I am not trying to address a problem. I am looking to tweak another YJ board and was looking to find an alternative capacitor to the TDK FK20X7R2E224K (x7r) in the bootstrap positions.
Thank you for the heads up with with capacitive losses due to bias.
Thank you for the heads up with with capacitive losses due to bias.
TI "advises" 220nF X5R (or better) in the datasheet..
Further:
Source: 3116 datasheet 7.3.6 page 16
http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf
The EVM is using "CAP SMD0603 CERM 0.22UFD 25V 10% X5R", PVCC is up to 26V.
So having an X7R in place, (16V or 25V) is more than enough, as X7R also doesn't derate that much than X5R in general (speaking for Murata parts).
Of course, there's an old "Preview" document, where X7Ris mentioned, but that dates before the official datasheet.
Source: http://www.ti.com/lit/ds/slos708/slos708.pdf
What's the sourche of 1.45xPVCC?
Saturnus once wrote:
With respect to Saturnus, this might be true for different bootstrap-configurations but, to me, not in this configuration. (There's no signal running through the bootstrap-cap.) The important part in this quote is, short distance, small current loops.
@hifix, irribeo:
To calculate the proper values, have a look into the equations shown:
http://www.silabs.com/Support Documents/TechnicalDocs/AN486.pdf
And here, section 3.1 following:
https://www.fairchildsemi.com/application-notes/AN/AN-6076.pdf
Last edited:
We don't need to calculate any value, TI gives us value for 3116, just like they do for 3251. If you haven't read TI's advice 1.45 times PSU voltage for bootstrap capacitor voltage rating, ask TI support. They also write in 3116 datasheet 1nF cap works for synch, so datasheet isn't always clear/correct. For 3251 if I remembered correctly max voltage bootstrap sees is ~60V, they pick 50V for EVM, I wouldn't and I think if you ask TI support they will also tell you there is a chance 50V isn't enough there. Is it likely I will ever need more than 50V bootstrap there, not likely, but you with all your sinewave testing might.
For YJblue, seriously, put gvdd cap on two gainresistor pcbholes closest to chip, you can move resistors back to gvdd cap pcbholes if you want better optics
http://www.diyaudio.com/forums/class-d/237086-tpa3116d2-amp-95.html#post3739580
(toprighthand side gainresistor is gnd, right side contact gvdd cap position is gnd plane)
http://www.diyaudio.com/forums/class-d/237086-tpa3116d2-amp-95.html#post3739580
(toprighthand side gainresistor is gnd, right side contact gvdd cap position is gnd plane)
Isn't it easier to provide a propper source rather than saying "ask TI"? 😕
The EVM is tested all the way up to full output power - sinewave, so those caps worked like they were implemented. That's why it is also called a "reference design". Would it make any sense to conceal informations to the customers and supply re reference design with suboptimal performance? Guess not.
Are you refering to "BST_X to GND" at the TPA3251D2 datasheet (page 5)? This is the maximum Pin-to-GND voltage (62.5V absolute maximum allowed), which is NOT the voltage across the BS-cap.
The actuall voltage across the BS-caps at TPA3251D2 is at ~12.5V. The reason why TI had choosen the GRM188R71H333KA61 (0603 0.033uF 50volts X7R 10%) here is, that theres no part available from Murata in this general purpose series at lower voltages at 33nF. GRM188R11E333KA01 as an alternative here, isn't sourceable.
🙂
But if you feel better to use a 250V part there with bigger physical dimensions, go for it. This is just my sight of things here, noone needs to adapt.
Peace.
Hi irribeo,
I "think" i get it. Would this happen to be it ?
Does this remove more noise from the gain setting input ?
The pitch of the capacitor looks abit different to the spacing of the resistor through holes.
Would any 1UJ63 capacitor do the job ?
I "think" i get it. Would this happen to be it ?
Does this remove more noise from the gain setting input ?
The pitch of the capacitor looks abit different to the spacing of the resistor through holes.
Would any 1UJ63 capacitor do the job ?
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Yes leadspacing is a little different and old cut resistors might be too big/leads too short for new position.
