Indeed, Im sure I should have done that in hindsight. However, I felt the cheap lytics had to go so went with my gut feeling on the feedback caps. I'll give things a day or 2 more to settle in and see how it sounds.
I'm using an old Marantz CD-46 as source connected via the headphone jack with a bodged interconnect made from a QED Quinnex 2. As I didnt have a 3.5mm stereo jack I used one off some old headphones.
Tonight I decided to re build it with a nutrick 3.5 jack. As the PC (and Marantz) arent earthed I decided to connect the shield at both ends. Soldered everything with Eichmann silver solder and used teflon tubing to insulate the sheild leadouts. It seems to have made a slight difference. Right now im listening to Pink Floyd 'Delicate sound of thunder' and the sound is rather good. I still have the issue of weak bass though. I suspect its due to the cap cahnge.
The input cap (and bypass) has a combined capacitance of 10.22uF. This should be plenty as the datasheet for the TEA2025 recomends 0.22uF minimum. I have 10uf OSCONS here with .22 Wimma's as bypass.
The feedback cap, though a different type (Tant bead) is of the same value as the original, so I cant see how this would effect the bass too much.
The same goes for the caps that link into the output. Same value just changed fron cheap lytics to Rubycon ZLH.
I'll try a few different combinations and also give the 100uF Nichion muse a try in different areas. Other than that, all thats left to try is adding capacitance to the main PSU resovouir cap. I dont think its worth adding high end resistors of upgrading the diodes. It probably wouldnt make a noticeable difference with this chip.
Stil looking for advice on a reasonable pot to use?
I'm using an old Marantz CD-46 as source connected via the headphone jack with a bodged interconnect made from a QED Quinnex 2. As I didnt have a 3.5mm stereo jack I used one off some old headphones.
Tonight I decided to re build it with a nutrick 3.5 jack. As the PC (and Marantz) arent earthed I decided to connect the shield at both ends. Soldered everything with Eichmann silver solder and used teflon tubing to insulate the sheild leadouts. It seems to have made a slight difference. Right now im listening to Pink Floyd 'Delicate sound of thunder' and the sound is rather good. I still have the issue of weak bass though. I suspect its due to the cap cahnge.
The input cap (and bypass) has a combined capacitance of 10.22uF. This should be plenty as the datasheet for the TEA2025 recomends 0.22uF minimum. I have 10uf OSCONS here with .22 Wimma's as bypass.
The feedback cap, though a different type (Tant bead) is of the same value as the original, so I cant see how this would effect the bass too much.
The same goes for the caps that link into the output. Same value just changed fron cheap lytics to Rubycon ZLH.
I'll try a few different combinations and also give the 100uF Nichion muse a try in different areas. Other than that, all thats left to try is adding capacitance to the main PSU resovouir cap. I dont think its worth adding high end resistors of upgrading the diodes. It probably wouldnt make a noticeable difference with this chip.
Stil looking for advice on a reasonable pot to use?
mikesnowdon said:
The feedback cap has an large effect on audio--exactly as much as the input filter (dc block) cap.
When using a bypass cap to a larger cap that's carrying an audio signal, its good to use a "necessary" bypass cap otherwise the results are frequenty either a decrease of clarity or a loud midrange boost, or both. With plastic or ceramic caps, one can use a smaller than anticipated size.
For either case, and more, its possible that you may have arranged for a midrange boost rather than decreased bass.
A reasonable pot is alpha A20k Dual Gang (audio type stereo) and its very cheap. Between that and the more expensive pots are differences of manufacturing variance and longevity.
However. . .
Over at Decibel Dungeon http://myweb.tiscali.co.uk/nuukspot/decdun/gainclonepre.html#gcprevc is a nice spot to tell you how to do a "law fake" for perhaps a B50k Dual Gang pot (linear), turning it into a 10k to 20k (or so) audio type. A law fake is merely an additional load on the "amplifier side" of a linear pot.
Pots.
I found 10k and 20k Alps pots on ebay going cheap, is it worth it?
Another option would be to use the pot I removed from the tone circuit as its identical.
Soundwise the amps producing quite nice sound for what it is. It seems to lack punch in the bass though, may be due to the cap changes. Im going to be paitent and see if things change after 1 more week. I decided to get the PSU kit from audioupgrades.co.uk. For the money its great value: PCB, Schottky diodes , 4700uF Panasonic FC, 220uF Rubycon ZA, and a 78/9** reg. All for under £20. I'll only use the diodes and 4700uF cap to begin with, adding the one I allready have to double up the smoothers. If I order the kit with 7809 reg I can try adding that later for experimentation.
I think first I should better understand the chip.
I need to be sure that the feedback caps are the middle 2, and I need to know what the last pair are doing (the ones that link into the output)??????
I found 10k and 20k Alps pots on ebay going cheap, is it worth it?
