Regarding Cmoraiu's improved amplifier schematic: http://www.diyaudio.com/forums/chip...ors-amp-tda7293-come-also-48.html#post4359289
I wonder, would it work the same way if you just omit the extra AB stage with the TIP41/42, and bias the main output transistors directly with the same bias string, in order to remove the (miniscule) crossover distrortion?
I don't have a simulation program, but i really want to try this the hardware way!
I wonder, would it work the same way if you just omit the extra AB stage with the TIP41/42, and bias the main output transistors directly with the same bias string, in order to remove the (miniscule) crossover distrortion?
I don't have a simulation program, but i really want to try this the hardware way!
#filthyone I'm sorry to disappoint you but it won't work. The TIP stage is designed only to drive the output around the crossover. Its bias loses stamina at high signal and isn't able to provide enough base current for final stage transistors.
#Moze I didn't make measurements for efficiency but having an additional AB stage it loses somehow in efficiency to increase linearity. Lowering distortions comes at a cost!
#Moze I didn't make measurements for efficiency but having an additional AB stage it loses somehow in efficiency to increase linearity. Lowering distortions comes at a cost!
Thanks for your answer cmorariu, but...
i just happened to find this http://www.diyaudio.com/forums/lounge/289869-stereo-chipamp-channel-separation.html#post4681162 which is exactly what i was thinking about and seems to do the job. Please, can you elaborate on why this wouldn't work as expected, because i didn't quite understand your last message.
I'm VERY interested in doing this as a mod to my gainclone, with two output pairs per channel under a beefy +-40v, regulated supply, for a stable 4ohm operation at about 200 watts.
Thanks again for any input!
i just happened to find this http://www.diyaudio.com/forums/lounge/289869-stereo-chipamp-channel-separation.html#post4681162 which is exactly what i was thinking about and seems to do the job. Please, can you elaborate on why this wouldn't work as expected, because i didn't quite understand your last message.
I'm VERY interested in doing this as a mod to my gainclone, with two output pairs per channel under a beefy +-40v, regulated supply, for a stable 4ohm operation at about 200 watts.
Thanks again for any input!
Hello
I know that the tda7294 can have up to 50-0-50 max supply voltage, I want to have 100 watt output on 8 ohm, what mods I need to do ?
Thank
Bye
Gaetan
I know that the tda7294 can have up to 50-0-50 max supply voltage, I want to have 100 watt output on 8 ohm, what mods I need to do ?
Thank
Bye
Gaetan
#gaetan It is not advisable to feed it with maximum voltage, it will blow up for sure. Theoretically +/-40V gives 100Wrms on 8ohm. To avoid clip distortions you need to feed the enhanced amplifier by +/-43V. Note that the bootstrap capacitor is connected to output pin of TDA7294 not as in my schematic.
#zanden30 We are talking about different things. I was talking about TDA + transistors amplifier, enhanced schematic as I posted on july 1, 2015. It easy can drive 200Wrms on low impedance loads.
Math is simple as long as the circuit can drive the necessary current.
For +/-40V the Vrms is 40/sqrt(2) = 28.28V. P=Vrms**2 / R = 100 Wrms, for R=8ohm.
Add +/-3V to be sure no clipping occurs.
Both TDA7293 or TDA7294 can be used to build the enhanced amplifier. The only difference in respect to this schematic is the connection of bootstrap capacitor. 7293 needs it between pins 6 and 12. 7294 between 6 and 14.
Math is simple as long as the circuit can drive the necessary current.
For +/-40V the Vrms is 40/sqrt(2) = 28.28V. P=Vrms**2 / R = 100 Wrms, for R=8ohm.
Add +/-3V to be sure no clipping occurs.
Both TDA7293 or TDA7294 can be used to build the enhanced amplifier. The only difference in respect to this schematic is the connection of bootstrap capacitor. 7293 needs it between pins 6 and 12. 7294 between 6 and 14.
That calc is not right mr. cmorariu.
The max amplitude of sinus = 33V (with Vcc = 35V).
