Okay, I downloaded the photo and I'm going through the checklist
My apologies.
This will not be pretty.
However, I hope that it helps.
Checklist:
1 GND (0v)
Check
2 -IN (requires fixed voltage divider)
Feedback-shunt and coupler cap, check
Feedback resistor???? Not shown clearly, please double check that a feedback resistor exists between pin2 and pin14.
Note--negative input coupler cap was sized small enough to be a bass blocker--consider using a larger cap.
3 +IN (requires input load)
22k check
input coupler is too small--try 0.68u or 1u
4 Small Signal Ground
Check
5 Don't connect at all
Check
6 BootStrap (requires ~33u cap from pin6 to pin14)
Check--the 22u is slightly too small, but shouldn't cause failure.
7 V+ (low current) (requires 1 or 2 of 220u caps--220u or 440u max)
FAIL--the caps are ringingly too large and too far away.
The 1000u caps too large and too far away may cause excessive heat and poor sound quality, but typically causes sketchy behavior rather than instant explode.
8 V- (low current) (requires 1 or 2 of 220u caps--220u or 440u max)
FAIL--the caps are ringingly too large and too far away.
The 1000u caps too large and too far away may cause excessive heat and poor sound quality, but typically causes sketchy behavior rather than instant explode.
9 Standby (requires 22k delayed by 10u)
FAIL--0 ohms is not the right value--severe overcurrent may cause distortion and/or damage--inopportune startup sequence may cause damage
10 Mute (requires 10k to 22k resistor)
FAIL--0 ohms is not in the right range--severe overcurrent may cause distortion and/or damage
11 Don't connect at all
Check
12 Don't connect at all
FAIL--please don't connect anything to N.C. pins.
13 V+ (high current) (requires RF bypass 3n~100n)
FAIL--the 100nF RF bypass caps are much too far away
14 Speaker output (other end of feedback resistor connects here)
Fail?--Feedback resistor connection not shown.
15 V- (high current) (requires RF bypass 3n~100n)
FAIL--the 100nF RF bypass caps are much too far away
Tab (assure that a shoulder washer and insulating pad are in use)
not checked (check it yourself with an ohmmeter between the tab and heatsink)
Another way to say this, is that although your board covers most of the needs of something like an LM3875 gainclone, unfortunately, the TDA7294 is a different part so it has different needs as well. (TDA7294 is not a gainclone)
Power board not shown--be sure to use a real power board so that the umbilical cable power input to the amplifier boards is nice smooth reasonably clean DC power. (TDA7294 is not a gainclone)
Repair list:
Make sure the feedback resistor is actually hooked up (2 to 14)
Cut off pin 12 (N.C. actually means Don't Connect At All)
Put one or tandem 220u cap(s) closely at pin 7 to stabilize small signal power
Put one or tandem 220u cap(s) closely at pin 8 to stabilize small signal power
Put one 100n cap closely at pin 13 power, for stability (retard RF)
Put one 100n cap closely at pin 15 power, for stability (retard RF)
Install a range 10k to 22k resistor in series to the mute pin 10
Install a 22k and a 10u for the standby pin 9, according to the datasheet.
Use larger signal coupling caps for both +in coupler and -in coupler to avoid bass blocking
*If after making the repairs the IC chips continue to explode, suspect counterfeit chips and therefore follow the TDA7296 datasheet instead, most especially the lower voltage.
Gain divider settings:
Here are some stability adjusted figures that you can try for gain divider instead of the 22k vs 680R.
37+37vdc: 115k vs 2k7 (high gain for excellent stability when at high power)
35+35vdc: 100k vs 2k7
32+32vdc: 91k vs 2k7
27+27vdc: 82k vs 2k7
25+25vdc: 75k vs 2k7
23+23vdc: 68k vs 2k7
19+19vdc: 56k vs 2k7 (low gain for high end imaging in a low power amp)
*The typical 680R has been replaced with 2k7 so that you can use a 220u -in coupling cap, which is more convenient to install than a DC tracker.
**The feedback resistor value has been varied according to voltage in an attempt to stay within a favorable current range because this is not automated inside the chip at all; however, the above is only an estimate and may require some adjustment of the values for fine tuning.
