Sounds good!
Very high efficiency speakers... Note that these class D amps do produce actualy comparatively more distorsions (still very low on the paper) at low power levels than at higher ones, but as said no problem with my 1W or so.
With that efficiency you could of course go Class A, but that's another story
Let us know how it sounds once done
Viel Spass
Claude
Very high efficiency speakers... Note that these class D amps do produce actualy comparatively more distorsions (still very low on the paper) at low power levels than at higher ones, but as said no problem with my 1W or so.
With that efficiency you could of course go Class A, but that's another story
Let us know how it sounds once done
Viel Spass
Claude
Hello Claude,
Class D has problems below 1W and most of these cheaper amps have rising THD above 6kHz. I have Neurochromes here too and maybe I try them above 2KHz.
Does the 3251 run on 600kHz using 10Kohm? Do you know that or is it possible to change to 600kHz changing the resistor if it is not so?
My 3255 boards came all with 30Kohms for 450kHz (TI sais normal adjust). 3251 has, acc. to datasheet, 600kHz as "normal". Maybe this is the reason why many say 3251 sounds a bit better.
Which Class A amps do you mean?
Have a nice evening.
Ralf
Class D has problems below 1W and most of these cheaper amps have rising THD above 6kHz. I have Neurochromes here too and maybe I try them above 2KHz.
Does the 3251 run on 600kHz using 10Kohm? Do you know that or is it possible to change to 600kHz changing the resistor if it is not so?
My 3255 boards came all with 30Kohms for 450kHz (TI sais normal adjust). 3251 has, acc. to datasheet, 600kHz as "normal". Maybe this is the reason why many say 3251 sounds a bit better.
Which Class A amps do you mean?
Have a nice evening.
Ralf
Hallo Ralf
Yep, that's what I said, at low levels distorsions are less good... TI provides you with various papers, that gives you hints on results. Albeit we don't know the 04 schematic they seem in that regard (on the power section) very close to the TI test board... at least on the 3 04amps I had on hands and tuned.
So we can only presume, with strong assumption, that it works at nominal frequency etc., but only you can tell how your unit really works as the manufacturer seems to take quite some liberties regarding production variations.
As of me, I do believe that, if correctly implemented, then indeed higher working frequency is better, at least in these frequency ranges. But then of course there must be many other variables - investigating all of them must be fascinating but I only "droped by" in the Class D section : quick in, quick out, my interest at the moment lies elsewhere.
Class A... my latest build was the initial Lottery Pass VFET, and that's what I use with the said Klipsch LS in my secondary system. I don't need more than 5W per channel and don't mind wasting 110W in SET Class A LOL, but that's me and a particular use.
I am also on the ACA mini for the fun, need to get that one started... and more seriously, for my own HIFI system in a dedicated room, on something far more complex than all these smallish amps: the FSSA-2, highly biased in Class A. That's another topic, quite a development as opposed to tweaking existing kits and (I hope) not comparable soundwise - well, only results will tell and that will take time as quite busy on many non audio projects LOL
Take care
Claude
Yep, that's what I said, at low levels distorsions are less good... TI provides you with various papers, that gives you hints on results. Albeit we don't know the 04 schematic they seem in that regard (on the power section) very close to the TI test board... at least on the 3 04amps I had on hands and tuned.
So we can only presume, with strong assumption, that it works at nominal frequency etc., but only you can tell how your unit really works as the manufacturer seems to take quite some liberties regarding production variations.
As of me, I do believe that, if correctly implemented, then indeed higher working frequency is better, at least in these frequency ranges. But then of course there must be many other variables - investigating all of them must be fascinating but I only "droped by" in the Class D section : quick in, quick out, my interest at the moment lies elsewhere.
Class A... my latest build was the initial Lottery Pass VFET, and that's what I use with the said Klipsch LS in my secondary system. I don't need more than 5W per channel and don't mind wasting 110W in SET Class A LOL, but that's me and a particular use.