Bootstrap caps and gvdd decoupling cap are two sides of same thing, same problem on YJblue. YJblue is only (3116) amp I ever heard that eliminates recorded audience noises from playback (referred to Teldec Maxim Vengerov cd befor), parts of the recording are really gone, no traces left. Replacing the bootstraps YJblue brought back most these noises or hints at noises, still YJblue seemed lifeless compared to other 3116's, putting gvdd decoupling in position you pictured changed most of that (I also mentioned earlier that electrolytic Jamicons on YJblue were dry, these were replaced earlier).
I don't have multiple YJblue's, can only share what this single ampboard seems to do, all other 3116 I have, I bought myself and always multiple boards, I like to A-B. YJblue was given to me after listeningsession (3116 comparison) at a friends place, because owner decided to buy another 3116 and wanted YJblue erased from his memory🙂
Bootstrap caps and gvdd decoupling cap are two sides of same thing, same problem on YJblue. YJblue is only (3116) amp I ever heard that eliminates recorded audience noises from playback (referred to Teldec Maxim Vengerov cd befor), parts of the recording are really gone, no traces left. Replacing the bootstraps YJblue brought back most these noises or hints at noises, still YJblue seemed lifeless compared to other 3116's, putting gvdd decoupling in position you pictured changed most of that (I also mentioned earlier that electrolytic Jamicons on YJblue were dry, these were replaced earlier).
I don't have multiple YJblue's, can only share what this single ampboard seems to do, all other 3116 I have, I bought myself and always multiple boards, I like to A-B. YJblue was given to me after listeningsession (3116 comparison) at a friends place, because owner decided to buy another 3116 and wanted YJblue erased from his memory🙂
(bootstrapcaps)C104, C108, C111, Ceramic, 0.22 µF, ±10%, X7R 0.22 µF 0603 C115 Voltage rating must be > 1.45 × V PVDD (main outputfiltercaps)C109, C110, C116, Ceramic, 0.68 µF, ±10%, X7R 0.68 µF 0805 C117 Voltage rating must be > 1.8 × V PVDD
This is from tas5756m design/implementation considerations datasheet, you can find other places TI did this, as you can find other places where TI tells you 16V will be enough for bootstrapcap voltage rating. If you're ordering capacitors it isn't very interesting to know why or which TI engineer is correct, just order component with higher voltage rating and move on🙂
This is from tas5756m design/implementation considerations datasheet, you can find other places TI did this, as you can find other places where TI tells you 16V will be enough for bootstrapcap voltage rating. If you're ordering capacitors it isn't very interesting to know why or which TI engineer is correct, just order component with higher voltage rating and move on🙂
Thank you for the heads up irribeo. 🙂
I will relocate the resistors and the capacitor tomorrow. I am looking foward to hearing the new board config in action!
TPA wise, i have only heard the YJ Blue in relation to the SURE board and the SMSL SA-36a Pro (TPA3118). Overall i prefered the YJ Blue. I think i will have a listen to more music with a headphone comparison to see what its leaving out. I have purchased a YJ Blue (direct from the manufacturer, via their website) with questionable parts. The blue input caps had the same "print" on them, but the caps were significantly shorter and the inductors had a smaller footprint as well. I have contacted them regarding this but i have not heard back.
Have you heard the YJ Green board ?
I will relocate the resistors and the capacitor tomorrow. I am looking foward to hearing the new board config in action!
TPA wise, i have only heard the YJ Blue in relation to the SURE board and the SMSL SA-36a Pro (TPA3118). Overall i prefered the YJ Blue. I think i will have a listen to more music with a headphone comparison to see what its leaving out. I have purchased a YJ Blue (direct from the manufacturer, via their website) with questionable parts. The blue input caps had the same "print" on them, but the caps were significantly shorter and the inductors had a smaller footprint as well. I have contacted them regarding this but i have not heard back.
Have you heard the YJ Green board ?
Last edited:
If YJgreen is the clone/copy of AC1308 (also sold on ebay by audiobah) then yes I bought some from AC direct.
