mikesnowdon
resU deretsigeR
As some of you may know I've been researching and asking questions regarding a DIY desktop amp I wanted to build. This thread is firstly to show off the new amp (assembled and running now) and secondly to ask for help on a few issues I still need to resolve.
A little background for those who havent read the orignal thread:
I had a pair of Desktop pc speakers with a built in amp and rubbish drivers. Sitting up in my loft were a lonely pair of Mission 700 series speakers which I wanted to use on the pc. Having no amp to drive them I decided 'just for a laugh' to rig the pc speakers internal amp and use it to drive the Missions. I was pleasntly suprised and, as is the nature of our hobby, I began to ponder what mods I could do the improve sound quality....
The first and most obvious mod was to improve the power supply. So I ordered a small toroidial transformer of nearly double the VA of the original. I also changed the resovoiur/smoothing cap from a 1000uF 16v 'junk' lytic, to a Pana FC 4700uf 25v. Of course the other big problem was the chassis. The amp module was housed in a thin plastic speaker enclosure and that combined with the rubbish original transformer meant the whole thing would resonate from the tx vibration. A new chassis was ordered from www.autocostruire.com.
The tone pot was removed and its associated componentry, and the '3D-Surround' switch and components removed as they were not required and I wanted to keep the signal path as simple as possible. The datasheet for the chip (TEA2025) stated a massive 470uf cap was needed for the output decoupling, so swapping for polyprops was out of the question due to size and cost. Instead I used a few of the holes (made available from the pot and switch removal) to mount a pair of 0.15 Mundorf MKP bypass caps to the 470uF lytics (linked underneath).
I had initaily played with 1uF Mundorf MKP's on the chip input, in place of the 100uF lytics which were there before. This had very good results but ufortunately I played around with the Mundorfs too much (adjusting their position) and broke one of the leads! I've put in a couple of 100uF tants for now.
I also had a Blackgate STD 220uf cap spare - so I added this in place of the 100uF decoupling cap on the chip. Also added a small heatsink to the chip as it gets rather hot driving the Missions!
I have some Tants, Rubycon, and Nichion Muse caps to play with in the feedback area (feel free to advise any better options here), and I also plan to get some more 1uF or 2.2uF Polyprops for the chip input. Rather than Mundorfs I want to try those small square ones (rookie alert!) as they will fit in the limited space better. Incidentaly, the chip datasheet suggests a value of 0.22uF on the chip input but the amp originally had 100uF lytics?
TEA2025 datasheet: http://www.datasheetcatalog.com/datasheets_pdf/T/E/A/2/TEA2025.shtml
There are only 2 manufacturers of this chip so as I allready have the UTC one I decided to buy a 'ST' one too. I'll try the other chip and see if theres a difference and I have a spare incase I cook one.
The original thread is here: http://pinkfishmedia.net/forum/showthread.php?t=50566
Well, the amps only been up and running for a few hours so I wont say much about sound quality. I can say that the beefed up power supply has made a diffrence. For a start the new Tx runs really quiet and the hiss/hum issues I had previously are solved. The sound has more guts now, sounding warmer and fuller.
Below is a picture of the amp as was before:
And heres the new version:
Internals:
Theres a couple of things to resolve before im happy the project is complete:
The 4700uF smoothing cap is too tall to mount on the PCB so Ive bodged a teddyreg board and wired it to the main PCB. Im not happy that this is doing any favours to the sound quality so I want to ask for advice on a better soloution. I cant mount the Pana on its side as it wont fit cleanly on PCB. Would it be better to connect the cap to the PCB with some solid copper wire, and ziptie it to the chassis, keeping it as close as possible to the PCB?
Im also unsatisfied with the hookup wire. I used some cheap Cambridge audio coax for the signal inpput and 1.5mm bellwire for the output to the speaker binding posts. What would be a better wire to use? (silver would be ott for this project), some kind of OFC stuff perhaps?
I havent connected the earth pin on the IEC to the chassis yet. Im wondering about a star earth (connecting all the grounds on the terminals together) but I dont know enough about that yet. Any advice?
Also the pot has become noisy but it was fine before, no idea why. Whats happened?
Lastly, is there a better soloution for the output decoupling caps? The main consideration here is that according to the datasheet anything less than 470uF will reduce bass output. Any better ideas than what I've allready done?
Please feel free to advise any other possible mods Ive not noticed.
