I got the replacement chip. The board works now but I'm getting a lot of oscillation. Must have something else wrong. I'm going to pull some to the components anyway and do the mods you have suggested. Maybe I'll find the issue along the way. I played some music through the other board in stock form and I was fairly impressed at the sound. Just using the +-25Vdc SMPS and it sounds clean with good bass. Pretty impressive for such a small amp.
Blessings, Terry
Blessings, Terry
I didn't read this whole thread, but I've been researching these switchmode amp chips for a while now, and I have a 2 cents to throw in.
If you replace the 220uF feedback cap with a 10uF polyprop cap, the low end will roll off at 6 HZ, and the PP cap may sound better than an electrolytic. Also I don't see any inductors in series with the output. They may not be "needed", but any switchmode circuit is likely to put out some Rf energy, so a good 2 pole output filter makes good sense to me.
Any circuit that deals with energy over roughly 200kHZ may benefit significantly from having a circuit board that has a ground plane on one side. Grounding technique and keeping all leads as short as possible would be very important, maybe the most important.
I'm not sure what that 33uF cap from output to the shield or heatsink is about. I would think that would be a disaster (?) Probably cause oscillations.
If you replace the 220uF feedback cap with a 10uF polyprop cap, the low end will roll off at 6 HZ, and the PP cap may sound better than an electrolytic. Also I don't see any inductors in series with the output. They may not be "needed", but any switchmode circuit is likely to put out some Rf energy, so a good 2 pole output filter makes good sense to me.
Any circuit that deals with energy over roughly 200kHZ may benefit significantly from having a circuit board that has a ground plane on one side. Grounding technique and keeping all leads as short as possible would be very important, maybe the most important.
I'm not sure what that 33uF cap from output to the shield or heatsink is about. I would think that would be a disaster (?) Probably cause oscillations.
The "shield" is the chip package, and the 33uF cap is going to pin 6. These aren't switchmode amps either.
I like the film feedback cap idea, but some say MKP causes hard sound. That is strange to me, but 10u MKP is also expensive. Not by much though once you've paid for the whole construction.
I like the film feedback cap idea, but some say MKP causes hard sound. That is strange to me, but 10u MKP is also expensive. Not by much though once you've paid for the whole construction.
Try this link instead: 
Why don't you try the board available at ebay at $6 + shipping. Works great and sounds great. You will have to make sure that you have the uPC 1237 protection chip. If you skip that you could use an external relay protection circuit. I've had several occasions when the protection tripped as I use a tube cathode follower at the input. That causes a dc surge as it comes on slowly. Just have to cycle the power switch again.
The amp seems to get better after time. I use a 275V 1uF polyester cap at the input. Didn't think it would be too good but it's amazing ! I do think it sounds better now after at least 50 hours of use. Was it the tube stage or the whole amp itself that "burned in" I can't say. But I am really happy I bought this board. I've bought some more for my experimental active system amp rack.
1pc TDA7294 TDA7293 Amplifier Amp DIY Board PCB | eBay
The amp seems to get better after time. I use a 275V 1uF polyester cap at the input. Didn't think it would be too good but it's amazing ! I do think it sounds better now after at least 50 hours of use. Was it the tube stage or the whole amp itself that "burned in" I can't say. But I am really happy I bought this board. I've bought some more for my experimental active system amp rack.
1pc TDA7294 TDA7293 Amplifier Amp DIY Board PCB | eBay
Yes, no matter which one of the TDA7293/4/5/6 chips it is, the lower voltage plus stabilized voltage will greatly improve the linearity.
It is like safety derating, but even nicer.
Also, it will change the ideal gain factor so that it will not match the examples posted (except for possibly TDA7296).
In combination, the news is that it will be much easier to dial in (a nice surprise) and that there will be some manual fine tuning steps for discovery of the relocated optimums (not a surprise).
Some observed and probable tips:
If you can find a 3k multi-turn cermet pot and rig it as a variable resistor (feedback-shunt resistor--the groundside partner of the feedback resistor), between pin2 and the nfb-shunt coupling cap + (use the 3K multi-turn cermet variable resistor instead of my 2k7), then not only can you dial in the gain, but you can fine tune it more suitably than a fixed resistor can do. The ideal is most usually an oddball value. It is okay to leave the multi-turn cermet trimpot in place after adjusting because cermet is a very fine quality material.
Somewhere earlier in this thread, I tried to find the optimums without using a variable resistor, and that really took a lot of time. I wish I'd used the trimpot. At end result I had to employ a trimpot anyway, because the "ideal" is an extremely narrow spot and varies by power voltage--very favorably at lower voltages (greater linearity is worth more than a lotta watts), so you might not have to go through as much effort as I did at higher voltage.
