Volt+ tpa3118 testing data
May i ask for THD vs. Pout (continuous plot for 20Hz/1k/6k) just for completeness? (4/8R)
Just like datasheet shows.
The board shows mainly H3/H5 (deadtime residuals)
May i ask for THD vs. Pout (continuous plot for 20Hz/1k/6k) just for completeness? (4/8R)
Just like datasheet shows.
The board shows mainly H3/H5 (deadtime residuals)
Last edited:
Identical frequency respons for 8 ohm load and earlier 4 ohm load.
Yeap. I looked at datasheet from tpa3116 at 1W vs 10W (fig 10 page 9) and difference of the THD+N is extremely low .
May i ask for THD vs. Pout (continuous plot for 20Hz/1k/6k) just for completeness? (4/8R)
Just like datasheet shows.
The board shows mainly H3/H5 (deadtime residuals)
I asked the team , they will do it.
Your 1 watt into 4 ohm and 1 watt into 8 ohm have same frequency respons, pages 8 of 11 for both.Yeap. I looked at datasheet from tpa3116 at 1W vs 10W (fig 10 page 9) and difference of the THD+N is extremely low .
@cgdsgames : Thanks for the work.
Excellent results @400Hz, confirm an excellent implementation 🙂
AP reports are not easy to read for non educated eyes. You should put your data together in a more readable form and publish on your website. You can be proud of them.
As diyers, we'd like to understand and could try new designs and mods endlessly just to find out about the differences...
Excellent results @400Hz, confirm an excellent implementation 🙂
AP reports are not easy to read for non educated eyes. You should put your data together in a more readable form and publish on your website. You can be proud of them.
As diyers, we'd like to understand and could try new designs and mods endlessly just to find out about the differences...
@cgdsgames : Thanks for the work.
Excellent results @400Hz, confirm an excellent implementation 🙂
AP reports are not easy to read for non educated eyes. You should put your data together in a more readable form and publish on your website. You can be proud of them.
As diyers, we'd like to understand and could try new designs and mods endlessly just to find out about the differences...
Alkasar , we appreciate every kind word . We worked hard on the Volt+ and its our great team from India that did the ground work with a bit of help from us.
Every comment like this is sent to them. Thx you.
We appreciate this community and will share our own "secret sauces" . We are all learning, we are all diyers.
A bottle of St-Emillion thats what waiting for me tonight 🙂
@cdsgames, is the RPi's 3.3V still not passed through?
One specific application where it would be handy to have the 3.3V pass through is for use with the Soekris dam1021. The Soekris dam1021 requires 3.3V input along with the i2s because it has an isolator that needs to be powered. So a "default" RPi can provide that natively, but it looks like using the Allo Kali would require a separate 3.3V supply.
(Granted, the Soekris has its own reclocker after the isolation, so in theory the Kali would be redundant anyway. Despite that, I've read subjective reports of improvements when using better I2S devices.)
I also found Soekris sound quality to be dependent on I2S signal quality. The better the I2S signal the better sound you get out of Soekris. Seems like soekris reclocking is inefficient.
I am also waiting for Kali version 2 with 3.3V pass through to power Soekris isolator.
Are the measurements posted here:
http://www.diyaudio.com/forums/vendors-bazaar/305243-volt-tpa3118-testing-data-5.html#post5028480
done at 19V or a higher voltage?
http://www.diyaudio.com/forums/vendors-bazaar/305243-volt-tpa3118-testing-data-5.html#post5028480
done at 19V or a higher voltage?
Hmm, Poutmax at 8R:
(19V*0.707)^2/8R = 22.5W
From the plots reading for 1% THD:
20Hz: 28W
1kHz: 26W
6kHz: 26W
Ok.
(19V*0.707)^2/8R = 22.5W
From the plots reading for 1% THD:
20Hz: 28W
1kHz: 26W
6kHz: 26W
Ok.
Hmm, Poutmax at 8R:
(19V*0.707)^2/8R = 22.5W
From the plots reading for 1% THD:
20Hz: 28W
1kHz: 26W
6kHz: 26W
Ok.
You are very precise by looking at that graph. 26W 🙂
I get slightly lower . But yes seems that they used higher voltage. I know the smps has a +- (10-15%) but frankly we work in different offices and a while back they told me we use 19v. Seems they used 5 to 10% more.
Yes, it's more like 25.25W. 😉
I did some more measurements on the Sanwu with different inductors, getting it quite close for loads >=8R with the right inductor to your pace. Will post that stuff in the other tread.
I did some more measurements on the Sanwu with different inductors, getting it quite close for loads >=8R with the right inductor to your pace. Will post that stuff in the other tread.
Close..but no cigar ?
Also I assume you are also using normal SMPS as power source to reach close like we do? Or is it a lab PSU/linear ..
Just to compare apples to apples.
Also I assume you are also using normal SMPS as power source to reach close like we do? Or is it a lab PSU/linear ..
Just to compare apples to apples.
Stepped attenuator for Volt+ released 🙂
10k , 0.1% thin film resistors , 25ppm
https://allo.com/sparky/stepped-attenuator.html
Next week we will test the Volt+ using the new attenuator. All test on Volt+ were done with POTs . Lets see if we get any improvement ..
10k , 0.1% thin film resistors , 25ppm
https://allo.com/sparky/stepped-attenuator.html
Next week we will test the Volt+ using the new attenuator. All test on Volt+ were done with POTs . Lets see if we get any improvement ..
Close..but no cigar ?
Also I assume you are also using normal SMPS as power source to reach close like we do? Or is it a lab PSU/linear ..
Just to compare apples to apples.
Close yes - because i don't have an AP to work with and i can't specify the uncertainty of measurement. I also don't know what integration time and window function the AP is using for the FFT. Beside this, i don't have exact 8R load resistors.
(Lastly, i compared at 1kHz only)
The results are reasonable as a colleague did measured his TPA3118 implementation in the same manner with "close" results.
Power supply is a Korad KA3005D. (Linear)
Last edited:
🙂Our testing is done with SMPS. I think its slightly unfair to use a linear power supply and compare the data versus one using SMPS..generally speaking
Second , you are comparing BTL with PBTL boards. You know that PBTL have better specs.
Even with that, from 0.1W to 22W our boards show better THD+N.
So we will take it.
Second , you are comparing BTL with PBTL boards. You know that PBTL have better specs.
Even with that, from 0.1W to 22W our boards show better THD+N.
So we will take it.
- Home
- Vendor's Bazaar
- New FIFO buffer for RPI/SBCs