My apologies.
This will not be pretty.
However, I hope that it helps.
Checklist:
1 GND (0v)
Check
2 -IN (requires fixed voltage divider)
Feedback-shunt and coupler cap, check
Feedback resistor???? Not shown clearly, please double check that a feedback resistor exists between pin2 and pin14.
Note--negative input coupler cap was sized small enough to be a bass blocker--consider using a larger cap.
3 +IN (requires input load)
22k check
input coupler is too small--try 0.68u or 1u
4 Small Signal Ground
Check
5 Don't connect at all
Check
6 BootStrap (requires ~33u cap from pin6 to pin14)
Check--the 22u is slightly too small, but shouldn't cause failure.
7 V+ (low current) (requires 1 or 2 of 220u caps--220u or 440u max)
FAIL--the caps are ringingly too large and too far away.
The 1000u caps too large and too far away may cause excessive heat and poor sound quality, but typically causes sketchy behavior rather than instant explode.
8 V- (low current) (requires 1 or 2 of 220u caps--220u or 440u max)
FAIL--the caps are ringingly too large and too far away.
The 1000u caps too large and too far away may cause excessive heat and poor sound quality, but typically causes sketchy behavior rather than instant explode.
9 Standby (requires 22k delayed by 10u)
FAIL--0 ohms is not the right value--severe overcurrent may cause distortion and/or damage--inopportune startup sequence may cause damage
10 Mute (requires 10k to 22k resistor)
FAIL--0 ohms is not in the right range--severe overcurrent may cause distortion and/or damage
11 Don't connect at all
Check
12 Don't connect at all
FAIL--please don't connect anything to N.C. pins.
13 V+ (high current) (requires RF bypass 3n~100n)
FAIL--the 100nF RF bypass caps are much too far away
14 Speaker output (other end of feedback resistor connects here)
Fail?--Feedback resistor connection not shown.
15 V- (high current) (requires RF bypass 3n~100n)
FAIL--the 100nF RF bypass caps are much too far away
Tab (assure that a shoulder washer and insulating pad are in use)
not checked (check it yourself with an ohmmeter between the tab and heatsink)
Another way to say this, is that although your board covers most of the needs of something like an LM3875 gainclone, unfortunately, the TDA7294 is a different part so it has different needs as well. (TDA7294 is not a gainclone)
Power board not shown--be sure to use a real power board so that the umbilical cable power input to the amplifier boards is nice smooth reasonably clean DC power. (TDA7294 is not a gainclone)
Repair list:
Make sure the feedback resistor is actually hooked up (2 to 14)
Cut off pin 12 (N.C. actually means Don't Connect At All)
Put one or tandem 220u cap(s) closely at pin 7 to stabilize small signal power
Put one or tandem 220u cap(s) closely at pin 8 to stabilize small signal power
Put one 100n cap closely at pin 13 power, for stability (retard RF)
Put one 100n cap closely at pin 15 power, for stability (retard RF)
Install a range 10k to 22k resistor in series to the mute pin 10
Install a 22k and a 10u for the standby pin 9, according to the datasheet.
Use larger signal coupling caps for both +in coupler and -in coupler to avoid bass blocking
*If after making the repairs the IC chips continue to explode, suspect counterfeit chips and therefore follow the TDA7296 datasheet instead, most especially the lower voltage.
Gain divider settings:
Here are some stability adjusted figures that you can try for gain divider instead of the 22k vs 680R.
37+37vdc: 115k vs 2k7 (high gain for excellent stability when at high power)
35+35vdc: 100k vs 2k7
32+32vdc: 91k vs 2k7
27+27vdc: 82k vs 2k7
25+25vdc: 75k vs 2k7
23+23vdc: 68k vs 2k7
19+19vdc: 56k vs 2k7 (low gain for high end imaging in a low power amp)
*The typical 680R has been replaced with 2k7 so that you can use a 220u -in coupling cap, which is more convenient to install than a DC tracker.
**The feedback resistor value has been varied according to voltage in an attempt to stay within a favorable current range because this is not automated inside the chip at all; however, the above is only an estimate and may require some adjustment of the values for fine tuning.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.