Hey all, I've finally finished by Sure TK2050 powered speakers, and they sound great! Here's a few pics, I've gone for function over form! 
Front:
Back (I'm listening to the one on the right...):
A quick test using my mp3 player:
I've only managed to listen to one at a time so far, as I still need to build a few XLR cables, but during the test using my Sansa Clip (with a flac file with HP and LP filters already applied) as a souce, they both sound good. Nice dynamics and also much cleaner bass compared to the previous LM3886 based amp that originally came with the speakers.
Once I get my XLR cables made and modded my DCX2496 I should be able to hear them in all their glory.
I'd like to say thanks to everyone in this thread, I wouldn't have managed to get things sounding so good without the great info contained in this thread.
Front:
An externally hosted image should be here but it was not working when we last tested it.
Back (I'm listening to the one on the right...):
An externally hosted image should be here but it was not working when we last tested it.
A quick test using my mp3 player:
An externally hosted image should be here but it was not working when we last tested it.
I've only managed to listen to one at a time so far, as I still need to build a few XLR cables, but during the test using my Sansa Clip (with a flac file with HP and LP filters already applied) as a souce, they both sound good. Nice dynamics and also much cleaner bass compared to the previous LM3886 based amp that originally came with the speakers.
Once I get my XLR cables made and modded my DCX2496 I should be able to hear them in all their glory.
I'd like to say thanks to everyone in this thread, I wouldn't have managed to get things sounding so good without the great info contained in this thread.
hi, im new to the forum and new to the whole diy audio thing in general and could do with a bit of help from those who know more than i do... i have one of these sure amps, well two actually but one is working fine so forget that one... the second, was working ok but... ive snipped off the wires to the fan as it added a lot of noise and i was running it on around 24v so it seemed unnececarry as it was barely warm even with the fan not attached to the heatsink.
After running it briefly like this, and with heatsink staying cool, the current and voltage on the psu began to fluctuate a bit and now it seems rather dead, ive got it hooked to a bench psu and where before it was running about 24v, now it wont take more than about 1.5v from the psu and hits the current limiting as the voltage is raised up from 0 where previously the current was tiny, and 0 before reaching around 10v. so i presume its shorting out somewhere. Neither the mute or power led's come on. Basically if anyone could give me an idea whether its totally buggered or whether its possible its something fixable, eg the voltage regulator the fan was connected to that would be great.
After running it briefly like this, and with heatsink staying cool, the current and voltage on the psu began to fluctuate a bit and now it seems rather dead, ive got it hooked to a bench psu and where before it was running about 24v, now it wont take more than about 1.5v from the psu and hits the current limiting as the voltage is raised up from 0 where previously the current was tiny, and 0 before reaching around 10v. so i presume its shorting out somewhere. Neither the mute or power led's come on. Basically if anyone could give me an idea whether its totally buggered or whether its possible its something fixable, eg the voltage regulator the fan was connected to that would be great.
Take a look
Take a look under the heatsink at the chips. You may see something obvious like missing legs or burns.
.
.
Take a look under the heatsink at the chips. You may see something obvious like missing legs or burns.
.
.
Summarizing thread
Thanks all for all your input in this thread,
I'm reading it for some time now but had no time to mod my own boards yet.
The thread is 165 pages long and is quite unusable now.
It would be great to start a new parallel thread just for each one to summarize its own mods and perhaps the ones he has rejected.
Nevertheless thanks again, hopefully I'll post my results soon.
Michel_MRS
Thanks all for all your input in this thread,
I'm reading it for some time now but had no time to mod my own boards yet.
The thread is 165 pages long and is quite unusable now.
It would be great to start a new parallel thread just for each one to summarize its own mods and perhaps the ones he has rejected.
Nevertheless thanks again, hopefully I'll post my results soon.
Michel_MRS
Am I correct in thinking that, because these chips are already bridged you cannot tie the 2 channels together using a device like this Balanced Input Phase Shifter?
Balanced
It wouldn't be like traditional bridging since neither of the speaker outputs is ground. It might be possible to parallel the two channels, with or without a balanced input. I may look into that at some point for use with a sub. It all depends on how close the dc off sets between the pairs of outputs can be adjusted towards zero. Then it might be possible with the addition of .1-.22 ohm output resistors.
It wouldn't be like traditional bridging since neither of the speaker outputs is ground. It might be possible to parallel the two channels, with or without a balanced input. I may look into that at some point for use with a sub. It all depends on how close the dc off sets between the pairs of outputs can be adjusted towards zero. Then it might be possible with the addition of .1-.22 ohm output resistors.
