The 3116 has the option of running two different modulation methods.
You might have to use a differential probe to get the correct output.
Review the data sheet and find out which mod scheme they are using.
http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf
page 21
You might have to use a differential probe to get the correct output.
Review the data sheet and find out which mod scheme they are using.
http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf
page 21
I guess that you need to make a balanced to unbalanced convertor, using a input attenuator and a well balance HV diff opamp. or a transformer or a differential probe or a isolation transformer for your scope power. To bad TI does not show instrumentation, but most probably AP equipment.
See attached, look how easy it is
You do not need the jfet input amp etc, it be made much simpler using one quad opamp (OPA1654 or equiv). Ckt. Power supply will have to float from earth ground, but a simple linear dual regulator (LM317/337) powered by a transformer with a centre tapped secondary, can do this. I saw one in the Elektor mag using NE5534's. If you want I can look it up for you.Paralleling the outputs allows you to share the current, so that you can drive lower impedance's such as 2 ohms.
So it is not necessary to parallel drivers for 4 ohm or larger. Paralleling drivers will reduce chip power dissipation, so you can use a smaller heatsink. It will increase reliability since that is proportional to die temperature.
Bottomline, you are not going to make anymore power unless you reduce the load impedance.
Okay enough from me. but thanks guys for sharing your experiences, I really enjoyed them.
Last edited:
Thanks a lot.
I'm building a subwoofer for a mobile soundsystem.
It is going to be an basshybrid and for the construction of my choice i only know two ulb-drivers that fit the required tsp.
One is an 8-ohmer (and little more inexpensive), the other an 4-ohmer (with a little bit better fitting numbers in VAS and Qts).
If i understand everything right i think i now have an additional reason to chose the 4-ohm-driver.
As long as i cannot really trust the numbers given for those ulb-drivers, taking care of the power efficiency may be the only valid argument since it is an accu-driven system.
I finally lashed my YJ 3116 board up. Running on 24v to a pair of EL70 full range speakers. These were being being powered by a Helder Hifi TA2020 MK2 until now. The sound is at least on par. I will have to take a closer listen tomorrow. The little 3-pin input is a pain. I ended up stripping the female crimped wire out of a similar plug and shrink tubing them.
Kyle
Kyle
Yes, I have and it sounds excellent. I need to get a new SLA because the one I have doesn't hold a charge for long.
I didn't get home until late last night. I do not have the full cased unit, just the populated board. The board dimensions are not even numbers in either metric, or inch units. My board measures 4.275 inches wide and 6.315 inches deep. There is a cutout across part of the back of the board to make room for the speaker connectors. Going by the pictures, the case looks just big enough to hold the board since the pots and input connectors are board mounted. The front panel looks to be at least 1/8 inch beep, maybe more. I'm going to guess that the box is 4.5 inches wide, by 6.5 inches deep. My board has 2 X 10,000 uF 35 volt capacitors in the middle of the board. The unit pictured has one 4700 uF 50 Volt cap. These caps are wired directly across the power supply.
I tried my board on a 22 volt 5000 mAH LIPO model helicopter battery. The old Sorensen linear power supply sounds almost as good. The peak currents drawn by these boards can cause some SMPS's to limit their output. wire a good sized choke in series, and a fat cap across the amps power input, and the SMPS begins to sound like a battery.
I tried my board on a 22 volt 5000 mAH LIPO model helicopter battery. The old Sorensen linear power supply sounds almost as good. The peak currents drawn by these boards can cause some SMPS's to limit their output. wire a good sized choke in series, and a fat cap across the amps power input, and the SMPS begins to sound like a battery.
Is there a workaround for SMSP 's that do not like to run into capacitive loads?
There are two reasons that some SMPS's don't like a large capacitive load.
1)A large capacitor draws a huge inrush current that trips the overcurrent protection circuit in the SMPS. In this case the SMPS will not turn on, or puts out one short pulse and shuts off.
2) The large capacitor creates a low frequency zero in the loop control dynamics causing the loop to become unstable. In this case the SMPS may start up, but output the wrong voltage, output a drifting unstable voltage, exhibit poor regulation, or shut down due to overvoltage on its output.