I am also on the ACA mini for the fun, need to get that one started... and more seriously, for my own HIFI system in a dedicated room, on something far more complex than all these smallish amps: the FSSA-2, highly biased in Class A. That's another topic, quite a development as opposed to tweaking existing kits and (I hope) not comparable soundwise - well, only results will tell and that will take time as quite busy on many non audio projects LOL
Take care
Claude
Hello Claude,
if you say that Low ESR is more important than big capacity may I better take 4x680uF (0,025Ohm) in parallel for the filter? I have some very good Nippons.
I read some time ago that the Neurochrome has 0,3W or so in Class A. Could be enough for moderate levels using horn-speakers with 110-118dB/1W.
What for is the 0,05Ohm resistor in the filter?
Thanks and have a nice weekend?
Ralf
if you say that Low ESR is more important than big capacity may I better take 4x680uF (0,025Ohm) in parallel for the filter? I have some very good Nippons.
I read some time ago that the Neurochrome has 0,3W or so in Class A. Could be enough for moderate levels using horn-speakers with 110-118dB/1W.
What for is the 0,05Ohm resistor in the filter?
Thanks and have a nice weekend?
Ralf
Hallo
The R is part of the RLC filter... for more information look at one half of Mark Johnson's excellent PO89ZB SMPS filter. 'Mine' is of course different re overall specs and cut frequency and attenuation etc., but the general principle is nearly identical to one half of it. There is a full thread on it and I guess far better explained than I could ever (the alternative being googling RLC low pass filter or running a simulation).
Re caps, you want some that are low ESR at high frequencies for the chip. In the past some (I remember Stereoplay's then to become AVM power amp kit in the 90s and many others) adviced to parallel caps to lower ESR and price. That was true back then, I did it a lot. But times change, no statement is frozzen re what is better. For my next amp I actually found out that today's big caps, if well chosen, can better that. So back to big caps on some occasions.
There is no clear cut here, it all depends on specs. Oh, and one last thing, on how you connect them. I knew some commercial A-B amps with 48 small caps in parallel, low equivalent ESR indeed (although one wants to look at different frequencies as ESR at low F is something of a problem sometimes), but can't help shaking my head as distance means the connecting wires or board circuit adds then a significant R itself which can counteract low ESR.
In short : best ESR for the appropriate frequency range as close as possible to the power need, means low impedance PS. From there on you can always add caps a bit further away, can be better or not, but rely then more on overall filtering effect than ESR reduction unless you run a simulation proving it is better. or of course you like it better to your ears/brain, as at the end that's what it is built for.
All IMHO and nice WE to you aswell
Claude
The R is part of the RLC filter... for more information look at one half of Mark Johnson's excellent PO89ZB SMPS filter. 'Mine' is of course different re overall specs and cut frequency and attenuation etc., but the general principle is nearly identical to one half of it. There is a full thread on it and I guess far better explained than I could ever (the alternative being googling RLC low pass filter or running a simulation).
Re caps, you want some that are low ESR at high frequencies for the chip. In the past some (I remember Stereoplay's then to become AVM power amp kit in the 90s and many others) adviced to parallel caps to lower ESR and price. That was true back then, I did it a lot. But times change, no statement is frozzen re what is better. For my next amp I actually found out that today's big caps, if well chosen, can better that. So back to big caps on some occasions.
There is no clear cut here, it all depends on specs. Oh, and one last thing, on how you connect them. I knew some commercial A-B amps with 48 small caps in parallel, low equivalent ESR indeed (although one wants to look at different frequencies as ESR at low F is something of a problem sometimes), but can't help shaking my head as distance means the connecting wires or board circuit adds then a significant R itself which can counteract low ESR.
In short : best ESR for the appropriate frequency range as close as possible to the power need, means low impedance PS. From there on you can always add caps a bit further away, can be better or not, but rely then more on overall filtering effect than ESR reduction unless you run a simulation proving it is better. or of course you like it better to your ears/brain, as at the end that's what it is built for.
