You are saying 6 resistors, 2 caps and two voltage sources are a good model of a TPA3255 class d amp? Or are there more things inside the box? Does TI have a model in Tina?
Because there was no model in Tina at the time, I created this model of my own.
Certainly It is very basic but with the single focus on loop stability it simulates surprisingly well.
And I did not like Tina that much.
Certainly It is very basic but with the single focus on loop stability it simulates surprisingly well.
And I did not like Tina that much.
Last edited:
Sure... I just had the EVM boards in mind. The TPA3255EVM in particular has nearly 100 uV output noise.The TPA3255 without PFFB and a good PSU had very low intrinsic noise. One of the quietest amps I have measured. I can press my ear to the speaker cone and cannot tell the amp is on. PFFB reduces the THD, but the background noise is already low.
This is the measured noise floor on my amp without PFFB using a MicroAudio SMPS650-SO PSU.
A lot of the noise is dependent on the quality of the PSU. The Class D switching noise still leaks through but, fortunately, 430kHz is well outside the bandwidth of speaker response or human hearing.
This is great that you developed such a simple model. How do you check to verify that the model works well?Because there was no model in Tina at the time, I created this model of my own.
Certainly It is very basic but with the single focus on loop stability it simulates surprisingly well.
And I did not like Tina that much.
If you're planning on using Faston tabs for output, how about spreading the L & R outputs apart a bit more and adding a central mounting hole to reduce PCB flex?Solve, fully agree that in-room freq response is so much more important than THD and that EQ, as well as room treatment and controlled dispersion is the key, but here we are talking about amps, and when designing a new amp it makes sense to try to obtain the best results also regarding THD and Noise 😉
Think I'm close to sending to JLCPCB for my first TPA3255 prototypes (10x10cm 4 layer) 😉
Let's see it it works. Seems to be room enough to implement a ADAU1701 and hopefully a PCM1808 ADC . The 3 opamps on the board can probably be ommited, they form the SE to BAL and Vdd/2 circuit. ADAU1701 can deliver fully balanced output.
But one thing at a time 😉
View attachment 1402815 View attachment 1402816
I have a TI tpa3255 evm board. On Texas Instruments website it is mentioned in the manual that the first edition needs 2 Power Supply. Is this the case? Attaching a clipping from the manual.
'
10 Power Supply Recommendations
10.1 Power Supplies
The TPA3255-Q1 device requires two external power supplies for proper operation. A high-voltage supply called
PVDD is required to power the output stage of the speaker amplifier and its associated circuitry. Additionally, one
mid-voltage power supply for GVDD_X and VDD is required to power the gate-drive and other internal digital and
analog portions of the device. The allowable voltage range for both the PVDD and the GVDD_X/VDD supplies
are listed in the Recommended Operating Conditions table. Ensure both the PVDD and the GVDD_X/VDD
supplies can deliver more current than listed in the Electrical Characteristics table.'
'
10 Power Supply Recommendations
10.1 Power Supplies
The TPA3255-Q1 device requires two external power supplies for proper operation. A high-voltage supply called
PVDD is required to power the output stage of the speaker amplifier and its associated circuitry. Additionally, one
mid-voltage power supply for GVDD_X and VDD is required to power the gate-drive and other internal digital and
analog portions of the device. The allowable voltage range for both the PVDD and the GVDD_X/VDD supplies
are listed in the Recommended Operating Conditions table. Ensure both the PVDD and the GVDD_X/VDD
supplies can deliver more current than listed in the Electrical Characteristics table.'
The TPA3255EVM does not require connecting two separate power supplies. Just the single PVDD is enough. I'm pretty sure the EVM has an onboard power supply to provide any other voltages needed.
Just to be sure, review the start-up guide for your device and see what connections they ask for.
Just to be sure, review the start-up guide for your device and see what connections they ask for.
I measured real life test circuits.How do you check to verify that the model works well?
The ones I build since.
Err...back to basic question, is PFFB worth it because it will reduce gain? because looks like the bucks bunny's PFFB just like TI reference after all
4 layer with 1 or 2oz copper for the outside layers? If only 1oz be mindful about max currents.Solve, fully agree that in-room freq response is so much more important than THD and that EQ, as well as room treatment and controlled dispersion is the key, but here we are talking about amps, and when designing a new amp it makes sense to try to obtain the best results also regarding THD and Noise 😉
Think I'm close to sending to JLCPCB for my first TPA3255 prototypes (10x10cm 4 layer) 😉
Let's see it it works. Seems to be room enough to implement a ADAU1701 and hopefully a PCM1808 ADC . The 3 opamps on the board can probably be ommited, they form the SE to BAL and Vdd/2 circuit. ADAU1701 can deliver fully balanced output.
But one thing at a time 😉
View attachment 1402815 View attachment 1402816
Yes and no. Remember we are not talking continuous currents at all. Have had absolutely no problems with same setup for +300W IRS2092 based amps 😉4 layer with 1 or 2oz copper for the outside layers? If only 1oz be mindful about max currents.
Will use one TPA3255 per side in my quad-amped system for the lower registers w/o using any input opamps. B_input and B-output should be permanently deactivated. B_input can be grounded either with cap or not. About the output it would be conveniant if the filter components could be dumped altogether, but I suspect they need to be kept togheter with a resistor termination after an output cap. If keeping the filter is a must, how small could the output cap be and how large the terminating resistor?
Unused Output B does not require the filter.
I would skip entirely the discrete input buffers, which by the way are more complex and inferior compared to simple op-amp buffer stage.
I would skip entirely the discrete input buffers, which by the way are more complex and inferior compared to simple op-amp buffer stage.
- Home
- Amplifiers
- Class D
- TPA3255 - all about DIY, Discussion, Design etc