In either case the solution is the same, you must decouple the large cap from the SMPS. Some experimentation is required, and some SMPS's will just not work no matter what you do. I have found that adding an Inrush Current Limiter between the SMPS and the cap helps the most. I use something from the GE CL series, but a different one will be needed in each case depending on the load and size of the SMPS. I have used the CL-30 or CL-40 with most small SMPS's. Adding a choke sometimes helps. Even though the SMPS runs in the 100+ KHz region, its control loop is usually running in the tens of Hz. A large iron core choke can break up the zero created near DC by adding the huge cap across the output. I have some old chokes from scrap power supplies, but the Triad C-56U should work. The big Triad choke is a bit larger than the ones I am using, so there is something else out there that should work.
I have a Triad SMPS that refuses to work right with a cap across it and nothing seems to work. It is also quite noisy electrically. It puts out more high frequency crud that the unshielded TI 3100 breadboard.
http://www.ge-mcs.com/download/temperature/920-325D-LR.pdf
http://www.alliedelec.com/images/products/datasheets/bm/TRIAD_MAGNETIC/70218081.pdf
my god. what TI is thinking. it's using the same little shielded ferrite.Have you guys seen the new TPA3131D2? I just got email from TI - it is a tiny 5 mm x 5 mm chip and looking at the data sheet can make 20 watts/ch with 0.07% THD into 4ohms with 19v supply.
i'm sure it'll have more 'cracking' distortion.
must change them to unshielded iron toroid immediately!
I do like battery powered audio (though there is conflicting research) but in this case I do not think of it as an efficient solution.
First, two batteries will be needed to achieve 24V. If someone is fine with the 12V power rating, I think would be better go with the Sure TPA3110. Smaller, cheaper and most likely better built.
But the problem is not there. Unless you spend some cash on big batteries, like 40-60AH car batteries. The usual suspect, 7AH 12V nobreak batteries will not provide a much long listening time. Maybe they will hold it for a couple hours, but will start dropping the voltage quickly. Plus the physical size of two big batteries, long time charging and the possibility of hazardous materials.
I am speculating based on the datasheet over current rating (7.5A) and what a normal use would require, perhaps 3-5A.
In the end, there isn't anything technical to stop anyone from doing it. It's about how much one is willing to spend on it. Batteries are great for low powered systems, but in this subject it makes more sense to invest in a good power supply in my view.
An 18ah battery is 35 incl. Shipping. 4a charger is 25 shipped. My battery currently last for 17 days and needs 10 hours of charge time I think that is pretty darn reasonable given the simplicity of setup and the relative small size of the battery. I am not sure why you believe that a battery will only last a few hours. the amp does not draw very much when no signal is coming through not to mention it is operational down to 4.5V. in fact I am confused as to where you are getting the 3 to 5 amps necessary rating.
Last edited:
i don't get the pbtl-mode running.
this is what i tried:
took an old 3.5mm-connector.
soldered 10k resistors onto th r+ and l+.
connected r+ and l+ to have a mono output from my pc.
cut off the 4th pin of a 4-port-connector to have another fitting 3-pin-connector.
connected the l and the g together with the ground of the previously soldered mono-output.
when i try connect rout+ and rout- for the bridged out+ and the lout+ and lout- for the bridged (as shown in the datasheet) i only hear ****.
when i connect rout+ and lout- as the mono output of the amp it sounds good.
but i cannot really if it is pbtl'ed....
this is what i tried:
took an old 3.5mm-connector.
soldered 10k resistors onto th r+ and l+.
connected r+ and l+ to have a mono output from my pc.
cut off the 4th pin of a 4-port-connector to have another fitting 3-pin-connector.
connected the l and the g together with the ground of the previously soldered mono-output.
when i try connect rout+ and rout- for the bridged out+ and the lout+ and lout- for the bridged (as shown in the datasheet) i only hear ****.
when i connect rout+ and lout- as the mono output of the amp it sounds good.
but i cannot really if it is pbtl'ed....