Really interesting results! I've been really happy with the opa1622 on adapter(the green pcb with virtual ground made from caps on resistors on the adaptor.) I recently bypassed those with a wire so opamp ground is connected directly to amp ground. I think it made an improvement and it measured better. Have any of you made this socket mod and then compared the opa1622 modded to the opa1656?
Poor measurements compared to ICEpower50ASX2SE
Got not very good results today with my Aiyima A04
On the Motu M4, even with 192khZ sampling, I do not see anything criminal in the waveform. Probably, still need an oscilloscope to figure out where the legs are growing: from a power source, LC-strapping TPA3251 or the wrong operating mode OPA1656 (which is most likely in unity gain judging by the very low gain)
Got not very good results today with my Aiyima A04
On the Motu M4, even with 192khZ sampling, I do not see anything criminal in the waveform. Probably, still need an oscilloscope to figure out where the legs are growing: from a power source, LC-strapping TPA3251 or the wrong operating mode OPA1656 (which is most likely in unity gain judging by the very low gain)
Attachments
i've also posted this on ASR
i purchased an A04 sometime back in 5/2020, received the tilear rev 2.1 board
recently purchased two A04s, HDX325X rev 3.1, and there is a massive power decrease
tests i just ran to demonstrate
and 'close' pics showing the detail
tilear A04 rev 2.1 (good) = 31.9Vrms with the above settings
new A04 rev 3.1 (bad) = 6.82Vrms with the identical settings!!!
the new A04 (bad) had to have the E30 attenuation set to -5.5dB to match ~31.2Vrms, 14.5dB difference
anyone have any ideas what could possibly have led to this massive decrease in output? change in caps? inductors? i was able to disassemble it but the two screws holding down the heat sink have some red glue/epoxy over them. will need a heat gun from a friend before i can attempt to get those off to see if they even are still using the tpa3251 chip or they bait and switched
i've attached some pics of the scope readings, and a pic from ASR of the 2.1 board, and a pic from me of the new rev 3.1 board
i purchased an A04 sometime back in 5/2020, received the tilear rev 2.1 board
recently purchased two A04s, HDX325X rev 3.1, and there is a massive power decrease
tests i just ran to demonstrate
- REW signal generator for 80hz @ -10dBFS
- E30 set to -20dB
- Right channel observed directly to scope, no load (can provide with a speaker load if needed)
- Volume knob on both units set to max, 14AWG ofc wire for testing
and 'close' pics showing the detail
tilear A04 rev 2.1 (good) = 31.9Vrms with the above settings
new A04 rev 3.1 (bad) = 6.82Vrms with the identical settings!!!
the new A04 (bad) had to have the E30 attenuation set to -5.5dB to match ~31.2Vrms, 14.5dB difference
anyone have any ideas what could possibly have led to this massive decrease in output? change in caps? inductors? i was able to disassemble it but the two screws holding down the heat sink have some red glue/epoxy over them. will need a heat gun from a friend before i can attempt to get those off to see if they even are still using the tpa3251 chip or they bait and switched
i've attached some pics of the scope readings, and a pic from ASR of the 2.1 board, and a pic from me of the new rev 3.1 board
Attachments
This is aiyima's explanation
i'm not entirely sure how to read that honestly
however, a) how was anyone suppose to know this, b) again, most DAC's commercially available are putting out 1.4-2Vrms, how is anyone making that higher? c) why now do we have to buy more equipment (wouldnt even know what to buy) to power an amp that at one point, was working perfectly fine with standard DACs?
i literally have 93db efficient speakers that i hooked this up to (DIYSG Volt6). At 6ft, i cannot calibrate to reference volume with a 2Vrms DAC without BOOSTING the digital signal (which would potentially cause clipping or hitting cLimiter in windows with EAPO)
EDIT: so some quick measurements, i have (2) 4.5ohm zister 20w resistors that i ran in series for a 9ohm load. on the rev 3.1, with 2Vrms (0dB on the E30 DAC, -3dBFS@80hz from REW) i'm getting 12.6Vrms output
on the rev 2.1 board, same setup, i'm getting about 21.4Vrms before i start to clip. note the volume knob is about 60% when this happens. if i push any higher than about 24-25Vrms, i hit some sort of limiter and it shuts down
on the rev 2.1 board, i maxed the volume on the amp, put E30 to -25dB, and pushed the volume up until i hit clipping around 21.4V which occurred at -10.5dB which measures 0.608Vrms of line output
since i have no way to get greater than 2Vrms of line output, i cannot get the required line it to max the rev 3.1
so it appears that in order to get max output:
rev 2.1 = 0.608V line input
rev 3.1 = unknown, >2V however
i'm not entirely sure how to read that honestly
however, a) how was anyone suppose to know this, b) again, most DAC's commercially available are putting out 1.4-2Vrms, how is anyone making that higher? c) why now do we have to buy more equipment (wouldnt even know what to buy) to power an amp that at one point, was working perfectly fine with standard DACs?