As a closing remark, I'll say that this is my first amp and Ive really learnt alot from it so far and its been fun. Thanks to all who helped on the original thread, in particular PD.
Cheers,
Mike.
A little background for those who havent read the orignal thread:
I had a pair of Desktop pc speakers with a built in amp and rubbish drivers. Sitting up in my loft were a lonely pair of Mission 700 series speakers which I wanted to use on the pc. Having no amp to drive them I decided 'just for a laugh' to rig the pc speakers internal amp and use it to drive the Missions. I was pleasntly suprised and, as is the nature of our hobby, I began to ponder what mods I could do the improve sound quality....
The first and most obvious mod was to improve the power supply. So I ordered a small toroidial transformer of nearly double the VA of the original. I also changed the resovoiur/smoothing cap from a 1000uF 16v 'junk' lytic, to a Pana FC 4700uf 25v. Of course the other big problem was the chassis. The amp module was housed in a thin plastic speaker enclosure and that combined with the rubbish original transformer meant the whole thing would resonate from the tx vibration. A new chassis was ordered from www.autocostruire.com.
The tone pot was removed and its associated componentry, and the '3D-Surround' switch and components removed as they were not required and I wanted to keep the signal path as simple as possible. The datasheet for the chip (TEA2025) stated a massive 470uf cap was needed for the output decoupling, so swapping for polyprops was out of the question due to size and cost. Instead I used a few of the holes (made available from the pot and switch removal) to mount a pair of 0.15 Mundorf MKP bypass caps to the 470uF lytics (linked underneath).
I had initaily played with 1uF Mundorf MKP's on the chip input, in place of the 100uF lytics which were there before. This had very good results but ufortunately I played around with the Mundorfs too much (adjusting their position) and broke one of the leads! I've put in a couple of 100uF tants for now.
I also had a Blackgate STD 220uf cap spare - so I added this in place of the 100uF decoupling cap on the chip. Also added a small heatsink to the chip as it gets rather hot driving the Missions!
I have some Tants, Rubycon, and Nichion Muse caps to play with in the feedback area (feel free to advise any better options here), and I also plan to get some more 1uF or 2.2uF Polyprops for the chip input. Rather than Mundorfs I want to try those small square ones (rookie alert!) as they will fit in the limited space better. Incidentaly, the chip datasheet suggests a value of 0.22uF on the chip input but the amp originally had 100uF lytics?
TEA2025 datasheet: http://www.datasheetcatalog.com/datasheets_pdf/T/E/A/2/TEA2025.shtml
There are only 2 manufacturers of this chip so as I allready have the UTC one I decided to buy a 'ST' one too. I'll try the other chip and see if theres a difference and I have a spare incase I cook one.
The original thread is here: http://pinkfishmedia.net/forum/showthread.php?t=50566
Well, the amps only been up and running for a few hours so I wont say much about sound quality. I can say that the beefed up power supply has made a diffrence. For a start the new Tx runs really quiet and the hiss/hum issues I had previously are solved. The sound has more guts now, sounding warmer and fuller.
Below is a picture of the amp as was before:
And heres the new version:
Internals:
Theres a couple of things to resolve before im happy the project is complete:
The 4700uF smoothing cap is too tall to mount on the PCB so Ive bodged a teddyreg board and wired it to the main PCB. Im not happy that this is doing any favours to the sound quality so I want to ask for advice on a better soloution. I cant mount the Pana on its side as it wont fit cleanly on PCB. Would it be better to connect the cap to the PCB with some solid copper wire, and ziptie it to the chassis, keeping it as close as possible to the PCB?
Im also unsatisfied with the hookup wire. I used some cheap Cambridge audio coax for the signal inpput and 1.5mm bellwire for the output to the speaker binding posts. What would be a better wire to use? (silver would be ott for this project), some kind of OFC stuff perhaps?
I havent connected the earth pin on the IEC to the chassis yet. Im wondering about a star earth (connecting all the grounds on the terminals together) but I dont know enough about that yet. Any advice?
Also the pot has become noisy but it was fine before, no idea why. Whats happened?
Lastly, is there a better soloution for the output decoupling caps? The main consideration here is that according to the datasheet anything less than 470uF will reduce bass output. Any better ideas than what I've allready done?
Please feel free to advise any other possible mods Ive not noticed.
As a closing remark, I'll say that this is my first amp and Ive really learnt alot from it so far and its been fun. Thanks to all who helped on the original thread, in particular PD.
Cheers,
Mike.