In theory it is possible to connect those ICs in parallel. In practice it takes well matched feedback circuits, if you don't want to run into trouble. It is not very likely that they are well matched, and since they are internal you cannot do anything about it. You may be better off connecting the 2 x 2 Ohm speakers in series, which gives you 4 Ohm.
The transistor solution is also not likely to work well, because the TDA1562 is internally a class H amplifier. You will need an additional (24 V?) power supply to get the corresponding higher supply voltage for the external transistors. The TDA1562 is also internally a bridged amplifier, which means you need double the amount of transistors as compared to a normal chipamp with transistors. In the end this will turn out to be pretty complicated and expensive.
The transistor solution is also not likely to work well, because the TDA1562 is internally a class H amplifier. You will need an additional (24 V?) power supply to get the corresponding higher supply voltage for the external transistors. The TDA1562 is also internally a bridged amplifier, which means you need double the amount of transistors as compared to a normal chipamp with transistors. In the end this will turn out to be pretty complicated and expensive.
Hi PacificBlue!
What do you think of LA4628 with 8 output transistors?
According to MCM, that chip is "often purchased with Sanken output transistors" however, other clues on this tempting prospect, were absent. How does one make such an amp?
I don't think much of adding transistors to power op amps. You need to add an external bias stage for the power transistors, although you already have one inside the IC and you don't make good use of the internal power transistors.
For designers who cannot or will not design their own driver stages, the industry offers driver ICs, e. g. TDA7250 or the LME49800 series.
The LA4628 is an internally bridged 20 W into 4 Ohm amplifier that runs from a 13,5 V single supply. You would need one pair of transistors and a bias stage for each output pin to get the two bridged output stages. The result would be increased output current and THD with decreased output voltage.
For designers who cannot or will not design their own driver stages, the industry offers driver ICs, e. g. TDA7250 or the LME49800 series.
The LA4628 is an internally bridged 20 W into 4 Ohm amplifier that runs from a 13,5 V single supply. You would need one pair of transistors and a bias stage for each output pin to get the two bridged output stages. The result would be increased output current and THD with decreased output voltage.
Wow man! That explanation as clear as a bell! Thanks!
Well, of course the output stage in this case would have to run higher voltage.
The effort to boost a chip amp is looking pretty silly because it requires the majority of a discrete amp to do it. lol! Oh, I get it now. You would normally expect more work for less results, except. . .
Personally, that wasn't my point. My reason for asking is that great sounding, ear-friendly, chips do exist, some are very clean, and some, like LM1875 come with a endearing performance that verges into areas we have only begun to measure. They have only one thing in common, and that is, insufficient output power. So, it seems that a good reason (maybe the only good reason) to use a power op amp as a pre-drive is that you happen to like its harmonic output. One would not expect the "booster" effort to make any more or any less audio degradation than adding a preamplifier to an audio system that doesn't require it.
Is this bit correct?
Yes, well, the gain is rather high. But are you saying that despite the inbuilt high gain, we'd also have to give it a shove with a preamp?
At that point, I wouldn't want to do it, because a simple JRC4560 could outperform it by far (without the circuit containing an extraneous power op-amp in the middle).
So, if i connect 2 or more parallel chips i will have trouble with feedback, if i connect external transistors, i need +-24V suply?
The preamp won't help. You need an additional VAS (voltage amplification stage) between the chipamp output and the power transistor input. Otherwise the output voltage cannot become higher than what comes out of the chipamps. The transistors alone only increase the ability to deliver current.
More or less. When you connect two amplifiers in parallel, their gains need to be matched tightly. E. g. the National datasheets recommend 0,1 % matched feedback resistors. Unmatched gains lead to different output voltage on each amp and that will lead to equalisation currents flowing from one IC to the other. Those currents heat up the chipamps and reduce the available current for the speaker.
For the external transistors you need a higher supply voltage, because you get a higher output voltage, when the circuit switches to class H operation. The TDA1562Q seems to use a built-in voltage doubler, so if you have a 12 V single supply for it, you will need a 24 V single supply for the transistors. And as described in post #8 you will need two transistors and a bias stage for each output pin. Then some kind of DC protection for the speakers in case one of the transistors blows. The TDA's built-in protection systems (short-circuit, overtemperature, overcurrent) will also not work on the transistors, so you may have to add those as well.
As for the TDA1562Q, its a "one trick" amp, and it has a good trick indeed. To increase power, add another speaker and another amp.
Or, you could use more efficient speakers--every 3db higher efficiency from the speaker is approximately representative of doubling the power of the amp.
Like this:
(wrong) 87db@1w speaker with your amp maxes at 106db*
-or this-
(better) 90db@1w speaker with your amp maxes at 109db*
-or this-
(prodigious) 93db@1w speaker with your amp maxes at 112db*
*TDA1652q's 70w rating requires a 4 ohm speaker, else subtract 3db from the maximum output figures above.
So, a 70w amp with a 93db@1w efficiency rated speaker does the the same decibel output as. . . a 140w amp with a 90db@1w efficiency rated speaker.
At this point, you might consider running your bass in stereo, which can provide greater average dynamics (than mono) and dynamics are so helpful in recovering the content of MP4, MP3, and other lossy sources that are in popular use today.
Or, you could use more efficient speakers--every 3db higher efficiency from the speaker is approximately representative of doubling the power of the amp.
Like this:
(wrong) 87db@1w speaker with your amp maxes at 106db*
-or this-
(better) 90db@1w speaker with your amp maxes at 109db*
-or this-
(prodigious) 93db@1w speaker with your amp maxes at 112db*
*TDA1652q's 70w rating requires a 4 ohm speaker, else subtract 3db from the maximum output figures above.
So, a 70w amp with a 93db@1w efficiency rated speaker does the the same decibel output as. . . a 140w amp with a 90db@1w efficiency rated speaker.
At this point, you might consider running your bass in stereo, which can provide greater average dynamics (than mono) and dynamics are so helpful in recovering the content of MP4, MP3, and other lossy sources that are in popular use today.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.