Chengzhi didn't mixup resistors, TI did in their datasheet 😛 If you remove the 100k you will have 20dB gain, no matter what other resistorvalue is. That will extend low frequencies....the ceramics might, but not certain about that, 3V bias plays funny tricks on x7r's, value could easily half because of bias, not unusual to lose even more. Search for voltage coefficent ceramics and you will see the 10% or 20% tolerance in datasheet is only about temperature change!!!!!!!! you could in fact have a 0.33uF capacitance using this 3.3uF capacitor with 3V bias applied.
To get that straight......just remove the 100k resistor and you have 20dB gain? The 5.6K is of no consequence - so the 'original' 20K does NOT have to be changed? What did TI think when asking for the 5.6K?
TPA3116/TPA3118/TPA3130 Gain Setting in Master and Slave
Posted by John Guy1 on Apr 04 2013 12:33 PMAn externally hosted image should be here but it was not working when we last tested it.
Prodigy100 points
Hi
Attempting to set 20dB gain on a TPA3118, and I believe there is an inconsistency in the datasheet.
- In the EC table, it says for GAIN(BTL) mode, for 20dB, connect R2 of 20k to ground, R1 open.
- In table 1, it says for GAIN(MSTR) mode for 20dB, connect R2 of 5.6k to grounb, R1 open.
For a gain of 20dB, what is the correct value for R2?
Thanks
John
Report Abuse
An externally hosted image should be here but it was not working when we last tested it.Posted by Don Dapkus on Apr 04 2013 22:28 PM
An externally hosted image should be here but it was not working when we last tested it.
Guru148335 points
Hi, John,
Nice catch, I thought we fixed that... It doesn't really matter the value of R2 in this case. The most important thing is if you use a divider, you don't draw too much current out of GVDD... Since in this case, no current is drawn from GVDD, you just want something to pull the pin to ground.
And, yes, it should be MSTR, not BTL...
I will get the d/s updated the next go-around.
-d2
-----
Don Dapkus
Audio Applications Engineering Manager
Dallas, TX USA
Thanks Irribeo for the helpful information! That makes it pretty easy to try the reduced gain method. Using the formula in the data sheet, the high-pass filter would be 2.66hz with 20db gain and 1uf input caps. Much better than the stock cutoff of 5.30hz.
Also, thanks for the information regarding ceramic caps and temperature change.
Ians505,
I like how you did your vertically integrated active bi-amp on your monitors. Mine are currently sitting in the open - glued to a piece of foam core board along with the miniDSP 🙂
Very professional looking box there - you must be working electrician to use boxes and conduits like that...
You can use big film caps - just make the legs stand up higher and get the pins to fit existing thru holes or SMT pads. Stabilize high mounted cap with some hot melt glue to some nearby structure. I use 3.3uF 250V Panasonic polyester film caps for about $1 ea from Digikey. Fat brown colored Chicklet candy shape ones...
I like how you did your vertically integrated active bi-amp on your monitors. Mine are currently sitting in the open - glued to a piece of foam core board along with the miniDSP 🙂
Very professional looking box there - you must be working electrician to use boxes and conduits like that...
You can use big film caps - just make the legs stand up higher and get the pins to fit existing thru holes or SMT pads. Stabilize high mounted cap with some hot melt glue to some nearby structure. I use 3.3uF 250V Panasonic polyester film caps for about $1 ea from Digikey. Fat brown colored Chicklet candy shape ones...
Okay, reduced the gain to 20dB and expected......well, an amp playing less loudly, but if you are using transformers on the input be prepared for a surprise: the amp suddenly plays a lot more three dimensional - one is able to look much deeper into recordings - highly recommended.
Thank you. If I remember correctly, you are using the Ydbz boards and like them? I ordered the chengzhi boards and Ydbz boards at the same time back in February. The chengzhi boards arrived in 1month... the Ydbz boards took 2.5months to get. I got anxious to listen to my setup and cased the chengzhi boards before comparing to Ydbz...still haven't found the time to try them. They may be better in stock form for my setup.
Thanks. No, just asked to imitate one when designing/fabricating equipment at work🙂
Thanks for your input on the polyester film caps. I like the hot glue idea.
Hahaha... you are right
Yes, I use the Ybdz ones and they are bone stock. After listening to them for 5 months still like them so have not had the big urge to mod them yet. I have all the parts to mod them just not enough activation energy because they sound so good already. I am driving some pretty efficient speakers with them (98 dB) so find that the 40 watts is plenty of power with a laptop brick.
Awesome finding. Man, I really want to try this input transformer business. You may have posted this already but can you please post a link of the transformer you use?
Finally got rid of R2 100K which set the gain setting in my YJ Blue/Black board to 20dB. One snip with a scissor was all it took. It reduced 30% of the hiss I had in the Beyma TPL-150 (~100dB 1W) tweeter and made the amp sound more balanced. Thanks!
Last edited:
Yes, and how did you connect the input wires to the board? Are you using input caps or did you remove them? I'm assuming you are using the transformer to convert single-ended source to differential, taking advantage of differential input stage of amp? In phase 2 of my modifications, I would like to use the balanced outputs from my minidsp for a higher source voltage ~4v to help drive the amp with gain set at 20dB - also taking advantage of the differential input stage. I'm working on figuring out how to do this.
Phase 3 would be to reconfigure the LC output filter and use ferrite beads since my speaker wire lengths are so short.
I have the same question. I believe it was Sharpi31 who posted about the Lundahl LL1540. What (admittedly little) searching I've done suggests those aren't readily available in the USA and are expensive (even used they're not cheap, at least not relative to a $20 PCB).
What other options are there? What specs are relevant when looking for a line-level input transformer for this application?
They are not even cheap in Sweden where they are manufactured. I used to live a couple of hours from the factory.
I wonder if a 560Q from Hammond works? It is 10k/40kohm impedance matching transformer with 30 Hz to 15kHz response. Epoxy potted PCB mounted part.
http://www.hammondmfg.com/pdf/5c0047-48.pdf
Have no idea what they cost but look like circa $15 part.
http://www.hammondmfg.com/pdf/5c0047-48.pdf
Have no idea what they cost but look like circa $15 part.
I've not tried them, as they are expensive to buy outside of the US, but have heard good things about Edcor transformers. These ones look suitable:
https://www.edcorusa.com/wsm15k-15k
https://www.edcorusa.com/xsm15k-15k
$10-12 each is worth a punt 🙂
https://www.edcorusa.com/wsm15k-15k
https://www.edcorusa.com/xsm15k-15k
$10-12 each is worth a punt 🙂
For the connection of those transformers: somewhere in this thread there is a drawing how to do that - according to sharpi's instructions 🙂
Generally using transformers means no input caps - i.e. the transformers' primary side is connected to the RCA jacks and the secondary to the board at the place where the input caps go towards the chip. It's really simple....
Generally using transformers means no input caps - i.e. the transformers' primary side is connected to the RCA jacks and the secondary to the board at the place where the input caps go towards the chip. It's really simple....
For the connection of those transformers: somewhere in this thread there is a drawing how to do that - according to sharpi's instructions 🙂
Generally using transformers means no input caps - i.e. the transformers' primary side is connected to the RCA jacks and the secondary to the board at the place where the input caps go towards the chip. It's really simple....
Am I right in thinking that the purpose of the transformer is just to eliminate DC on the input - no other function? So if there is no DC on the device providing the input to the amp (DAC etc.) there is no need for transformer of cap? For example there is a DC blocking cap on the DAC output or some other way to nullify DC.
Or am I missing something here?