Ac1308 was in comparison, YJclone/copy of AC1308 didn't exist then. AC1308 was stock. I ordered some after comparison, https://controls.world.taobao.com/, ~$7 per ampboard ex shipping.
Ac1308 sounded clearer than YJred/blue. Bass was little weaker. Comparison was with preamp, which AC1308 needs the most, it has very low inputimpedance.
Power on/off pop is strongest selling point AC1308. YJcopy differs from AC1308 in decouplingvalues and outputfiltervalues, and ofcourse, YJ sourced all parts for clones.
Ac1308 sounded clearer than YJred/blue. Bass was little weaker. Comparison was with preamp, which AC1308 needs the most, it has very low inputimpedance.
Power on/off pop is strongest selling point AC1308. YJcopy differs from AC1308 in decouplingvalues and outputfiltervalues, and ofcourse, YJ sourced all parts for clones.
Hey thanks for that! I've got some inductors on the way... and after another couple of builds and rounds of experimentation... i might give the green one a go.
Hello
For a project at school I need to make a class D amp. I've chosen the tpa3130 to use as base. (I found later that the 3116 has more output, but the chip was already ordered).
Anyway, this is my first attempt at building such an amp and I'm going to use it for a portable 2way system with a Fane Sovereign 8-225 8Ohm woofer and an eminence apt80 8Ohm tweeter with a crossover frequency of 3.5kHz. This all is going to get powered by a 12V 7Ah (or 10Ah, not sure yet) battery.
I've made my eagle schematic and I'd love if someone here can give his opinion on it.
Zippyshare.com - schema, snubbermod.sch
Some more explanation:
-first there is a 4th order active Butterworth highpass filter @70kHz to cut the lower frequencies. this is to limit the lower frequencies and save some battery
-after the highpass, i've added a bass/treble/volume control circuit. Both the highpass and this circuit use the same opamp chip (3 needed, 4 present on the chip)
than there is the 3130 itself:
- the am inputs are connected to jumpers so I can change the config if needed
- i've used the the giancarlo circuit for the faultz, sdl and mute pins
- i've used the bootstrap snubber mod, i'm just not sure if i can connect positive and negative pin of each output before or after the LC filter. atm I need 2 inducturs instead of 4.
- the LC filter is tuned at 28kHz with an impendance of 1.7Ohm
- I'm using this chip in mono, and pbtl mode
The most important components (ordercodes are from farnell):
-for DC power cap:
1x
Electrolytic Capacitor, FM Series, 2200 µF, ± 20%, 25 V, 12.5 mm, Radial Leaded
Ordercode: 1219472
2x
Electrolytic Capacitor, FM Series, 470 µF, ± 20%, 35 V, 10 mm, Radial Leaded
Ordercode: 1219475
-for input dc de-coupling caps:
2x
capacitor 10µF, >10V, smd, >x5r
-bootstrap caps:
4x
0.22 µF, ± 10%, X7R, 250 V, 1210
Ordercode: 2112911
or
Ordercode: 1828959
-for snubber:
2x
SMD Multilayer Ceramic Capacitor, CGA Series, 3.3 µF, ± 10%, X7R, 50 V, 1210
Ordercode: 2211157
-for LC filter:
2x
Surface Mount Power Inductor, AEC-Q200 SRR1280A Series, 10 µH, ± 20%, Shielded, 0.0195 ohm, 6 A
Ordercode: 2374152
Thanks for your time 🙂
For a project at school I need to make a class D amp. I've chosen the tpa3130 to use as base. (I found later that the 3116 has more output, but the chip was already ordered).
Anyway, this is my first attempt at building such an amp and I'm going to use it for a portable 2way system with a Fane Sovereign 8-225 8Ohm woofer and an eminence apt80 8Ohm tweeter with a crossover frequency of 3.5kHz. This all is going to get powered by a 12V 7Ah (or 10Ah, not sure yet) battery.
I've made my eagle schematic and I'd love if someone here can give his opinion on it.
