Wow, that did really not answered my question. 😀
Hmm. :/
What is this "much"? Is it "sonic signature" (suppression)? Maybe i don't get it because of the language barrier?
Hmm. :/
I bought this 2 TPA3118 PBTL.
TPA3118 PBTL 60W Mono Digital Audio Power Amplifier Board Module DC 8 24V T8L9 | eBay
Based on the datasheet, if I want to parallel 2 PBTLs to create stereo amp, I havet to connect both synch pin (PIN 16) with 10K resistor and create ground connection with added 1nF capacitor.
You can see my diagram plan here:
https://drive.google.com/open?id=0B_U4OLmXFhUjOHRWNmVPVmtVWFU
Is this correct thing to do?
Thanks for your help.
TPA3118 PBTL 60W Mono Digital Audio Power Amplifier Board Module DC 8 24V T8L9 | eBay
Based on the datasheet, if I want to parallel 2 PBTLs to create stereo amp, I havet to connect both synch pin (PIN 16) with 10K resistor and create ground connection with added 1nF capacitor.
You can see my diagram plan here:
https://drive.google.com/open?id=0B_U4OLmXFhUjOHRWNmVPVmtVWFU
Is this correct thing to do?
Thanks for your help.
Last edited:
No those datasheet synch values do not work and you need to put your R amp in slavemode (gainsetting). Simulate datasheet values as a filter and you'll see why. Use EVM values for example those do work. (4k7, 47pF I think, but I am half sleeping 🙂
I just treat them as separate amps, one for left channel, one for right channel.
Also, in the event that you have noise issues, try experimenting with the gain by changing R27(R1 on the table) and R28 (R2 on the table). 26dB and 20dB works well.
Here's a couple of problems:
All else except bass and bari is so much louder that it is difficult to hear the bass and bari, especially the bari.
The louder higher frequencies have distortion so that they're not as useful as needed.
Here's some possibilities:
Input filters, output filters, voltage amp stability, and power decoupling are all possible locales for the problem; so, those are all possible locales to repair.
However,
Achieving lowest distortion results requires situating the repair as close as possible to the problem. Unfortunately, I haven't found the locale(s) of the problem. Where is it?
Are you sure the problem is just not your speakers ? I think the amp will probably be fairly flat response except for falloff of bass by 1dB or so below 100Hz. The excessive midrange you talk about sounds more like the speaker doesn't have its baffle step compensation set right. That is a 3dB to 6dB effect of excessive mids and highs on a speaker with no BSC circuit. A frequency response measurement would be helpful.
If you have -1dB at 100Hz I would also look at your speaker for that.
8 minutes into video you get a different impression of 1uF input capacitors combined with 26dB and slightly peaking (in audioband, can get ugly outside audioband, see shutdown)outputfilter into 8 ohm resistor!!
https://www.youtube.com/watch?v=s3PTPEfmRRw
8 minutes into video you get a different impression of 1uF input capacitors combined with 26dB and slightly peaking (in audioband, can get ugly outside audioband, see shutdown)outputfilter into 8 ohm resistor!!
https://www.youtube.com/watch?v=s3PTPEfmRRw
I meant starting below 100Hz - my 3118 is about -0.8dB at 40Hz relative to my class AB Dx but that may be input cap being only 1uF.
Thanks for link to Danzz Blue/Black amp. That amp used to be the "go to" 3116 when it came out. Measures pretty well for not having the best layout. I would love to have one of those HP distortion analyzers that the guy used in the video.
1uF chinese ceramic could be identical to 1uF chinese Epcos filmcap, but could be very different too 🙂
Your PSU could be different too. You use same Amazon for the Dx ?
Your PSU could be different too. You use same Amazon for the Dx ?
Last edited:
Almost the same 19.5v 4.6a but dual rail for Dx. This is temporary until I build 250va transformer based dual rail 32v.
That's the thing with class AB amps is the expensive big iron dual rail supplies. The 4700uF caps just arrived so can start making the ps board finally.
That's the thing with class AB amps is the expensive big iron dual rail supplies. The 4700uF caps just arrived so can start making the ps board finally.
While you're at it, put in a "ripple eater" too. The PS can be the difference between a "good" sounding amp and a "great" sounding one.
I am actually making the Dx power supply which looks like it has a CLRC filter in it.
An externally hosted image should be here but it was not working when we last tested it.
Applying an input load of 22k seemed to help the TPA3118 a little bit.
I actually, have 470R series then 22k load, so it is very slightly quieter as well.
I actually, have 470R series then 22k load, so it is very slightly quieter as well.
Not excessive Midrange sonic signature like a normal amp, but somewhat higher pitched and not less loud, but merely less observable because higher than ear sensitively peak frequencies.
Again, the bass of this amp, does not lack, but there seems to be a problem that it does a few things artificially louder.
Thanks so much for the reminder! I had meant to use this amp with a specific speaker, not the huge old test speaker in the workshop. Well, I connected the intended speaker, and it does have a BSC circuit onboard, mainly because it is half as big.
That was even worse!
BRB. Doing this in live time, soldering irons hot...
Next, I tried the TDA8932 in BTL mode, as shipped. Sure, it produces higher resolution in SE mode, but I tried it in BTL for comparison.
(this sample has the slightly better decouplers as shipped recently, and the 47u added for standby timer.)
I was SO surprised!
Oh my word!
Problem solved.
Is it faultless? You'll have to wait till I have recovered from the surprise of how well it did. The only news I have to report is that I am not currently in need of any TPA3116/TPA3118 chips, not now and not soon. I am grateful that it was informative. The efficiency was different; however, the output was also different. Fair enough, because effectiveness is mandatory. Signing off... This not mine!!!
Ok, so saying that stock arrangement is best? Isn't it PBTL mono and not BTL mono? I think it is mislabeled as Irribeo pointed out earlier.
Can't decipher cryptic language.....Danielwritesbac, what are you trying to say ? I just reread some posts and they seem to have gone through Google translate 😉
Last edited:
I was willing to accept the lesser efficiency of the TDA8932 in trade for the output that I liked better. And it is also relevant that, the extent/amplitude of difference was a surprise. I make less sense when surprised.
It is possible that it would depend on your application. Stock arrangement won't be for best; however, it is merely more noteworthy to observe where they have practiced cost cutting versus where they chose to spend more.
Just now, a difference in decoupling cap quality, the lesser "K4 100 355" versus the better "100 35V VT" and I wonder about the cost difference. This does NOT sound the same. Fortunately for me, the latter blew away its competitors straightaway, so I got my project done already.
It didn't involve using a TPA3118/TPA3116. They tried well and they weren't bad, but I wasn't able to use them.