All IMHO and nice WE to you aswell
Claude
Just want to contribute something after reading all this mods in aiyima A04 in this forum
I think Aiyima place wrong value in 12v op-amp capacitor, 470uf (2 cap with green dot) is storage to much energy for anti pop to react if the amp is shut down, so i decide to replace them with 220uf Pana FC (you can use any brand) and the pop is gone! I'm only hear a fainth noise if you shutdown the amp without music in the background (if you play with music there is no pop out at all!). Well, It bother me since i bought this amp and i hope this help other people that have same problem with this amp. Sorry for my bad english, its not my main language.....
I think Aiyima place wrong value in 12v op-amp capacitor, 470uf (2 cap with green dot) is storage to much energy for anti pop to react if the amp is shut down, so i decide to replace them with 220uf Pana FC (you can use any brand) and the pop is gone! I'm only hear a fainth noise if you shutdown the amp without music in the background (if you play with music there is no pop out at all!). Well, It bother me since i bought this amp and i hope this help other people that have same problem with this amp. Sorry for my bad english, its not my main language.....
In aiyima a04 and a07, the power supply circuit for 2 opamps ic still works even if main switch is off, remaining 12v is supplied to VDD pin and GVDD will be closed through main switch. The capacitors you marked are the decoupling capacitors that feed the ic opamps as well as the VDD and GVDD pins
I don't want to screw up the A07 I am currently using for listening, so I just bought a second A07 to experiment with mods. Spend about $100 on components and shipping.
That got me to thinking. Why don't folks buy TI's eval board instead of the boards from China with cheaper/counterfeit components? Yes quite a bit more expensive, but you get XLR + RCA inputs, and direct signal into amp, no cheap "preamp" circuitry in the signal path.
I think I should have probably done that. And I wonder why ASR doesn't test those boards? One had to figure TI would know how to get the most out of their own chip?
That got me to thinking. Why don't folks buy TI's eval board instead of the boards from China with cheaper/counterfeit components? Yes quite a bit more expensive, but you get XLR + RCA inputs, and direct signal into amp, no cheap "preamp" circuitry in the signal path.
I think I should have probably done that. And I wonder why ASR doesn't test those boards? One had to figure TI would know how to get the most out of their own chip?
All valid points. The board seems to be designed very well - not surprisingly... The only thing that would prevent me from buying would be having to spend time and effort to place it in a nice case, plus the power supply. Also, it is not FCC-approved for resale... this probably means very little to potential manufacturers... but still, it could be a problem. Anyway, the board does look very nice and provides many options... as you have already identified. With a very nice power supply, the whole package (inside the case) could cost in excess of $500... not counting the time spent to get everything over the finish line.
Of course, if you are a DIY-er with time on your hands... things are somewhat different.
Of course, if you are a DIY-er with time on your hands... things are somewhat different.
I believe the Meanwell 36v 9.6 amp smps is < $70. TI spec sheet for the eval board specifically touts the sound quality... I would not care about a case but was thinking I might try to fit it into my ClassDAudio case and perhaps even use the linear power supply that came with that amp. That amp's board is somewhat obsolete now, but linear supply not so much...
I'm tempted to return unopened the second A07 arriving tomorrow (or just give it away) and just go for the eval board.
And yeah, I am retired now so nothing but time to play now that Covid nixed my community volunteer projects that I had going...
I'm tempted to return unopened the second A07 arriving tomorrow (or just give it away) and just go for the eval board.
And yeah, I am retired now so nothing but time to play now that Covid nixed my community volunteer projects that I had going...
Hi guys,
A few points, quite quickly, as this has been posted more or less in the past...
First off, as per title this is a 04 specific thread: if related to 07 could you please keep it that way? There are already existing 07 threads...