Zippyshare.com - schema, snubbermod.sch
Some more explanation:
-first there is a 4th order active Butterworth highpass filter @70kHz to cut the lower frequencies. this is to limit the lower frequencies and save some battery
-after the highpass, i've added a bass/treble/volume control circuit. Both the highpass and this circuit use the same opamp chip (3 needed, 4 present on the chip)
than there is the 3130 itself:
- the am inputs are connected to jumpers so I can change the config if needed
- i've used the the giancarlo circuit for the faultz, sdl and mute pins
- i've used the bootstrap snubber mod, i'm just not sure if i can connect positive and negative pin of each output before or after the LC filter. atm I need 2 inducturs instead of 4.
- the LC filter is tuned at 28kHz with an impendance of 1.7Ohm
- I'm using this chip in mono, and pbtl mode
The most important components (ordercodes are from farnell):
-for DC power cap:
1x
Electrolytic Capacitor, FM Series, 2200 µF, ± 20%, 25 V, 12.5 mm, Radial Leaded
Ordercode: 1219472
2x
Electrolytic Capacitor, FM Series, 470 µF, ± 20%, 35 V, 10 mm, Radial Leaded
Ordercode: 1219475
-for input dc de-coupling caps:
2x
capacitor 10µF, >10V, smd, >x5r
-bootstrap caps:
4x
0.22 µF, ± 10%, X7R, 250 V, 1210
Ordercode: 2112911
or
Ordercode: 1828959
-for snubber:
2x
SMD Multilayer Ceramic Capacitor, CGA Series, 3.3 µF, ± 10%, X7R, 50 V, 1210
Ordercode: 2211157
-for LC filter:
2x
Surface Mount Power Inductor, AEC-Q200 SRR1280A Series, 10 µH, ± 20%, Shielded, 0.0195 ohm, 6 A
Ordercode: 2374152
Thanks for your time 🙂
That's what I wouldn't call proper engineering. 🙄
But hey, why think yourself if others do, careless if they're right or wrong. The religious way of "life" - praise the lord, oh yeah.
Peace.
Since you also do not care enough to question your 16V, welcome to the club!
Than you should read my postings again. Beside you, I measured the questioned voltages and also picked my parts regarding the needs from calculation and measurements.
Would guess, you actually don't know how the bootstrapping really works, but, also don't care.
I'm tired to discuss aspects of proper implementation, because no one really cares but just thrust their ears as the ultimate "measuring equipment", with super low distortion and no mind bias..
Keep posting your crude suggestions, I'm done with it.
I'm sorry, but I can't keep you seriously.
🙂
Would guess, you actually don't know how the bootstrapping really works, but, also don't care.
I'm tired to discuss aspects of proper implementation, because no one really cares but just thrust their ears as the ultimate "measuring equipment", with super low distortion and no mind bias..
Keep posting your crude suggestions, I'm done with it.
I'm sorry, but I can't keep you seriously.
🙂
Yes 20uF you calculated per channel also is more than enough, so a 7mOhm polymer should be easy to implement and prove you wrong, however you are not interested in proving your own calculations wrong.
Without measurements and trying yourself, you also endorsed AVCC/PVCC split using an RC on AVCC, did you by any chance ever tried such an ampboard by now? Care to share your measurements or thoughts, or do you fear to show you were wrong again?
Without measurements and trying yourself, you also endorsed AVCC/PVCC split using an RC on AVCC, did you by any chance ever tried such an ampboard by now? Care to share your measurements or thoughts, or do you fear to show you were wrong again?
I haven't seen any measurements of you here in DiyAudio? So if i'm wrong in some way, there's no problem to correct me. My measurements are posted here and on my website, where's yours?
But yeah, i should try to listen more, measure less. -.-
But yeah, i should try to listen more, measure less. -.-
Or one can use common sense if you want. Figure 3 shows a 7 V zener diode paralell to bootstrap capacitor, so there is no chance for high voltage. It must be a copy - paste error.
http://www.ti.com/lit/ds/symlink/tas5756m.pdf
doctormord!
Congrats for your webpage, nice work!