Justblairs Audio and Electronic's Pages

So far I have modified several aspects of the Sure TA2024 board. In this article I look again at the input to the device and swap over the input capacitor.

The Sure Electronics TA2024 Amplifier Module sounds reasonable out of the box. Offering excellent value for money, it is a tempting prospect to the DIY builder. However as I reported in the mini review, the board does not sound as clear as it could.

So far I have:

• Removed the DC output correction from the board
• Removed the low pass filter on the input line
• Removed the Protection Diode on the power line
• Added 3000uf of rail capacitance to the VDD1 and VDD2 power lines
• Experimented with the voltage supplied to the chip

These modifications have made a difference to the sound of the board, some making more of an improvement than others.

Next aspect of the design to receive some attention is the input capacitor. The input capacitor is used on the amplifier to block DC voltage. On designs that I have worked on previously, i.e. Lm3875 or LM3886 based “Gainclones” the DC blocking capacitor serves the function of blocking DC voltages from further up the signal chain being amplified.

The input capacitor basically acts with the following input resistor as a high pass filter, blocking DC.

On Tripath based amplifiers, the amplifier inputs in fact are biased to a higher voltage, usually around 2.4v. The DC blocking capacitor in this configuration protects the components further up the input chain such as pre-amplifiers or the outputs to a CD player. However it still in conjunction with the input resistor (R8 and R14) acts as a high pass filter.

Looking at the Sure Electronics supplied schematic for the amplifier the input capacitors are identified as C13 and C21. Fair play to Sure for supplying the schematic. According to the schematic, C13 and C21 are specified as being 2.2uf. With input resistance of 20K for R8 and R14, the Sure Electronics boards have a high pass filter with a -3db point tuned to 3.62hz. Below 5hz is considered a good value for this high pass filter. All looks good then for the Sure design.

Having a closer look at the boards we can see the capacitors on the board. The capacitors are very small and would appear to be ceramic or film capacitors. Being generous we will say film. Ceramic capacitors have a reputation for not sounding so good. Polyester film is better.

For the purposes of my experiment, I wished to try out a polypropylene capacitor. In my parts box, I have 4.7uf Sonicap Gen 1 capacitors and I wished to try one out on the line in.

First step is to remove the input capacitors from the line in on the Sure Electronics boards. I used my KADA 852D smt solder station for this, heating the little capacitors and removing them with tweezers. The capacitor was absolutely tiny. I was suspicious that this little capacitor could have 2.2uf of capacitance in its tiny shell. I decided to test it using my capacitance meter. My suspicions proved to be correct. The capacitor supplied is in fact a 1uf capacitor.

The effect of the 1uf capacitor in combination with the 20K input resistor is to raise the value of the created high pass filter from the 3.62hz value (remember under 5hz is good) to 7.96hz. Now that is the -3db point that is measured. The filter will start to reduce the bass at a higher frequency.

Now in my final design, I intend to bridge the pads that are left, and attach the very large Sonicap Gen 1 capacitors off of the board rather than try to mount them on the board. However for my first listening test I wanted something quick and dirty to try out. I decided to solder my Sonicaps directly on to the board. This is by no means a permanent solution and if you choose to make this mod. The size of the Sonicaps is more than the solder pads will take under anything other than the utmost of care in handling them. Others (such as audio1st) across at www.diyaudio.com have used physically smaller epcos 2.2uf caps on their boards which are far more compact and can be attached this way without problems.

Soldering the legs of the Sonicap Gen 1 capacitors on to the board proved to be easy. I just held them in place and applied the soldering iron with a touch of solder on the tip. This was enough to make a temporary join to the boards.

Now I could compare the input capacitor modified board with the stock Sure supplied Capacitor.

Listening

Powering up the amplifiers and testing them with my Mission 753 speakers, it did not take long to notice some differences in the sound. The sound was richer sounding in the bass regions. Playing double bass pieces through the amplifiers the amp with the Sonicap input capacitor sounded more like a real instrument. The high point of these amplifiers is for me in the higher frequencies, but now with the capacitor swapped the lower frequencies were just as enjoyable to listen to. The overall balance of the amplifier sounds more “right” to my ears.

Playing some tracks with female vocals other differences were apparent. The transition through the frequencies was more fluid with the new capacitors in place. Female vocals also became more melodic.

I don’t want to over egg the pudding with my listening reports. Essentially the good points of a Tripath 2024 based amplifier are the good points. Changing a component over like this is not going to revolutionise the sound of the amplifier. However in my setup, the swap of the capacitors was a worthwhile modification to make. I have de-soldered them from the pads as I don’t want them damaged before I get a chance to bridge them.

Also of note is that many people (including myself) believe that capacitors take some hours of play before they give their best. The Sonicaps have a reputation for requiring lots of play before they perform at their best. I am looking forward to getting the Sure Electronics Boards into their eventual project destination so that I can give all components a decent burn in.

Realistically if you wish to try this modification, you should not need to go as high as 4.7uf as a replacement capacitor value. 2.2uf will give a high pass filter -3db point of 3.62hz, 3.3uf a value of 2.41hz and a 4.7uf gives 1.69hz. Others have reported good results using 2.2uf and 3.3uf values. Should you wish to change over the input resistors to a different value, my Tripath 2024 calculator here will do the frequency and gain calculations for you.

If you like making micro-improvements, how about using the roulette winning strategy from GamblingCity.net to increase your roulette payout when you play online roulette? Stop in today for some online roulette gambling at GC's internet roulette tables!