Another option would be to use the pot I removed from the tone circuit as its identical.
Soundwise the amps producing quite nice sound for what it is. It seems to lack punch in the bass though, may be due to the cap changes. Im going to be paitent and see if things change after 1 more week. I decided to get the PSU kit from audioupgrades.co.uk. For the money its great value: PCB, Schottky diodes , 4700uF Panasonic FC, 220uF Rubycon ZA, and a 78/9** reg. All for under £20. I'll only use the diodes and 4700uF cap to begin with, adding the one I allready have to double up the smoothers. If I order the kit with 7809 reg I can try adding that later for experimentation.
I think first I should better understand the chip.
I need to be sure that the feedback caps are the middle 2, and I need to know what the last pair are doing (the ones that link into the output)??????
Caps:
Capacitors don't change their sonic signature greatly after break-in. There are few exceptions. For most caps, if it doesn't sound better in 15 minutes, its not going to, and if it sounds bad immediately, it won't stop doing that; however, if it sounds good immediately, then it will probably keep doing that.
So, if you've made a no-bass arrangement, then the task is to figure out which one of the changes or which several of the changes was responsible for that modification. This won't suddenly change (back) by itself. The power supply, feedback cap, output cap, and input filter caps are suspect as the cause.
Hotrod with caps is possible:
For increasing the impact of an audio AC pulse, there are some capacitors made for it. Those that help break the drywall in nightclubs everywhere are also good for hot-rodding fun. That's Nichicon's ES, Blackgate's N, Elna's Cerefine, and very few others. Although made for audio signal, most also do amazing things in power supplies.
All of these are more difficult to apply than a plain wire, because, with any capacitor, your application may not work with the ideal bandwidth of your purchase.
If an input filter cap misses the high end slightly, there's the "bypass cap" practice for hours of tinkery fun. Its also possible to put a resistor (avg value 20R) onto the input side of a very small value MKT bypass cap (avg value 180nF) in order to sink the mids without reducing perceived clarity--eventually resulting in more apparent bass response.
Volume control:
Alps makes some of the best potentiometers and some of the worst too. The value of a volume control is the technology inside, successfully applied to the application.
I'm pretty sure that Alpha's cheap carbon beats Alps cheap carbon by a landslide, so I wouldn't pay any more than $2 for a carbon pot.
However, Alps also makes pots from more advanced materials like plastics, ceramics and perhaps pixie dust? Many other manufacturers also have advanced materials options.
A step ahead of the carbon or plastic pots is the "stepper" style, involving a rotary switch and a lot of good quality resistors.
Next up are the more interesting technologies of the lightbulb>CDS, the Joshua, and the transformer based volume controls that don't cost more than the highest "zoot factor" pot, but can outperform them all (depends on application).
The usage of the highest quality volume controls is somewhat regulated by system symmetry (whether or not it benefits the application) although quality input is always a good starting place.
Testing:
In theory, speakers are an awful thing to test with for home projects; however, headphones can do well for testing amplifiers. If using headphones to test, then, next time you move or buy new speakers, then you don't get the task of re-doing the amplifier. Here's a great article by Rod Elliot: http://sound.westhost.com/project100.htm
Capacitors don't change their sonic signature greatly after break-in. There are few exceptions. For most caps, if it doesn't sound better in 15 minutes, its not going to, and if it sounds bad immediately, it won't stop doing that; however, if it sounds good immediately, then it will probably keep doing that.
So, if you've made a no-bass arrangement, then the task is to figure out which one of the changes or which several of the changes was responsible for that modification. This won't suddenly change (back) by itself. The power supply, feedback cap, output cap, and input filter caps are suspect as the cause.
Hotrod with caps is possible:
For increasing the impact of an audio AC pulse, there are some capacitors made for it. Those that help break the drywall in nightclubs everywhere are also good for hot-rodding fun. That's Nichicon's ES, Blackgate's N, Elna's Cerefine, and very few others. Although made for audio signal, most also do amazing things in power supplies.
All of these are more difficult to apply than a plain wire, because, with any capacitor, your application may not work with the ideal bandwidth of your purchase.
If an input filter cap misses the high end slightly, there's the "bypass cap" practice for hours of tinkery fun.
Its also possible to put a resistor (avg value 20R) onto the input side of a very small value MKT bypass cap (avg value 180nF) in order to sink the mids without reducing perceived clarity--eventually resulting in more apparent bass response. Volume control:
Alps makes some of the best potentiometers and some of the worst too. The value of a volume control is the technology inside, successfully applied to the application.
I'm pretty sure that Alpha's cheap carbon beats Alps cheap carbon by a landslide, so I wouldn't pay any more than $2 for a carbon pot.
However, Alps also makes pots from more advanced materials like plastics, ceramics and perhaps pixie dust? Many other manufacturers also have advanced materials options.
