www.justblair.co.uk

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.

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!

Comments

Add New Search

Write comment

Name: Email:  do not notifynotify Website: Title: UBBCode:         -color-aquablackbluefuchsiagraygreenlimemaroonnavyolivepurpleredsilvertealwhiteyellow -size-tinysmallmediumlargehuge     Please input the anti-spam code that you can read in the image.

Nelson Fernandes  - TA2024 Input Capacitor |18-09-2008 22:05:45 Hello Justblair, First of all I would like to congratulate you for the great job that you've made, guiding me improving my amplifier board. I'm almost going to do this MOD, but I'm uncertain if it's a good idea to use some small 2.2uF electrolytic capacitors that I have. I don't wish to enlarge my board above the original dimensions, since I intend to put it on a small aluminum box about the same size. Also, for VDD1 and VDD2 , I'm going to put 4 X 1000uF and I don't intend to use premium capacitors... The PS that I'm going to use is regulated, has transformer and can give pulses at about 2,5A. Isn't much but I trust on it and the capacitors are going to be a good help since the output stage of the PS doesn't have more than 10uF (input capacitor has 4700uF). Now the big question: I'm thinking about to build a small protection for the amplifier, based on a PIC microcontroller. It would enable... Reply | Quote 0 0

Nelson Fernandes  - TA2024 Input Capacitor |18-09-2008 22:08:50 It would enable power to the amplifier just when the input voltage is between 11 and 13V, or something like that, to protect it for overvoltage... Is it a good idea? Limiting the input voltage like this would affect the sound of it in a perceptible way? Thank you for your attention, Best Regards, Nelson Reply | Quote 0 0

Justblair |18-09-2008 22:16:22 I think that you could easiy use some nice electrolytics on the input. Polyprops are big. Mind though in my eventual design the big caps will be in a more manageable place. I mounted then that way just for testing purposes... Electrolytics come in all flavours, but one thing to look out for is the polarity of the caps. As the bias on the tripath chip is +ve, you would orientate appropriately. There are plenty of brands such as Black Gates et all that are favoured by others. Reply | Quote 0 0

Justblair |18-09-2008 22:20:23 As for your second question, yes I am sure that you could do that. But it seems to me to be a bit of a long way round for a shortcut. How much overvoltage do you anticipate? A clamping diode could be used instead. I suspect though that unless you are getting big fluctuations your chip will tolerate the overvoltage for brief pulses. I think I have reached the end of my knowledge on that one! Reply | Quote 0 0

Nelson Fernandes |20-09-2008 11:59:45 Hello again, Thank you for your quick reply. Yesterday I did successfully all mods shown on your site. I didn't found those premium components, but regarding the input electrolytic caps, I've tested their values before applying them. As for the big electrolytic caps, I've used 1000uF/16V from the best quality I found at my regular shop. Also Offset went down to 66mV and 76mV. Reply | Quote 0 0

Nelson Fernandes |20-09-2008 12:08:34 My next stage is testing a LM7812 if it doesn't heat up too much... Also I will fit everything in a small box, like this one: http://cgi.ebay.com/Aluminum-Project-Box-34-Ins... If I need the PIC solution, maybe a bigger box will be better Also, I'm using medium speakers, 30W 4Ohm and I'm thinking also applying a pair of tweeters that don't need crossover. Still, do I need some kind of filter applied to the speaker? Thank you for attention to my newbie questions. Best regards, Nelson Reply | Quote 0 0

Justblair  - Congratulations |23-09-2008 02:21:07 Hi Nelson Congratulations on completing the modifications. Could you tell me f you feel they were worth it? Incidently I have added another power line mod. I am very interested in hearing if this was a success for anyone else. As for your tweeter, I don't have experiance of tweeters that do not require a crossover. But I suppose some enterprising person will have built a unit with built in crossover. Matching it to your own system might be hit and miss though. Reply | Quote 0 0

Nelson Fernandes |25-09-2008 22:16:24 Hi, Thanks Yes, I think they were worth. But my hearing is very bad for these evaluations... I noted more beat and perhaps a more crystalline sound at higher frequencies. About the tweeters... I'm not going to worry about that right now. I'm going to build a aluminum box, maybe with the Tripath logo marked on the top. Then I'm going to think if it's worth to buy somethiing like the mission 753 but much more underpowered Thanks for help and guidance! Best regards, Nelson Reply | Quote 0 0

Justblair |25-09-2008 22:29:38 Glad to hear it went well for you, Yes these amps deserve some nice speakers to go with them. The Mission 753 is an old model now, but there are plenty of good speakers around. Next set of speakers for me will be DIY, but I have a lot more enjoyment to take from my Missions just yet. Reply | Quote 0 0

Nelson Fernandes |26-09-2008 01:00:22 What is your opinion about the Mission M30I? Or some other kind of speakers but still in the same style? Can you recommend me a pair with a good quality/price relation? All I can say about the M30I is that I liked very much their design. Reply | Quote 0 0

Justblair |26-09-2008 15:01:20 I don' have any experiance with this model. I used to sell HiFi many years ago, but I have not auditioned speakers for a while. I like the appearance of the speakers, very nice design. Mission has changed hands since my models were produced, Looking on the internet, they seem to get favourable reviews, though other speakers may sound better for the same price. Perhaps someone else can advise? Reply | Quote 0 0

adam |16-11-2008 05:35:16 I broke r14 off when attempting this mod. do you know what to replace this with? Thanks Reply | Quote 0 0

Justblair  - R14 Values |16-11-2008 12:25:17 Hi there. The value that you need to replace R14 will depend on what value Sure have put on the board. In the 5 boards that I have Sure have gone for either of two values 22K or 20K. If you look at the TA2024 Gain calculator http://www.justblair.co.uk/tripath-2024-gain-ca... you will see the effect of different values in this location. You can either measure R8, or even read its value of the top of the resistor 203=20k, 223=22K. If you are replacing the resistor, can I suggest that you replace R8 at the same time and try and find resistors with as low a tolerance (measured as a percentage) as possible. Reply | Quote 0 0

Powered by !JoomlaComment 4.0 beta1

!joomlacomment 4.0 Copyright (C) 2009 Compojoom.com . All rights reserved."