P block/music = 33 x 33 /8 = 136 W
P RMS = 1/2 x 136 = 68 W.
This is NS specs calculation.
if you go to the limit Vcc = 40 V (I would not advise to go to 47V !!!) than the power is
P RMS = 38 x 38 / 8 = 90,25 W.
100 W in 8 Ohm is not possible in a safe way. Not due to the current limitations (the boost transistor will not help you in this case), but because of the voltage limitation. That's why BTL is the solution.
With BTL at 100 W you need I max = 5A.
P = 5 x 5 x 8 = 200 W (block)
P RMS = 1/2 x 200 = 100W. Thus the booster is not necessary. Although it might be better for cooling/dissipation reasons....
A Vcc of 22V is sufficient in this case! Choose 25V for some tolerance.
The max amplitude of sinus = 33V (with Vcc = 35V).
P block/music = 33 x 33 /8 = 136 W
P RMS = 1/2 x 136 = 68 W.
This is NS specs calculation.
if you go to the limit Vcc = 40 V (I would not advise to go to 47V !!!) than the power is
P RMS = 38 x 38 / 8 = 90,25 W.
100 W in 8 Ohm is not possible in a safe way. Not due to the current limitations (the boost transistor will not help you in this case), but because of the voltage limitation. That's why BTL is the solution.
With BTL at 100 W you need I max = 5A.
P = 5 x 5 x 8 = 200 W (block)
P RMS = 1/2 x 200 = 100W. Thus the booster is not necessary. Although it might be better for cooling/dissipation reasons....
A Vcc of 22V is sufficient in this case! Choose 25V for some tolerance.
Last edited:
#zanden30 The calc is perfectly right and you did it in the same manner. The only difference is that you assumed a +/-40V given power supply, and assumed that at +/-38V no clipping occurs.
For +/-38V of course the RMS power is 90,25W as you computed.
I computed starting from the requirements of #gaetan8888 who needs an output power of 100W.
Using same formula as you did, a +/-40V peak to peak sinusoidal on output will deliver 100Wrms on 8ohm. To be sure no significant clipping occurs I added 3V (as you deducted 2).
Discussion remains if TDA7294 will stand +/-43V. I would say yes, if one original chip, last revision, is used. I made tests feeding the enhanced amplifier with +/-45V and it worked just fine.
For +/-38V of course the RMS power is 90,25W as you computed.
I computed starting from the requirements of #gaetan8888 who needs an output power of 100W.
Using same formula as you did, a +/-40V peak to peak sinusoidal on output will deliver 100Wrms on 8ohm. To be sure no significant clipping occurs I added 3V (as you deducted 2).
Discussion remains if TDA7294 will stand +/-43V. I would say yes, if one original chip, last revision, is used. I made tests feeding the enhanced amplifier with +/-45V and it worked just fine.
ST datasheet for TDA7293 states under "Absolute maximum ratrs" that supply voltage must not exceed +/-60V in absence of signal. Under "Electrical characteristics" supply range is supposed to be between +/-12V and +/-50V. So +/- 43V is far from limit.
But, allways is a "but", definitely the chip heats more if the voltage increases. According to same datasheet the thermal protection, TDA7293 mutes at 150oC and shuts down at 160oC. And this is not radiators temperature but inside chip. So to have the amp working at full load without interruptions a rather big radiator is needed, better a ventilated one.
But, allways is a "but", definitely the chip heats more if the voltage increases. According to same datasheet the thermal protection, TDA7293 mutes at 150oC and shuts down at 160oC. And this is not radiators temperature but inside chip. So to have the amp working at full load without interruptions a rather big radiator is needed, better a ventilated one.
Assuming the supply is at ±43Vdc when delivering maximum power and you get a sinewave output of 40Vpk, then you are indeed getting 100W into 8r0.
Now turn down the input and the voltage from the supply rises.
Expect it to rise by at least another 3Volts, but more likely 5V to 8V of rise.
Let's say you are now showing 43+5V at the power pins.
Now apply a mains voltage variation of +6% (some countries have +10% as their upper limit).