A step ahead of the carbon or plastic pots is the "stepper" style, involving a rotary switch and a lot of good quality resistors.
Next up are the more interesting technologies of the lightbulb>CDS, the Joshua, and the transformer based volume controls that don't cost more than the highest "zoot factor" pot, but can outperform them all (depends on application).
The usage of the highest quality volume controls is somewhat regulated by system symmetry (whether or not it benefits the application) although quality input is always a good starting place.
Testing:
In theory, speakers are an awful thing to test with for home projects; however, headphones can do well for testing amplifiers. If using headphones to test, then, next time you move or buy new speakers, then you don't get the task of re-doing the amplifier. Here's a great article by Rod Elliot: http://sound.westhost.com/project100.htm
Just thought I'd end off on this one and post a final piccy of the amp as it is now.
I rescued the Pana FC'c and fitted a 4700uF Rubycon PK. Added some Oscons for signal decoupling duties. Left the rest of it including the ALPS pot as it was secondhand anyway. Theres still quite a few nice caps in there which I'll probably salvage when I need them for more worthy equipment. The amp sounds allright though, not hi-fi but listenable and it suits the Missions for nearfield listening. It has a groovy nature and being used nearfield it relly grabs your attention. Surprisingly it does draw you in to the music.......(as long as you dont push it too much - distortion)
As I said, this project was fun and I learnt from it. I had a ball with it.
I thought of a name for it: OTTA (as in otter, small and cute, OTTA also stands for something - Over The Top Amp! )
And this is what I started with:
I rescued the Pana FC'c and fitted a 4700uF Rubycon PK. Added some Oscons for signal decoupling duties. Left the rest of it including the ALPS pot as it was secondhand anyway. Theres still quite a few nice caps in there which I'll probably salvage when I need them for more worthy equipment. The amp sounds allright though, not hi-fi but listenable and it suits the Missions for nearfield listening. It has a groovy nature and being used nearfield it relly grabs your attention. Surprisingly it does draw you in to the music.......(as long as you dont push it too much - distortion)
As I said, this project was fun and I learnt from it. I had a ball with it.
I thought of a name for it: OTTA (as in otter, small and cute, OTTA also stands for something - Over The Top Amp! )
And this is what I started with:
Forgot to mention.
I had experimented with a adding a second 4700uF Panasonic FC. I was hoping for better bass but what I got was overly revealing midrange. With a better quality chip the additional cap might have had a positive effect. I opted for the Rubycon in the end as it works better with the chip. I also added shottkey diodes which improved the dynamics and made the bass better defined. The ground now runs from the star point at the terminals to the negative side of the main smoother. This cleaned up the sound a little.
I still have a hankering to try putting the Tants back and possibly add a small 100uF Bypass cap on the main smoother at some point. But for now Im enjoying a very listenable sound. Its not the smoothest but it has good timing and reasonable detail. Its a little bright though so I might pad down the tweeters in the speakers in using. One thing I really like is the punchy nature it has. Basslines really get you moving.
I think the grainy sound is due to the chip and all the good components are revealing its shortfalls. My inkling is that putting the tants back in for the feedback and output (currently using Rubycon ZLH) will warm things up a bit.
I had experimented with a adding a second 4700uF Panasonic FC. I was hoping for better bass but what I got was overly revealing midrange. With a better quality chip the additional cap might have had a positive effect. I opted for the Rubycon in the end as it works better with the chip. I also added shottkey diodes which improved the dynamics and made the bass better defined. The ground now runs from the star point at the terminals to the negative side of the main smoother. This cleaned up the sound a little.
I still have a hankering to try putting the Tants back and possibly add a small 100uF Bypass cap on the main smoother at some point. But for now Im enjoying a very listenable sound. Its not the smoothest but it has good timing and reasonable detail. Its a little bright though so I might pad down the tweeters in the speakers in using. One thing I really like is the punchy nature it has. Basslines really get you moving.
I think the grainy sound is due to the chip and all the good components are revealing its shortfalls. My inkling is that putting the tants back in for the feedback and output (currently using Rubycon ZLH) will warm things up a bit.
I tried putting the Tants back in the feedback, added warmth but also gave me a distorted top end. Replaced the Rubycon ZLH's in the feedback and tried the Tants on the bootstrap. Also I added a 220uF Rubycon ZL to the main smoother.
Much better sound was achieved! Detail seperation has improved nicely. Top end extension is better and more detailed. The amp seems to do a better job at reproducing the 'air' of instruments. Bass is more natural, still goes quite deep but has better punch now, tighter. The amp sounds a bit toppy though but this could be because I have my Mission speakers upside down. I'll have a bit more of a play with the speakers later. (Planning to mod them a bit sometime. Polyprop caps and new resistors (probably a slightly higher value on the tweeter).