Regarding eval boards a few general points, whatever boards (amps, DACs etc.) :
Can't comment on 07, but regarding 04, the main burden here as already posted is the price as when getting a 04 amp you get efectively a working TI evalk board. Indeed, the 04 we worked on where copy-paste of the eval boards regarding the (core / sound path" schematic (all the workin rest having been conviniently sorted), with some minor variations regarding the initial non balanced OPA gain stage. Apparently that changed somewhat during production... and may have changed as who knows how the 04 evolved it its production.
In short, when purchasing a 04 you effectively get a non expensive (if majour problem just buy another one!) DIYer friendly TI eval board.... at least regarding the schematic, let's go into details. But it is very cost effective... it works... and sounds good!
Details are the following:
Let's of course not forget safety regarding casing etc, conviniently done when going the 04 route
Dare I say a tuned 04, used with reasonable power limits, sounds as good or even better than an eval board?
That's probably the case, you to find out, let us know!
All IMHO
Claude
A few points, quite quickly, as this has been posted more or less in the past...
First off, as per title this is a 04 specific thread: if related to 07 could you please keep it that way? There are already existing 07 threads...
Regarding eval boards a few general points, whatever boards (amps, DACs etc.) :
- they are often very expensive
- they are sometimes not very user friendly, not just regarding case, PS, connections etc., but also sometimes they need some at each start-up some manual sequences to be triggered to work safely. These need to be automated to make them HIFI user friendly outside of a bench
- some are a pain to work on, depending on solder use etc. Having said that, the relatively 3 early 04 amps we worked on proved to be a pain aswell for some components, these being glued and soldered in a very average way making parts replacement sometimes a pain, but well...
Can't comment on 07, but regarding 04, the main burden here as already posted is the price as when getting a 04 amp you get efectively a working TI evalk board. Indeed, the 04 we worked on where copy-paste of the eval boards regarding the (core / sound path" schematic (all the workin rest having been conviniently sorted), with some minor variations regarding the initial non balanced OPA gain stage. Apparently that changed somewhat during production... and may have changed as who knows how the 04 evolved it its production.
In short, when purchasing a 04 you effectively get a non expensive (if majour problem just buy another one!) DIYer friendly TI eval board.... at least regarding the schematic, let's go into details. But it is very cost effective... it works... and sounds good!
Details are the following:
- it is very possible TI used better tracks and larger / better quality tracks and boards, enabling better high power output... you only have to look at the boards...
- It is very possible 04 use fakes, but then as posted already these fakes, apart from a few bottlenecks we addressed, turn out to be so good - or not placed at the sound relevant locations - that at the end the amp sounds VG and comparable to some non fake units. In short, with very little work, for HIFI use you get with 04 again a cost effective eval board. Tat means 04 has a good layout etc. aswell
- Yep, there are some bottlenecks, and some improvement that could possibly also be carried out on the TI eval boards when it comes to sonic appreciation and not working bench use... these has been addressed in this thread, that was its purpose.
Let's of course not forget safety regarding casing etc, conviniently done when going the 04 route
Dare I say a tuned 04, used with reasonable power limits, sounds as good or even better than an eval board?
That's probably the case, you to find out, let us know!
All IMHO
Claude
I haven't read all 41 pages in this thread, but I wonder if someone did a schematic including all the changes in capacitors, resistors and inductors that are recommended to be made.
There's one above by trunddtmc that includes isolated copper wire used as solid tracks, and some bypassing on the input ICs that would be useful to know where they are placed, and have some schematic of the solid wires.
One thing that intrigues me is why no one seems to have considered arranging a larger box to put a larger heatsink and even put the power supply inside, with a better power connector than the A07's original one, that is very cheesy, and prone to malfunction.
There's one above by trunddtmc that includes isolated copper wire used as solid tracks, and some bypassing on the input ICs that would be useful to know where they are placed, and have some schematic of the solid wires.
One thing that intrigues me is why no one seems to have considered arranging a larger box to put a larger heatsink and even put the power supply inside, with a better power connector than the A07's original one, that is very cheesy, and prone to malfunction.