I'm using a 225va 18-0-18 toroid, and silver plated teflon wire. I'm using the BC 1% resistors from digikey and 1200uF HFQ panasonic caps (predecessor of FA I think). Should be decent 
Oh, I doubt I need input caps (yes no?) since I'm taking the output of the project 88 preamp...which has 2 paralleled 1uF solens per channel. No extra caps needed? Thanks!
-Matthew K. Olson
pics comin soon
Oh, I doubt I need input caps (yes no?) since I'm taking the output of the project 88 preamp...which has 2 paralleled 1uF solens per channel. No extra caps needed? Thanks!-Matthew K. Olson
pics comin soon
I still have this heat problem on the chip, with low volume
Can someone help me? please
That's my system:
2*18V torroid
2 1000V 10A silicon diodes bridge
one 2200µF cap directly to the bridge, per rail (+ to 0, and - to 0)
a 0.1µF cap directly on the chip, on each rail
10K input pot
10K input resistor
220K feedback resistor
1 TDA7294
1 tiny heatsink (3*5*1cm)
and a crappy loudspeaker
Can someone help me? please
That's my system:
2*18V torroid
2 1000V 10A silicon diodes bridge
one 2200µF cap directly to the bridge, per rail (+ to 0, and - to 0)
a 0.1µF cap directly on the chip, on each rail
10K input pot
10K input resistor
220K feedback resistor
1 TDA7294
1 tiny heatsink (3*5*1cm)
and a crappy loudspeaker
Bricolo,
Start by trying a zobel, or an output resistor (.22 ohm) or an output resistor in parallel with an inductor. People have been telling you that you are probably oscillating, so try to solve it. If one of the classic solutions works, THEN try to optimize.
Just because some people have built amps that are stable without any output networks doesn't mean the design is unconditionally stable in all environments, with all loads.
It is possible that it is not the load that is causing oscillation but rather HF noise coupled by the cable into the amp. Is the circuit stable into 10R using the same cable as the oscillating loudspeaker load?
Start by trying a zobel, or an output resistor (.22 ohm) or an output resistor in parallel with an inductor. People have been telling you that you are probably oscillating, so try to solve it. If one of the classic solutions works, THEN try to optimize.
Just because some people have built amps that are stable without any output networks doesn't mean the design is unconditionally stable in all environments, with all loads.
It is possible that it is not the load that is causing oscillation but rather HF noise coupled by the cable into the amp. Is the circuit stable into 10R using the same cable as the oscillating loudspeaker load?
tiroth said:
what do I nead for a zobel?
Right now, I'm trying with a 10R in series with the loudspeaker, the chip is still hot, can't leave my hand on it more than 5 seconds, but it seems colder than before
Some people told me it was oscillating, but I can't verify it (no scope) and I don't know how to fix sur a problem
Bricolo,
You might try going to a larger pot on your input. Try a 25k or a 50k log pot.
Check your grounds for good connection to the single star ground and make sure your output isn't shorting to ground somewhere like at the binding post. Other than the heat and and a hum you still have music coming out of the speaker, yes? How sure of the internals of the test speaker are you?
You might try going to a larger pot on your input. Try a 25k or a 50k log pot.
Check your grounds for good connection to the single star ground and make sure your output isn't shorting to ground somewhere like at the binding post. Other than the heat and and a hum you still have music coming out of the speaker, yes? How sure of the internals of the test speaker are you?
Philo said:
I'll test with a 100k in 5 minutes, but I don't think this will reduce the oscillation
my grounds are ok, all are going to the star point. Oh, just a thing: my chip has 3 pins connected to ground (gnd, +in, and mute) they are connected together with a wire, and I use a single wire to connect this to the star point. Is this correct, or must I use a wire to star ground, per pin?
Yes, I have music, quite good one! Quality is ok, power too. Hum isn't noticeable with my computer running.