Hello
I'll try something like that
http://oi57.tinypic.com/33cabux.jpg
1) no charge at the output Master (R9 need adjustment) juste used as V amp
2) Separate power supply between the Master and slave ( may be statilised)
3) Minimum gain applied to the TDA (stable over 15db), complement with the OPA.
4) no cap on feedback
I'll try something like that
http://oi57.tinypic.com/33cabux.jpg
1) no charge at the output Master (R9 need adjustment) juste used as V amp
2) Separate power supply between the Master and slave ( may be statilised)
3) Minimum gain applied to the TDA (stable over 15db), complement with the OPA.
4) no cap on feedback
Last edited:
Digora,
Chip power op amps may handle low gain when in inverting mode; however, they don't handle low gain in non-inverting mode.
For non-inverting TDA7293, here's estimated gain factor related to voltage
(less gain may exhibit either minor or major instability)
36+36vdc, 38X
32+32vdc, 34X
30+30vdc, 32X
28+28vdc, 30X
26+26vdc, 28X
24+24vdc, 26X
22+22vdc, 24X
20+20vdc, 22X
For non-inverting TDA7294, here's estimated gain factor related to voltage
(less gain may exhibit either minor or major instability)
36+36vdc, 42X
32+32vdc, 38X
30+30vdc, 35X
28+28vdc, 33X
26+26vdc, 30X
24+24vdc, 27X
22+22vdc, 25X
20+20vdc, 23X
18+18vdc, 22X
Notes:
Gain, is the primary stability control for a chip amplifier.
The gain setting has a Dramatic effect on stability and therefore a very noticeable effect on tone as well.
The above figures are ballpark estimates; however, The ideal setting is within a very narrow range so that a trimmer is useful at feedback-shunt locale.
It may be useful to adapt the gain divider's current to avoid hindering linearity of microscopic small signal input transistors.
Choose Non-inverting mode if you want high gain (as listed above) OR choose inverting mode if you want low gain (unlike what is listed above).
Chip power op amps may handle low gain when in inverting mode; however, they don't handle low gain in non-inverting mode.
For non-inverting TDA7293, here's estimated gain factor related to voltage
(less gain may exhibit either minor or major instability)
36+36vdc, 38X
32+32vdc, 34X
30+30vdc, 32X
28+28vdc, 30X
26+26vdc, 28X
24+24vdc, 26X
22+22vdc, 24X
20+20vdc, 22X
For non-inverting TDA7294, here's estimated gain factor related to voltage
(less gain may exhibit either minor or major instability)
36+36vdc, 42X
32+32vdc, 38X
30+30vdc, 35X
28+28vdc, 33X
26+26vdc, 30X
24+24vdc, 27X
22+22vdc, 25X
20+20vdc, 23X
18+18vdc, 22X
Notes:
Gain, is the primary stability control for a chip amplifier.
The gain setting has a Dramatic effect on stability and therefore a very noticeable effect on tone as well.
The above figures are ballpark estimates; however, The ideal setting is within a very narrow range so that a trimmer is useful at feedback-shunt locale.
It may be useful to adapt the gain divider's current to avoid hindering linearity of microscopic small signal input transistors.
Choose Non-inverting mode if you want high gain (as listed above) OR choose inverting mode if you want low gain (unlike what is listed above).
thank you for your explanations, I'll check it possible to use theTDA in inverter
However, on the gain and stability, look at this document:
http://jimmy.thomas.free.fr/temp/TDA7293/TDA7293-MESURES.pdf
edit :
In fact what you are trying to make me understand is that it is useless to try the inverter because anyway the TDA does not support a low gain?
However, on the gain and stability, look at this document:
http://jimmy.thomas.free.fr/temp/TDA7293/TDA7293-MESURES.pdf
edit :
In fact what you are trying to make me understand is that it is useless to try the inverter because anyway the TDA does not support a low gain?
Last edited:
Well, no I did not say that. It is doable. However, it is not the easiest way.
In fact, the easiest way is to 1) build one normal tda7293, 2) after that, then modify that same amplifier to slave mode to create a power buffer with no gain, 3) push it with an LME driver chip.
http://www.diyaudio.com/forums/atta...0-optimizing-tda7294-output-tda7293-hi-fi.gif
Hi Daniel,
what do you recommend as other changes in this diagram ?
Mute and STBY are optimal ?
In fact, what is the best TDA7293 diagram uncompromising ?
Thanks
I found B&W DM605 service manual using TDA7294 as inverting mode with only 4X GAIN !
BW DM605 S2 Service Manual free download, schematics, eeprom, repair info for electronics
Am I wrong?
BW DM605 S2 Service Manual free download, schematics, eeprom, repair info for electronics
Am I wrong?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.