Sorta...
But yes.
No signal passes through them in single ended operation. However they can be used for balanced operation if they are not just grounded off.
rhing, I have a question about your X2 cap. My smsl tpa3118 has no turn on noises, but it does have a minor turn off pop. Did your other mods exaggerate the on/off transients?
Also, I have not yet opened my smsl enclosure for close inspection of the components. What is the lead spacing and largest possible body dimension of the 470uF/25V caps? I'll probably add the FMs to my next order.
BK
Only if Forte's still have stock crossover/tweeter. Higher frequency impedance is what will differ stock/modified Forte's.
Ceramics drop when DC is applied. On input there is DC from chip. The 3.3uF won't ever be 3.3uF there, since they are classII ceramics. Unless you measured them while amp was playing LOL Forte's being 100dB LOLOL, when measured with big resistor on speakeroutput is possible
Without load you should not turn amp on, if you try, it will probably sound different after, filter could burn
PBTL is selected if SMSL inputcapacitors both positive and negatve righthandside looking from front, if both are connected to ground befor the capacitor, so if chipinput goes straight to ground.
Ceramics drop when DC is applied. On input there is DC from chip. The 3.3uF won't ever be 3.3uF there, since they are classII ceramics. Unless you measured them while amp was playing LOL Forte's being 100dB LOLOL, when measured with big resistor on speakeroutput is possible
Without load you should not turn amp on, if you try, it will probably sound different after, filter could burn PBTL is selected if SMSL inputcapacitors both positive and negatve righthandside looking from front, if both are connected to ground befor the capacitor, so if chipinput goes straight to ground.
I did a couple of times, no matter the cansize/voltage on lower efficiency speakers, when amp has to produce a little power, high frequencies suffer real bad with all Panasonics. Not with ZA hybrids and probably not with Oscons, both being Sanyo's I guess, not Panasonics.
I made a SPICE script to sim the transfer function of the LC filter with a driver as reactive load. The 680nF and 10uH looks pretty good. More details here:
http://www.diyaudio.com/forums/multi-way/258487-why-crossover-1-4khz-range-76.html#post4027841
I was referring to the eight 25V/470uF used on the smsl TPA3118 board. Are you referring to the same?
BK
I was referring to what Panasonic FM/FR and what others told me Panasonic FC too, did to high frequency detail on Hiampmini3116, Sure3110 (little) and audiobah3116 pcb for me. Every time cymbals become noise, lack detail, completely flat and are generally attenuated a little compared to other electrolytics tried. (Hybrid Panasonic ZA also don't have this flaw on same ampboards).
Using 8x 470uF/25V Panasonic FM caps in the SMSL SA-36A Pro does not add harshness, distortion or loss of articulation to the sound quality. However, for the electrolytic caps closest to the TPA31xx IC, I would not recommend Panasonic FM caps, because as Irribeo stated, the sound quality can degrade with use of the Panasonic FM. I cannot say the same for Panasonic FC or Panasonic FR cap, because I hadn't tried them. The caps closest to the TPA31xx IC are where I used 330uF/25V Panasonic SEPF OSCON caps with positive results.
Do you happen to know the lead spacing and body dimensions of the 8x 470uF/25V that fit the SMSL SA-36A Pro board?
I already have 2x Panasonic SEPF OSCONs on hand for the spot near the chip.
BK
KJA 2013- great looking little amp
Thanks Prezden! I like the way it looks too. I've even had a request to make another one just like it.
Panasonic EEU-FM1E471, 470uF/25V, 5mm LS, 10mm dia. x 16mm length. You can get these from Mouser or Digikey.
As for the minor turn-off pop, this diminished once I installed the 8x Panasonic FM caps and the Epcos X2 safety cap. I installed these at the same time, so I cannot tell you which contributed to reducing the pop.
Another thing to note is the solder on the SMSL amp board barely reflows with heating using my 40W Metcal soldering iron. Be careful not to rip out any thru-hole eyelets when removing the old caps and solder. I had to flow some Cardas quad-eutectic solder onto the existing solder just to get my solder sucking tool to pick up anything. Even then, it took some time and patience to get all the thru-holes cleared to install the Panasonic FM caps. I even had to resort to using a stainless steel needle to ream open some of the thru-holes while heating them.
Needless to say, I retouched almost every solder joint on that amp in case there were any cold solder joints.
Great info - thank you!
I have a Haako desoldering gun and a 130W JBC solder station, so hopefully those will ease the task. Sounds like tricky lead free solder?
BK
Ya, lead free is mostly required for international shipping these days. Sometimes it isn't lead free, but anymore expect it.
Lead free can be made to flow with special stuff you add. Personally I just add a gob of leaded solder and problem is over. But these days I wish solder wick wasn't a God damn fortune.....
Lead free can be made to flow with special stuff you add. Personally I just add a gob of leaded solder and problem is over. But these days I wish solder wick wasn't a God damn fortune.....
This whole lead free RoHS thing costs billions in losses and hassle. Did you know that pure tin soldered components grow whiskers spontaneously and they short out perfectly soldered electronics? This is really bad for $400M spacecraft meant to last 15yrs in orbit. Not to mention it is a pain to solder/desolder.
Yep, and not all solder wick is created equal. Cheap stuff is just that and causes more problems than it solves. I use Gootwick these days especially around SMD.
There are some exception. I'm not sure how or why you can get them, but I know I've received a mix of lead free and leaded well after the fact.
Still, yes, I wish there was a ROFL smiley face.
The only people that really have a stake in the lead free is those who work on electronics in an environment where they don't want to provide ventilation or anything at all. It's not really an environmental contaminate I'm aware of except when 3rd world counterfeit jerks wash the PCB boards in rivers....
Luckily for us all we can buy leaded solder and continue life...
What can be frustrating is no-flow Lead-free solder on a board with poorly constructed boards where thru-hole eyelets and traces easily pull-off the board when heating and modding. Sometimes, you only have one chance to get it right, or you have to scrape away varnish and reconstruct a trace with wire and solder.
Sure TPA3116 suitable for 8 ohms load
I'm planning to order sure TPA3116 board :
Sure Electronics' webstore 2 x 50 Watt 4 Ohm Class D Audio Amplifier Board ? TPA3116
Is the board suitable for 8 ohms load, their site annonce TPA3116 delivers 2 x 50W into 4Ω loads. No informations concerning 8 ohms load.
For which impedance is the output filter designed ?
I'm planning to order sure TPA3116 board :
Sure Electronics' webstore 2 x 50 Watt 4 Ohm Class D Audio Amplifier Board ? TPA3116
Is the board suitable for 8 ohms load, their site annonce TPA3116 delivers 2 x 50W into 4Ω loads. No informations concerning 8 ohms load.
For which impedance is the output filter designed ?
From KJA's excellent photo of the 3116 we can see that the inductor is 10uH so I assume it is designed for 4ohms load. However, my experience and spice sims show there will be hardly any discernible difference in SQ when used with 8ohm speakers. It has to do with maybe 1dB difference in 18khz and above range. Also, I think using 8ohm speakers with 10uH has more linear phase response according to spice model.
For Tina simulations resistance can be replaced by datasheet Le/Re values speaker like Trevor Marshall did
Yes, I did that and it actually improves things. The LC filter was designed with a speaker's reactive load in mind and the gain and phase both improved when a more realistic speaker is modeled with TS params. This model includes more than Re and Le also has Cms and Rms.
The gain and phase error with a typical driver are now essentially insignificant within the 20kHz bandwidth.