The speaker is one from my low end stereo (an akai, 60w 6ohm, 3 way)
Bricolo said:I still have this heat problem on the chip, with low volume
Can someone help me? please
That's my system:
2*18V torroid
2 1000V 10A silicon diodes bridge
one 2200µF cap directly to the bridge, per rail (+ to 0, and - to 0)
a 0.1µF cap directly on the chip, on each rail
10K input pot
10K input resistor
220K feedback resistor
1 TDA7294
1 tiny heatsink (3*5*1cm)
and a crappy loudspeaker
Hi,
you must connect C1,2,5 (according to the original pdf) for proper work of TDA7294.( you didnt list these caps). For first time heatsink is enough, but I think that TDA7294 have gain depend of temperature. (say bigger heatsink, better work)
I was construct several amps with TDA7294, and all work immediately (with original schematic). Try this first. IMHO
Regards
I concur with moamps, about the gain. The datasheet is pretty sketchy about the gain but I think your feedback resistor might to large, thus raising your gain to high. And you mught have and impedance match problem by omiting the resistor between the inputs. Try going to a 22k for both like in the datasheet.
bricolo,
you are using the TDA7294 ? I am not sure...
This is an intirely different circuit to the LM3875.
Even if you use it in inverting mode (it seems not to be designed for), you must connect it differently:
You probably forgot the bootstrap cap (22yF) from pin 6 to pin 14 (output).
This chip has FOUR supply rails (pin 7 +Vs and 13 +PWVs, pin 8 -Vs and 15 -PWS) which all must be bypassed with caps individually even if you connect plus and minus supplies together. I would recommend the bigger caps to close the power and the smaller ones close the other pins.
Pin 1 is the standby ground, connected individually to ground. Pin 9 is standby, if you are not using it,leave it open.
Pin 4 is the second non-inverting input, typically connected to ground for mute, switched in by mute at pin 10, if you are not using it, leave it open.
I strongly recommend also a bigger heatsink and for the inverting setup the value of the input pot is too low.
Please check the datasheet carefully. The TDA7294 can be very stabile, but only with a clear layout.
Klaus
ps: You guys write fast...
you are using the TDA7294 ? I am not sure...
This is an intirely different circuit to the LM3875.
Even if you use it in inverting mode (it seems not to be designed for), you must connect it differently:
You probably forgot the bootstrap cap (22yF) from pin 6 to pin 14 (output).
This chip has FOUR supply rails (pin 7 +Vs and 13 +PWVs, pin 8 -Vs and 15 -PWS) which all must be bypassed with caps individually even if you connect plus and minus supplies together. I would recommend the bigger caps to close the power and the smaller ones close the other pins.
Pin 1 is the standby ground, connected individually to ground. Pin 9 is standby, if you are not using it,leave it open.
Pin 4 is the second non-inverting input, typically connected to ground for mute, switched in by mute at pin 10, if you are not using it, leave it open.
I strongly recommend also a bigger heatsink and for the inverting setup the value of the input pot is too low.
Please check the datasheet carefully. The TDA7294 can be very stabile, but only with a clear layout.
Klaus
ps: You guys write fast...
Just Speculating.
Bricolo,
Try the simple change I posted. There is the possibility that connecting the Vin+ with the power ground over such a "long run" allows surge power currents acting as signals to cause the amplifier to begin to oscillate. This is also addressed in the data sheets for most of these IC's.
Bricolo,
Try the simple change I posted. There is the possibility that connecting the Vin+ with the power ground over such a "long run" allows surge power currents acting as signals to cause the amplifier to begin to oscillate. This is also addressed in the data sheets for most of these IC's.
I'm using Peter's minimised gainclone shematic, adapted to a 7294:
+in to ground (pin 3)
in+ mute to ground (pin 4)
stdby gnd to ground (pin 1)
+Vs and +PWVs connected together, to bridge's +, and bypassed with a 0.1µF on the chip, and 2200µF on the bridge
the same for -Vs and -PWVs
pin 6 (bootstrap) is connected to out (pin 14) with a 22µF cap
-in is connected to out with a 220k, and to the input pot's whiper with 10k
mute and stdby are connected to the bridge's +, with resistors (according to the datasheet)
the bridge's + and - are to 24.5V DC
and with 0.15V AC ripple (measures with my multimeter)
+in to ground (pin 3)
in+ mute to ground (pin 4)
stdby gnd to ground (pin 1)
+Vs and +PWVs connected together, to bridge's +, and bypassed with a 0.1µF on the chip, and 2200µF on the bridge
the same for -Vs and -PWVs
pin 6 (bootstrap) is connected to out (pin 14) with a 22µF cap
-in is connected to out with a 220k, and to the input pot's whiper with 10k
mute and stdby are connected to the bridge's +, with resistors (according to the datasheet)
the bridge's + and - are to 24.5V DC
and with 0.15V AC ripple (measures with my multimeter)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.