@sendler:
Your work on the TK2050 output filter is really interesting. I'm building some 41Hz amps, and I was going to try "porting" your filter on the Sure to a 41Hz amp out of curiosity. The premise make a lot of sense, i.e. "Buy" a cleaner audio-band filter by "paying" in out-of-band peaks.
I'm not sure the filter implementation there is exactly the same as in the Sure amps. I saw you posted a schematic of the modified output filter's values here: http://www.diyaudio.com/forums/clas...nics-new-tripath-board-tc2000-tp2050-101.html
My question is: Do you have the corner frequency and Q values for this filter? (Q values for, say, 4 and 8 Ohm.)
Thanks!
Your work on the TK2050 output filter is really interesting. I'm building some 41Hz amps, and I was going to try "porting" your filter on the Sure to a 41Hz amp out of curiosity. The premise make a lot of sense, i.e. "Buy" a cleaner audio-band filter by "paying" in out-of-band peaks.
I'm not sure the filter implementation there is exactly the same as in the Sure amps. I saw you posted a schematic of the modified output filter's values here: http://www.diyaudio.com/forums/clas...nics-new-tripath-board-tc2000-tp2050-101.html
My question is: Do you have the corner frequency and Q values for this filter? (Q values for, say, 4 and 8 Ohm.)
Thanks!
New filter sym
Any bridged class D amp can use this filter as long as you choose your coils to cover the expected peak currents. The Wurth XXL coils can do 10 amps. The new filter sounds great but does increase idle current consumption a bit so If you are an efficiency maniac you will want to stay above 10uH for the coils. You can see from the sym that the peaking really is no big deal. The impovement in sound with the wide open filter goes beyond simple changes in frequency response. There is something else going on with the lower value coils that makes them sound much better. The p and r around the coils are not real components. They were just included for the sake of a thorough sym as was the small amount of inductance of the tweeter in addition to the 8 ohm load voice coil. 4 ohm loads will have a smaller peak and be just as flat at 20K. The zobel really knocks down the peak's interaction with the driver inductance and could use a smaller resistor for 4 ohm loads. Set the resistor by ear for the best trade off for any given speaker between making the sound too aggressive or too boring.
Any bridged class D amp can use this filter as long as you choose your coils to cover the expected peak currents. The Wurth XXL coils can do 10 amps. The new filter sounds great but does increase idle current consumption a bit so If you are an efficiency maniac you will want to stay above 10uH for the coils. You can see from the sym that the peaking really is no big deal. The impovement in sound with the wide open filter goes beyond simple changes in frequency response. There is something else going on with the lower value coils that makes them sound much better. The p and r around the coils are not real components. They were just included for the sake of a thorough sym as was the small amount of inductance of the tweeter in addition to the 8 ohm load voice coil. 4 ohm loads will have a smaller peak and be just as flat at 20K. The zobel really knocks down the peak's interaction with the driver inductance and could use a smaller resistor for 4 ohm loads. Set the resistor by ear for the best trade off for any given speaker between making the sound too aggressive or too boring.
Attachments
My completed desktop amp
Hi All,
For the past several weeks I've been working on a compact desktop amplifier for my computer based on the Sure TK2050 2 channel board. Before I did too much, I'd read through every post here (150 pages at the time!) and I have to say the info here has been invaluable. I especially want to thank Scott for his detailed coil trials. I ended up going with the old-style Wurth 6.8uH, since I did not want to add a full Zobel and the 4.7 are no longer available in the US anyway. For the input cap, I went with the Janzten poly cross cap bypassed with a Dayton foil. For the limited space available in my application, this seemed like a good compromise.
I'm sure further gains are available for both sections if cost, labor and particularly space were not limits. While I wanted as good sound as possible, space and reliability were equally high priorities. In any case, many of my music files are compressed anyway (until I get around to re-ripping everything lossless). I do have a Total Bithead running as the DAC, which is a vast improvement over onboard sound. For the power supply, I went with the ubiquitous and cheap HP printer brick. That way, I can tuck it out of the way. Since it's at an already crowded workspace, I wanted everything to be as inconspicuous as possible. The amp itself is actually mounted on one of the desk's lower panels. Nothing is visible up top except the speakers (era D4s - great little speakers)!
I'm very happy with results. I think the sound is just about as good as possible given the source and small speakers. Maybe not a match for my Cary and Magnepans, but so good that they might not get used as much. Real hi-fi the size of a paperback book - I would not have thought it possible!
