page two there are curves that plot Voltage or Amperage vs load resistance and external temperature. You will surprise yourself if you look at these curves, since many products out there, even from tier one manufacturers (Intel I7 cpus are sometimes at real usage slower that i5 due to thermal throttling) will loose a significant portion of performance under load outside the lab.
Sizing the power supplies to your needs by probing the current load seems to me a more effective approach than buying a very oversize power supply just to comply with an unrealistic manufacturer specs, unless (such as in my Sure module), your goal is to drive 3 Ohms speakers with 10% distortion at close to 2Vpp test tones.
That said, buying two power supplies of smaller power for two modules can cost (for the power supplies) less than buying a single larger one. Performance not use is just overpriced heat. I personally like the LED waterproof bricks as they need not cooling and they are designed to operate 24/7/365 at full power.
If a power supply requires a cooling fan to operate, some can only provide full power for a short time. After that the cooler kicks is. Even if at that time the cooling is not enough, the power supply will derate itself, reducing output to be able to comply with thermal specs.
Derating is many times ignored, and we like to just look at the cover page of the specs, but if the part has a good manual, on page two there are curves that plot Voltage or Amperage vs load resistance and external temperature. You will surprise yourself if you look at these curves, since many products out there, even from tier one manufacturers (Intel I7 cpus are sometimes at real usage slower that i5 due to thermal throttling) will loose a significant portion of performance under load outside the lab.
Yep, thanks, I went for the aformentioned as worst case it gets an easy job driving one amp... and as we just decided to go biamping it will indeed in real life still have enough room as I don't believe the claims of several amperes, but well.
It is better in that case to go one brick as we may have to dismantle both amps to build with them and the PS a dedicated DAC / amp... which will have to fit inside a 70's amp case as it is sentimental.
Thanks again
Claude
It is better in that case to go one brick as we may have to dismantle both amps to build with them and the PS a dedicated DAC / amp... which will have to fit inside a 70's amp case as it is sentimental.
Thanks again
Claude
Hi Rich,
As that one is more intented to you as you have been there... but anyone feel free to step in.
I will of course follow your advice and modify the output filter.
My friend's LS have an impedance between 9 and 11 Ohms from 10 to 20kHz, so somewhere one can feel that L = 12.5uH is probably better suited in his case than the 7uF.
Having said that, I have no clue on what is in the unit now as I haven't even seen it. Could you please enlighten me?
1- What ouput filter configuration is Aiyima using? I presume (?) BTL common mode perhaps?
2- What is the value of the existing Caps they use? Is that 0.68uF x 4 for both channels... is that the same in all units?
It might be convenient to keep the caps, but on the other hand I might have to change them for other values to have the right filter, depending on what I find. So might have to play with L and C, of course perhaps I am just reinventing the wheel... if so just shout!
3) Will play with the TI filter designer tool... have just downloaded it, looks nice...what is it they call Load factor (note: not Rload or speaker load)
Many thanks!
Claude
As that one is more intented to you as you have been there... but anyone feel free to step in.
I will of course follow your advice and modify the output filter.
My friend's LS have an impedance between 9 and 11 Ohms from 10 to 20kHz, so somewhere one can feel that L = 12.5uH is probably better suited in his case than the 7uF.
Having said that, I have no clue on what is in the unit now as I haven't even seen it. Could you please enlighten me?
1- What ouput filter configuration is Aiyima using? I presume (?) BTL common mode perhaps?
2- What is the value of the existing Caps they use? Is that 0.68uF x 4 for both channels... is that the same in all units?
It might be convenient to keep the caps, but on the other hand I might have to change them for other values to have the right filter, depending on what I find. So might have to play with L and C, of course perhaps I am just reinventing the wheel... if so just shout!
3) Will play with the TI filter designer tool... have just downloaded it, looks nice...what is it they call Load factor (note: not Rload or speaker load)
Many thanks!
Claude
If you already have the PS and you bought them for full manufacturer specs, keep what you have. In real life you have a saefty / utilization factor of 2.5 times your need.
I find that in designing modular amps, you have to balance budget with "specs" and expectations. Once you round the cost of the case, connectors, power supplies, amp modules, soft starts, power filters and other miscellaneous staff, a project can be rather pricey, to the point that a similar commercial good amp can be more cost effective or a better value for a little more (specially if you consider the risk of electrical fires and hazard and all the hours you will invest on).
That is the same for the enclosed aiyima blocks. Depending on your mods (sophisticated op-amps, chokes, caps, quality power supply), your cost can easily exceed $200 for a $50 amp not counting the hours it takes for a comprehensive component upgrade. I am not so sure if at that price the aiyima 3251d2 is a good value
Last edited:
Indeed... well, hours I won't count them as I charge 4 figures a day in my professional life, so let's assume pure hobby and gift for a friend
Cost has to be reasonable, so that's 40E per amp, 30E for a PS that can easily feed both should we want to go biamping. Op amp rolling, I do the boards myself and have many flavours for him to try, then once decided it will be a few euros each as I have some lying around so at it will be cost (be it OPA1656, AD825 Biased in Class A or ADA4827-1).
Then the coils, indeed say 25E per amp, a few cap tweaks here and there for a few euros as I know where to get them and, should it fit, a better pot for less than 15E.
So yes, at the end the mods can double the price of the amps and 30E for PS and say 10 for cables etc. Make that less than 140E for 1 full amp and less than 240E for biamping if needed. That's the max I want to go as from 500E upwards I would go Class A or A/B DIY., or used top amps, or tweak other amps etc. (and that would indeed cost a little fortune withe proper connectors, soft start, DC protect LOL)
But back to topic: Question is: how good will it sound for 140E / 240E? We have some commercial amps available for comparison And my friend is willing to play for that kind of money, bearing in mind worst case it is used in one of his secondary systems
But let's not get carried away indeed with extras, but as long as VFM and fun is there... let's tweak that small beast.
Claude
Cost has to be reasonable, so that's 40E per amp, 30E for a PS that can easily feed both should we want to go biamping. Op amp rolling, I do the boards myself and have many flavours for him to try, then once decided it will be a few euros each as I have some lying around so at it will be cost (be it OPA1656, AD825 Biased in Class A or ADA4827-1).
Then the coils, indeed say 25E per amp, a few cap tweaks here and there for a few euros as I know where to get them and, should it fit, a better pot for less than 15E.
So yes, at the end the mods can double the price of the amps and 30E for PS and say 10 for cables etc. Make that less than 140E for 1 full amp and less than 240E for biamping if needed. That's the max I want to go as from 500E upwards I would go Class A or A/B DIY., or used top amps, or tweak other amps etc. (and that would indeed cost a little fortune withe proper connectors, soft start, DC protect LOL)
But back to topic: Question is: how good will it sound for 140E / 240E? We have some commercial amps available for comparison
And my friend is willing to play for that kind of money, bearing in mind worst case it is used in one of his secondary systemsBut let's not get carried away indeed with extras, but as long as VFM and fun is there... let's tweak that small beast.
Claude
Based on this statement I just added up my purchased mod list...
So far my Aiyima TPA3251, PSU, and all mod parts comes out to roughly ~$162. I'm guessing about 4 to 5 hours of time for my build, because of limited space some parts will need to be moved to other locations.
Rob43
Last edited: