Hi Christian,
It is not clear (to me at least) which ripple you want to measure?
For the SMPS: if you have such two safety capacitors forming a mid-point at the chassis, and if you have earth terminals in the power outlets where you live, you can modify the 230V connection to modern standards:
Open the SMPS, find the two capacitors at the input from the power cable and remove the two capacitors.
Replace the probably two wire net cable with a three wire net cable and a power plug with earth.
For the three wire net cable, connect the yellow/green to chassis, the brown or red to phase input and the blue to neutral input.
Then you have no more tingling when touching the chassis. It would be a pity to discard a working SMPS just because it is made to old standards.
Enjoy Graz!
It is not clear (to me at least) which ripple you want to measure?
For the SMPS: if you have such two safety capacitors forming a mid-point at the chassis, and if you have earth terminals in the power outlets where you live, you can modify the 230V connection to modern standards:
Open the SMPS, find the two capacitors at the input from the power cable and remove the two capacitors.
Replace the probably two wire net cable with a three wire net cable and a power plug with earth.
For the three wire net cable, connect the yellow/green to chassis, the brown or red to phase input and the blue to neutral input.
Then you have no more tingling when touching the chassis. It would be a pity to discard a working SMPS just because it is made to old standards.
Enjoy Graz!
Last edited:
You mention burn-in.
Burn-in, I have been taught, is important to eliminate defective items following the "bath-tub curve" showing the frequency of failure is dropping rapidly in the first two days. After these two days, the frequency remains low and stable for long, until at the end it increases from an aging effect.
Certain items, like speakers and perhaps vacuum tubes, may change characteristics during burn-in. For speakers, the cone will become more soft from initial use.
But, I have never heard that semiconductors change characteristics (if they do not fail) during burn-in. Do you do burn-in for a better sound?
Burn-in, I have been taught, is important to eliminate defective items following the "bath-tub curve" showing the frequency of failure is dropping rapidly in the first two days. After these two days, the frequency remains low and stable for long, until at the end it increases from an aging effect.
Certain items, like speakers and perhaps vacuum tubes, may change characteristics during burn-in. For speakers, the cone will become more soft from initial use.
But, I have never heard that semiconductors change characteristics (if they do not fail) during burn-in. Do you do burn-in for a better sound?
Last edited:
do not operate a class D am without load!!! --> connect a speaker or a resistive load on each channel.
the internal frequency is pushing through the output filter and heat them up.
for me the burning phase is switch on and of for more then 5 times overall. e.g. over some days switch off the psu and the amp 5 times and let it run 5-10 hours.
this is a kind of getting the things older and more stable.
at this amp its a different if its very new---needs a week.... but if you change just the op amp i guess on day burn in is good enough. ...my experience
hi FF
i want to measure the ripple current/voltage of this smps. good smsps rated 100-200mV.
i get once this strange buzzy feeling, but change something on the smps ...i will see if i have time to change.
thank you very much
Speakers connected of course. tpa3116 for example cold as ice in same conditions
I can't imagine a reason for semiconductors to change or have a break-in period. Yes, tubes would, as there are physical/chemical changes going on in the tubes and on the anode/cathode. Speakers would, of course, as they are physically moving and the materials are loosening up. But I suspect that semiconductors and digital chips do not have a break-in and that a perception thereof is a confirmation bias. The mind is powerful.
jop...i agree...i realized this at my brothers P6 from hifiakademie
at the fx502spro i really dont care or realized this.
I understand measuring a signal in the order of 100mV when you may have ground-currents from a 116Vac signal can be difficult with a normal oscilloscope. You previously told me you do not have a battery operated oscilloscope. What about relying on an AC measurement with a multi-meter? It is floating potential wise.
As also explained by chermann, even without any audio output signal the amplifier is constantly switching between the two supply rails at a frequency of 300-500KHz. This we won't hear because of the frequency being far outside the audio band and the output filter capacitor absorbing far the most of the ripple current circulated in the LC-circuit (operated above the resonance frequency). But for the output filter chokes it is a life as usual in a low-pass resonant circuit: current going one way and then current going the other way etc. Each cycle leaves losses, in particular in the choke cores (you can read about core-losses on the Net) and with 300000-400000 cycles per second the loss energy adds up to heating you will notice. The larger losses per cycle, the worse.
This current will evidently also pass the TPA3116 output switches. But, the TPA3116/TPA3118 has got very low impedance switches: 120mOhm stereo and 60mOhm BTL (better than TPA3255 normal!) So, the TPA3116 chip can stay cool while the heat builds up in the cores. Cores are generally not good at passing heat on to the air.
Last edited:
I looked at the TPA3255 datasheet and it seems the chip uses both modulators in BTL, thus, separate modulators for each output pin. Logically, TI could have done the control of the second output by inverting the logical signals for the first output, but TI chose the other solution. TPA3116 is "born" BTL, TPA3255 is not. In TPA3116 I guess they just invert the signals for the second output. For PBTL the only really safe way is to generate the control signals at the logical level.
I got the LME49720NA chips from the TI Store, they are the dip8 format. This seems like the best deal out there when there is free shipping.
How do people characterize the difference in sound? I put them into one of my FX502SPRO amps, but I can't tell the difference because I can't really do a test where I switch back and forth quickly on the same song. I think confirmation bias would be very probable in trying to evaluate these chips, so how does one do a valid test?
hi swank42
dont panic...listen to the music should not force in stress
take your time for about 4-6 favorite songs and play this at your favorit volume level.
as often as you want -i think 3-4 time playing all song or sequences of it is enough.
then switch back to the other amp
start the same procedure.
i personally do this switch back 2-3 times so its a lot of wokr
...but dont forget -dont be stress if you dont hear a big step of differrence the first time.enjoy the music what you like....maybe at special sequences you can hear the difference first.
i listened since 30 years to audio..(yes i started with 17...my friends think i get stupid
)...so my brain is a little bit trained for this diffrence
chris