i literally have 93db efficient speakers that i hooked this up to (DIYSG Volt6). At 6ft, i cannot calibrate to reference volume with a 2Vrms DAC without BOOSTING the digital signal (which would potentially cause clipping or hitting cLimiter in windows with EAPO)
EDIT: so some quick measurements, i have (2) 4.5ohm zister 20w resistors that i ran in series for a 9ohm load. on the rev 3.1, with 2Vrms (0dB on the E30 DAC, -3dBFS@80hz from REW) i'm getting 12.6Vrms output
on the rev 2.1 board, same setup, i'm getting about 21.4Vrms before i start to clip. note the volume knob is about 60% when this happens. if i push any higher than about 24-25Vrms, i hit some sort of limiter and it shuts down
on the rev 2.1 board, i maxed the volume on the amp, put E30 to -25dB, and pushed the volume up until i hit clipping around 21.4V which occurred at -10.5dB which measures 0.608Vrms of line output
since i have no way to get greater than 2Vrms of line output, i cannot get the required line it to max the rev 3.1
so it appears that in order to get max output:
rev 2.1 = 0.608V line input
rev 3.1 = unknown, >2V however
yes, the 32V 5A was used, i apparently forgot to update the forum post as well
i had the 10x selector on my scope by accident
here are the accurate measurements, also noted in another post
"so some quick measurements, i have (2) 4.5ohm zister 20w resistors that i ran in series for a 9ohm load. on the rev 3.1, with 2Vrms (0dB on the E30 DAC, -3dBFS@80hz from REW) i'm getting 12.6Vrms output
on the rev 2.1 board, same setup, i'm getting about 21.4Vrms before i start to clip. note the volume knob is about 60% when this happens. if i push any higher than about 24-25Vrms, i hit some sort of limiter and it shuts down
on the rev 2.1 board, i maxed the volume on the amp, put E30 to -25dB, and pushed the volume up until i hit clipping around 21.4V which occurred at -10.5dB which measures 0.608Vrms of line output
since i have no way to get greater than 2Vrms of line output, i cannot get the required line it to max the rev 3.1
so it appears that in order to get max output:
rev 2.1 = 0.608V line input
rev 3.1 = unknown, >2V however"
Not saying that there couldn´t be a reason for this difference but with regards to the "limiter": I had a similar problem with one of my A04´s.
With ~0.6VRMS input the amp would switch off (blinking red LED, muted) at about 70% of the volume knob.
I tried real loads, dummy loads, no load. Made no difference.
(mind you, the gain/output voltage on my A04s were the same though)
Aiyima wanted a video of that happening and following I had a new amp in the mail.
yeah, Amir on ASR confirmed there is an internal limiter on the Aiyimas
also, i was able to confirm this behavior on 2 brand new A04s as i returned the first thinking it was defective
Do you happen to know where I can find that comment?
If you have to search yourself then don´t bother and I´ll try to find it.
(maybe you know the thread where that was posted?)
Admittedly I didn´t follow the subject any further when I received the working amp.
Now both of my A04´s are working in an active setup, volume knob at 3 o´clock and no problems.
I prematurely concluded that there was one of the TPA3251´s internal protections kicking in (there are several) but that really shouldn´t happen even at full volume. (see datasheet TPA3251 at 9.4.6 for the error reporting; I guess those outputs are simply connected to the LEDs as I couldn´t see any external fault handling on the board)
I offered Aiyima to find the fault myself if they would help and reveal the schematic but they didn´t want to go that far and sent the new amp.
There are a few parts that (I guess), when populated with wrong values, could trigger the internal protections (like the OC_ADJ Resistor for example).
And yes, even after that defect amp I still regard them as fine amps with an excellent price-quality-ratio.
Nevertheless I exchanged opamps to LM4562/OPA1656 and output inductors to Coilcrafts to sleep better ;-)
Last edited:
AIYIMA A07 TPA3255 Review (Amplifier) | Audio Science Review (ASR) Forum
"As expected, with low impedance of 4 ohm we are current limited so the 5 amp stock power supply does better, turning in 77 watts. Yes, the 300 watt rating is a fantasy but maybe it can do better with momentary peaks. Alas, its aggressive protection circuit would shut it down and not let me test max and peak power."
yeah, i can see how it can be read both ways: the power supply giving out or the protection circuit of the amp
I can say, with the rev 2.1 board, and a 2Vrms input, you can get it to hit a shut down circuit as well. again, this is with the 32v5a ps. i'm not sure the 3251 goes much past 36v? but i can't remember