I would not be to sure about that diagnosis. The very strange behaviour of the speaker impedance may have some impact on overall frequency response. Class-D LC filter output may be an issue here, depending on characteristics of the post filter feedback. Can you provide a frequency response measurement with a) nominal resistive dummy and b) no load at all?
+1
I can confirm this
But also depending on your speakers and room
1/ TPA3255 non PFFB with bypassed op-amps - image more forward and a bit harsh, on the edge
2/ TPA3255 with PFFB and op-amps - soundstage less in your face and smoother highs
Some people may like option 1
Now i am very happy with option 2
Enjoy..
D
Deleted member 148505
I appreciate that people like JLesterP, drMordor, etc, seem to be advancing the optimal use of TPA325x amplifiers, but until this development settles, I'm really enjoying my ICEpower 50ASX2 (+ jfet buffer).
On the subject of buffers, it seems that getting the "recipe" right for TPA325x amps involves using just the right opamp as a buffer, and the current favourite appears to be the OPA1656 ...
and this makes me wonder - most DAC's have an opamp-based output stage, so if that opamp happens to be an OPA16xx (such as iancanada's ES9028Q2M/ES9038Q2M DAC) maybe that would be a good candidate for direct input to a TPA325x amp, thus reducing the total number of opamps in the signal chain?
On the subject of buffers, it seems that getting the "recipe" right for TPA325x amps involves using just the right opamp as a buffer, and the current favourite appears to be the OPA1656 ...
and this makes me wonder - most DAC's have an opamp-based output stage, so if that opamp happens to be an OPA16xx (such as iancanada's ES9028Q2M/ES9038Q2M DAC) maybe that would be a good candidate for direct input to a TPA325x amp, thus reducing the total number of opamps in the signal chain?
Last edited:
PCM5242 definitely has balanced outputs.
If drMordor's TPA3251 is configured for BTL stereo, and there is no opamp front-end, its inputs can only be balanced.
Could someone help me out a little bit.
I believe I read somewhere that Class D can be sensitive towards an uneven load, such as not having a speaker connected on one output.
If I recall correctly it was a matter of it's BTL or not.
The reason I ask is because I'm going to be making driver measurements later to adjust the DSP.
I've got both the 3251 4 channel and 3255 2 channel boards from 3E.
I believe I read somewhere that Class D can be sensitive towards an uneven load, such as not having a speaker connected on one output.
If I recall correctly it was a matter of it's BTL or not.
The reason I ask is because I'm going to be making driver measurements later to adjust the DSP.
I've got both the 3251 4 channel and 3255 2 channel boards from 3E.
I think the combination of TPA325x and PCM5242 is an exercise in technical simplicity - to get digital sources to amplification with the minimum of devices and circuitry. The PCM5242 requires no I/V stage so its (differential) outputs can go straight to the unbuffered inputs of a TPA325x amp.
In fact this is virtually the same as TI's own equivalent product - the TAS3251, except that drMordor has possibly provided the PCM5242 with better clocks and better caps.
the TAS3251 is good alternative if you are looking for a digital input.
Checkout the TAS3251 running @home :
Upload Video Online - Free Video Hosting - Streamable