The supply voltage goes up even more, probably another 3V to 4V depending on how you built it.
You now have 43+5+3.5V = ±51.5Vdc at the power pins with a small audio signal at the input.
Now turn down the input and the voltage from the supply rises.
Expect it to rise by at least another 3Volts, but more likely 5V to 8V of rise.
Let's say you are now showing 43+5V at the power pins.
Now apply a mains voltage variation of +6% (some countries have +10% as their upper limit).
The supply voltage goes up even more, probably another 3V to 4V depending on how you built it.
You now have 43+5+3.5V = ±51.5Vdc at the power pins with a small audio signal at the input.
Last edited:
It's obvious, to avoid the worst scenario don't plug it in on Fridays, 13.
Let's go backwards, suppose that 50V is for 10% more mains supply which is very improbable.
For the regular mains voltage it will be 45.45V. Under load it will go down to 40.45 if the transformer is not so strong.
According to TDA7293 datasheet at +/-40V you will get 100W on 8 ohm at 10% distortion.
According to same datasheet "Continuous output power" at the same time you will get 1% distortion at 80W. If safety is the main reason stick with +/-40V under maximum load and live with the distortions at full load.
Let's go backwards, suppose that 50V is for 10% more mains supply which is very improbable.
For the regular mains voltage it will be 45.45V. Under load it will go down to 40.45 if the transformer is not so strong.
According to TDA7293 datasheet at +/-40V you will get 100W on 8 ohm at 10% distortion.
According to same datasheet "Continuous output power" at the same time you will get 1% distortion at 80W. If safety is the main reason stick with +/-40V under maximum load and live with the distortions at full load.
Just to continu on an old post that I found some extra info for:
I found a schematic to the link I posted above. I had never seen it before, but it generally takes 10+ minutes to load (!!!), so maybe I never took the time to wait. I am assuming sangram DID see it and commented on it. It seems the TDA7294 is only driving the transistors and not sharing the load (no "sync" resistor) like in dr frost dk's original first post of this thread.
I know too little about amplifiers to comment on the implementation this way. I'll stick with dr frost's postings.
Link to bare PCB
Link to complete board
I found a schematic to the link I posted above. I had never seen it before, but it generally takes 10+ minutes to load (!!!), so maybe I never took the time to wait. I am assuming sangram DID see it and commented on it. It seems the TDA7294 is only driving the transistors and not sharing the load (no "sync" resistor) like in dr frost dk's original first post of this thread.
I know too little about amplifiers to comment on the implementation this way. I'll stick with dr frost's postings.
Link to bare PCB
Link to complete board
I've been very lazy about this build, primarily because it's a lot of work (active crossover, OB correction and mating multiple amps to a single power supply is never fun). My PCBs have arrived a few weeks ago, but I've not yet got around to testing them because of nothing in particular.
I have a different application in mind, 2 ohm driving capability with very low power supply voltage for highly sensitive paralleled 12" woofers. I'm not particularly comfortable with the TDA alone, and some help is probably needed even though I'm probably not going to cross 20V on the rails.
The output devices in this application are switched by current through the power stage. I'm using the feature of rail separation, leaving the driver side connected to the main rails and using series droppers in the power side rails to switch the devices when excess power is drawn. The dropper size determines at what current the transistors begin to contribute to the load. I say switch but in reality most transistors have a 'knee', which suits this operation mode really well.
I need to test a board pronto, maybe will solder up one soon.
I have a different application in mind, 2 ohm driving capability with very low power supply voltage for highly sensitive paralleled 12" woofers. I'm not particularly comfortable with the TDA alone, and some help is probably needed even though I'm probably not going to cross 20V on the rails.
The output devices in this application are switched by current through the power stage. I'm using the feature of rail separation, leaving the driver side connected to the main rails and using series droppers in the power side rails to switch the devices when excess power is drawn. The dropper size determines at what current the transistors begin to contribute to the load. I say switch but in reality most transistors have a 'knee', which suits this operation mode really well.
I need to test a board pronto, maybe will solder up one soon.