@JPS64 you laid out the Symlph and it measures very well on ASR https://www.audiosciencereview.com/...ics-sylph-d200-amplifier-module-review.25295/
I have no doubt that your designs will be a real 'banger'!
As an aside regarding DSP and what not. A lot of DIY seems to resist change. My JBl P308 MK2s destroy my Fusion 10 pures at half the cost per unit and the sound they output makes my poor DIY speakers weep. That's why I bought two PA3s and a driverack PA2 [to replace AVR amping and Fusion 10 XO]. Affordable active + DSP offers the greatest advancement we have seen since neo magnets and advanced polymers became the norm in speaker construction. It can pay for itself in Xo parts and room correction experiments; it allows compensation for hearing damage, etc. Powered speakers with DSP destroyed all my preconceived notions.
ETA: The bit in [ ] for clarity.
ETA2: Topping is not replying to my firmware query. If anybody is really familiar with STM8S microcontroller programming, let's talk XD
I have no doubt that your designs will be a real 'banger'!
As an aside regarding DSP and what not. A lot of DIY seems to resist change. My JBl P308 MK2s destroy my Fusion 10 pures at half the cost per unit and the sound they output makes my poor DIY speakers weep. That's why I bought two PA3s and a driverack PA2 [to replace AVR amping and Fusion 10 XO]. Affordable active + DSP offers the greatest advancement we have seen since neo magnets and advanced polymers became the norm in speaker construction. It can pay for itself in Xo parts and room correction experiments; it allows compensation for hearing damage, etc. Powered speakers with DSP destroyed all my preconceived notions.
ETA: The bit in [ ] for clarity.
ETA2: Topping is not replying to my firmware query. If anybody is really familiar with STM8S microcontroller programming, let's talk XD
Just got these.
The cheapest i could find(€20 with shipping)
It plays right out of the package.
No noise or hum.
I will make some baseline measurements before i start to modify.
It has line in, bluetooth in and USB in autoselect.
2 channel.
TL072 input buffers.
Switching 5v supply.
The cheapest i could find(€20 with shipping)
It plays right out of the package.
No noise or hum.
I will make some baseline measurements before i start to modify.
It has line in, bluetooth in and USB in autoselect.
2 channel.
TL072 input buffers.
Switching 5v supply.
Incredible price for what you get. But.. how does it sound?!
Small coils and is that a 4 gang potentiometer!?
Small coils and is that a 4 gang potentiometer!?
No,it is a stereo pot with on/off switch.
First impression on small JBL speakers is good. No hum, no hiss.
https://nl.aliexpress.com/item/1005..._list.0.0.5d9979d26R2E4F&gatewayAdapt=glo2nld
The curse of MA12070 is the destination for mass production. A sheer amount of necessary caps don’t make it brighter for DIY but any endeavour should be encouraged.
Mmm, in case you want to look up the datasheet of the BT receiver:
https://doc.zh-jieli.com/vue/#/home
Probably a cousin of this one, there exist many from this manufacturer:
http://www.chenbingdom.com/uploadfile/20172151033299946.pdf
https://doc.zh-jieli.com/vue/#/home
Probably a cousin of this one, there exist many from this manufacturer:
http://www.chenbingdom.com/uploadfile/20172151033299946.pdf
Last edited:
This evening connected my Dynaudio Audience 122. Now the MA12070 sound good by low soundlevel at some higher soundlevel distorsion kicks in. It has not enough clean power. My 20Watt tube amp does a lot better job. The Purifi 1et400a modules are much better. This amp is not for big power hungry speakers. I shall modified it to mono PBTL and change 5v powersupply so i can use 24v. Now you can not use more then 19,5V or the 5V onboard powersupply blows up.
It is not the chip but the implementation on this 20 Euro board. Probably a combination of too small coils, layout issues, part quality (the caps, opamps) and such. Maybe the PSU you use is not suitable for this task either. PSU voltage is chosen low so less output power. Many variables. Here a Sabaj A10a (single MA12070, 300W SMPS of unknown voltage but probably 26V) plays with 6 Ohm 88 dB/1 meter loudspeakers. No such issues but I did notice a very faint whine from the SMPS when you are with your ear almost against the loudspeaker.
Purifi 1ET400A (425W output power) costs 329 Euro/piece so that is at least 660 Euro so at least a factor 33 more expensive. It would be very strange if a 20 Euro module would be equal or better.
Purifi 1ET400A (425W output power) costs 329 Euro/piece so that is at least 660 Euro so at least a factor 33 more expensive. It would be very strange if a 20 Euro module would be equal or better.
Last edited:
I use a good lab power supply who can deliver 10A at 19V. Amps draws about 2A when it start to distort. But i will remove tl072 inputbuffer and use my own design output filter.
Further i will make better 5v supply and ad film capacitors. And run it in PBTL with 24V with balanced input.
The Purifi is my reference and i do not expect that the ma12070 will do a better job. But it is fun to look what is possible.
Further i will make better 5v supply and ad film capacitors. And run it in PBTL with 24V with balanced input.
The Purifi is my reference and i do not expect that the ma12070 will do a better job. But it is fun to look what is possible.