Might try 100uF Tants on the input with the Wimma polyprop bypass caps. Currently I have 10uF Oscons with Wimma bypass. Adding the Larger value Tants might improve bass and give a bit of extra warmth to the sound.....
Overall there seems to much better clarity and air to the sound, and detail has improved further. Im happy.
Mike.
Much better sound was achieved! Detail seperation has improved nicely. Top end extension is better and more detailed. The amp seems to do a better job at reproducing the 'air' of instruments. Bass is more natural, still goes quite deep but has better punch now, tighter. The amp sounds a bit toppy though but this could be because I have my Mission speakers upside down. I'll have a bit more of a play with the speakers later. (Planning to mod them a bit sometime. Polyprop caps and new resistors (probably a slightly higher value on the tweeter).
Might try 100uF Tants on the input with the Wimma polyprop bypass caps. Currently I have 10uF Oscons with Wimma bypass. Adding the Larger value Tants might improve bass and give a bit of extra warmth to the sound.....
Overall there seems to much better clarity and air to the sound, and detail has improved further. Im happy.
Mike.
You might see some improvement by removing the pot and using a DPDT switch with a few select resistor values set to your normal and low listening levels. I have had much success in the past by doing so and also doing the shunt mod. (Not the law fake shunt mod. This one http://www.wduk.worldomain.net/forum/showthread.php?t=198) The problem you will run into with the shunt is it will attenuate signal some and I don't think you have much to spare.
The alps blue is a good pot but it's not better than no pot.
The alps blue is a good pot but it's not better than no pot.
Hmm, 2.5w, good bass, slightly toppy, smooth clear mids. . . That's just the thing for full range speakers. Perhaps some of the tang band? There's a goodly amount of available documented projects for those. There's often a notch filter involved, but a splurge on the inductor gauge (use thick wire inductor) will help preserve the efficiency.
I think that we all need a cute little "relax" system that's breathlessly clear, realistic, and takes up almost no space. . . because the "listening room" is not the exclusive location for music.
In my opinion, I think that you should preserve this project as a reference point, and I think that you should employ it at a location in your office or home, where it can be enjoyed regularly.
Congrats on the fine project!
With such good attention to detail, I can't wait to see what your next project will look like.
Perhaps some of the tang band? There's a goodly amount of available documented projects for those. There's often a notch filter involved, but a splurge on the inductor gauge (use thick wire inductor) will help preserve the efficiency. I think that we all need a cute little "relax" system that's breathlessly clear, realistic, and takes up almost no space. . . because the "listening room" is not the exclusive location for music.
In my opinion, I think that you should preserve this project as a reference point, and I think that you should employ it at a location in your office or home, where it can be enjoyed regularly.
Congrats on the fine project!
With such good attention to detail, I can't wait to see what your next project will look like.
Thanks for the ideas on volume controls. I'd also like to add a extra input with a small toggle switch.
Further tweaks I have planned are to add 10uF Wima MKS2-XL around the chip - particularly on the input. I also want to see what effect it has to remove the caps on the output.
Ive thought about buffering with a good op-amp. What do you guys reckon?
My next project is to make something special out of my Cambridge Audio 340A SE. Its a LM3886 based amp with (upgraded) LM4562 gainstage. The biggest compromise it has is solid state input switching. I have an idea to use the channel LED outputs from the microprocessor to control relays for input switching. So far I have added some nice caps/diodes and a better opamp with fairly good results. (Id like to post a link to the thread but I fear you will not post due to some of the arguments there have been).
Mike.
Further tweaks I have planned are to add 10uF Wima MKS2-XL around the chip - particularly on the input. I also want to see what effect it has to remove the caps on the output.
Ive thought about buffering with a good op-amp. What do you guys reckon?
My next project is to make something special out of my Cambridge Audio 340A SE. Its a LM3886 based amp with (upgraded) LM4562 gainstage. The biggest compromise it has is solid state input switching. I have an idea to use the channel LED outputs from the microprocessor to control relays for input switching. So far I have added some nice caps/diodes and a better opamp with fairly good results. (Id like to post a link to the thread but I fear you will not post due to some of the arguments there have been).
Mike.
Do you still have this TEA2025 project? I've been planning to build one for years to work with small speakers and high-power headphones.
Did you find it necessary to implement capacitors on the input? According to the datasheet they are only used to remove noise caused by cheap potentiometers. I had planned on building my unit with no volume adjustment as I planned on using it with a PC.
Can you show what your final values were for every capacitor? And what did you do for a power-supply?
Did you find it necessary to implement capacitors on the input? According to the datasheet they are only used to remove noise caused by cheap potentiometers. I had planned on building my unit with no volume adjustment as I planned on using it with a PC.
Can you show what your final values were for every capacitor? And what did you do for a power-supply?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.