Hopefully some images are attached which will tell the rest of the build details better than I could in words. Perhaps it will be useful to someone else. I'm happy to answer any questions.
Anyway, just wanted to thank all the regular posters here for sharing their experiences and insights. I've been very, very impressed with the level of testing and experimentation as well as the helpful attitudes here. I can't say I'll do this again soon, but it was fun. Thanks.
-Mark
Hi All,
For the past several weeks I've been working on a compact desktop amplifier for my computer based on the Sure TK2050 2 channel board. Before I did too much, I'd read through every post here (150 pages at the time!) and I have to say the info here has been invaluable. I especially want to thank Scott for his detailed coil trials. I ended up going with the old-style Wurth 6.8uH, since I did not want to add a full Zobel and the 4.7 are no longer available in the US anyway. For the input cap, I went with the Janzten poly cross cap bypassed with a Dayton foil. For the limited space available in my application, this seemed like a good compromise.
I'm sure further gains are available for both sections if cost, labor and particularly space were not limits. While I wanted as good sound as possible, space and reliability were equally high priorities. In any case, many of my music files are compressed anyway (until I get around to re-ripping everything lossless). I do have a Total Bithead running as the DAC, which is a vast improvement over onboard sound. For the power supply, I went with the ubiquitous and cheap HP printer brick. That way, I can tuck it out of the way. Since it's at an already crowded workspace, I wanted everything to be as inconspicuous as possible. The amp itself is actually mounted on one of the desk's lower panels. Nothing is visible up top except the speakers (era D4s - great little speakers)!
I'm very happy with results. I think the sound is just about as good as possible given the source and small speakers. Maybe not a match for my Cary and Magnepans, but so good that they might not get used as much. Real hi-fi the size of a paperback book - I would not have thought it possible!
Hopefully some images are attached which will tell the rest of the build details better than I could in words. Perhaps it will be useful to someone else. I'm happy to answer any questions.
Anyway, just wanted to thank all the regular posters here for sharing their experiences and insights. I've been very, very impressed with the level of testing and experimentation as well as the helpful attitudes here. I can't say I'll do this again soon, but it was fun. Thanks.
-Mark
Attachments
Thanks. Not that warm, actually. About body heat (98 F) according to touch and my IR thermometer. I'm using the stock heatsink, btw, with the fan removed. So the top of the case is contacted by just the outer fins, but that's enough. I'd planned to replace it with a machined block of solid aluminum but that turned out to be more trouble than necessary. Before mating to the case it would run about 135, so the large surface of the case really helps dissipation.
I'd also meant to mention that the case is model AD01 from Asia Engineer on eBay.
-Mark
I'd also meant to mention that the case is model AD01 from Asia Engineer on eBay.
-Mark
Strange noise
That is a strange noise. Two boards, two power supplies, one channel of the source on one board for left, the other channel from the same source on the other board for right, results in only normal hiss with your ear to the tweeter.
.
Two boards, connected to the same terminals of one power supply, not even daisy chained through the boards, as soon as you connect the second source cable, a tone starts in the speakers louder than the hiss. Even if you lift the audio ground from one source cable. They observe the same problem on the web page for the Charlize. Very weird.
.
I didn't come up with a solution as it doesn't seem like anything simple like a star ground will take care of this. I wonder if you swap C19, 21, 22, 27 in one amp to 470p if the noise would move out of the audio band?
.
Good thing the Meanwell power supplies are cheap. I bought enough so that each board can have it's own when I start using 3-4 boards per side when I finally go active cross. The cost aside, it does start to get quite bulky.
That is a strange noise. Two boards, two power supplies, one channel of the source on one board for left, the other channel from the same source on the other board for right, results in only normal hiss with your ear to the tweeter.
.
Two boards, connected to the same terminals of one power supply, not even daisy chained through the boards, as soon as you connect the second source cable, a tone starts in the speakers louder than the hiss. Even if you lift the audio ground from one source cable. They observe the same problem on the web page for the Charlize. Very weird.
.
I didn't come up with a solution as it doesn't seem like anything simple like a star ground will take care of this. I wonder if you swap C19, 21, 22, 27 in one amp to 470p if the noise would move out of the audio band?
.
Good thing the Meanwell power supplies are cheap. I bought enough so that each board can have it's own when I start using 3-4 boards per side when I finally go active cross. The cost aside, it does start to get quite bulky.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.