You could wait with the PBTL till the other modifications are done just to see if the issues are solved. I used OPA2156 in the BRZHiFi with good results compared to the TL072s that it had. No one knows what parts are good quality in a 20 Euro module and if they are not fake/recycled.
You seem to confuse the IC with the module. For now the Nvarcher module sucks (as could be expected). You would have had more success with the 50 Euro PL-AD-160 module where quite some things are done right.
You seem to confuse the IC with the module. For now the Nvarcher module sucks (as could be expected). You would have had more success with the 50 Euro PL-AD-160 module where quite some things are done right.
Last edited:
@Koifarm 's soeakers are the Dynaudio Audience 122 whhich are 6-Ohm, 86dB sensitivity. I think he would probably be disappointed with any MA12070 not in PBTL and may still be a bit disappointed depending on room size, typical listening loudness, and distance from the speakers.
Comparing it to a tube amp is also not entirely fair. They are both lower wattage (than say a 100-200w class AB) but I personally find tubes to be a lot kinder to the ears than chipamp distortion. Beam forming vs switching and silicon breakdown.
THe current handling capability will just depend on the dielectric substrate and transistor design baked into the chip itself. I think the only way to coax more power out of the MA12070 would be some sort of sub-ambient cooling, so I think @Koifarm's plans to go PBTL and higher voltage will probably get him a few more watts than the 20W tube amp. Maybe liquid cooling would get a few more watts out of the chip since Infineon shows it at saturation being 125C?
But ASR tests show ~15WPC into 4-Ohms before distortion skyrockets which maps to ~80C on Infineon's application notes.
So liquid cooling might allow more headroom with power. In theory that barrier should be stronger at lower temperatures in every silicon structure. Infineon shows a fairly high thermal resistance in every configuration but a decent waterblock with very good EPAD contact (assuming the chip is well bonded to the same epads) should achieve under 1*C/W. A single 120mm PC radiator with fan could likely cool 2-4 chips assuming the pump used could achieve the required head pressure.
So who is going to try it out? 😀
Great info Jean-Paul.
I like the changes i did today🙂
Today did the Pbtl mono mod and removed the TL072. Further i used/build my own output filter.
Used 24V , no problem with 5v onboard supply. I can substantial drive my dynaudio loudspeakers louder.
Filter change giver better bass control but also the mids/hights are more in balance.
Removing the TL072 and drive the MA12070 Gives less distorsion and noise. But when i used unbalanced setup the noise get ups(10 a 15db). Also when nothing is connected the input pickup a lot of noise.
Next i will use opa1632 input board(€ 3,-) and try to remove the input capacitors with the vcom correction of the opa1632.
Then i will change chargingcaps and when still in use inputcaps.
After that it is time for modify the 5V supply.
I like the changes i did today🙂
Today did the Pbtl mono mod and removed the TL072. Further i used/build my own output filter.
Used 24V , no problem with 5v onboard supply. I can substantial drive my dynaudio loudspeakers louder.
Filter change giver better bass control but also the mids/hights are more in balance.
Removing the TL072 and drive the MA12070 Gives less distorsion and noise. But when i used unbalanced setup the noise get ups(10 a 15db). Also when nothing is connected the input pickup a lot of noise.
Next i will use opa1632 input board(€ 3,-) and try to remove the input capacitors with the vcom correction of the opa1632.
Then i will change chargingcaps and when still in use inputcaps.
After that it is time for modify the 5V supply.
Yes the chip hisses when not driven actively/terminated and it needs to be driven balanced as well hence the active UNBAL/BAL conversion.
I suspect those TL072s in cheap amplifiers to be Chinese copies as the difference with OPA2156 just was too large.
I suspect those TL072s in cheap amplifiers to be Chinese copies as the difference with OPA2156 just was too large.
Last edited:
Today i used the opa1632 on the input. This is a lot better. No noise when no cable connected. No noise(<-100db) whit balanced input, still a little more noise( -88db) with RCA input. Also tried Lundahl LL1521 input trafo but this gave a lot of noise/static.
Tried also to remove the inputcapacitors(2u2) by using Cref(MA12070) and Vcom(OPA1632), but these are not compatible. Just a pumping amp.
The idea was to get DC on output OPA1632 on same level as DC input MA12070.
Tried also to remove the inputcapacitors(2u2) by using Cref(MA12070) and Vcom(OPA1632), but these are not compatible. Just a pumping amp.
The idea was to get DC on output OPA1632 on same level as DC input MA12070.
I have boxed in one at chassie, one of them the other it's outside, chassie to small right now.
Shorted some cables and twisted.
Noise?
Yes if I get as close to the elements.
I have some EMI as jean-paul say.
But buy two new Dac's for balanced?
I wait.
Best sound ever I have had.
Not much power but enough for me.
But it's there!
JPS64 and jean-paul make an board for us!
🙂
Shorted some cables and twisted.
Noise?
Yes if I get as close to the elements.
I have some EMI as jean-paul say.
But buy two new Dac's for balanced?
I wait.
Best sound ever I have had.
Not much power but enough for me.
But it's there!
JPS64 and jean-paul make an board for us!
🙂
Last edited: