Does anyone have that graph?
ST did not include it in their datasheet.
For example, here is the graph for LM3886:
ST did not include it in their datasheet.
For example, here is the graph for LM3886:
Last edited:
Given the data they provide and the huge number of amps and manufacturers that successfully used the tda7293/94 it seems that diy'ers opinion don't really matter.I built a 7294 amp 18 years ago and it worked for about 3 years before I sold it making a lot of people happy...ST is a really big and trusty manufacturer.They don't sell sh*t and they don't need audiophile auras around their heads to be trusted.
In my humble experience there's always a relevant graph missing in any datasheet.
In my humble experience there's always a relevant graph missing in any datasheet.
Last edited:
I am interested in that chip mostly because of the shortages of LM3886. Another reason could be because of JFET output transistors - that supposed to sound "warmer" whatever that means.
Yeah, getting measurements for each given TDA7293 sample is not the most efficient way to produce big quantities of boards. Otherwise what is the guarantee that one out of 10 or even 100 is not going to give proper result? I must agree on that point.
Yeah, getting measurements for each given TDA7293 sample is not the most efficient way to produce big quantities of boards. Otherwise what is the guarantee that one out of 10 or even 100 is not going to give proper result? I must agree on that point.
DMOS, MOSFET, not JFET. And "warmer" is more about implementation than device.
Spend $8-$12 and measure it yourself? (Subject to tomchr's objections about single-sample error.)
You buy one and measure it.
The easiest way to measure AVOL is to configure the chip as an inverting amplifier. Then measure the AC voltage between its inputs as function of frequency.
Personally I would use a network analyzer for this or (maybe) an audio analyzer, but nothing stops you from measuring with an AC voltmeter, though you need one with fairly high bandwidth, which usually means a bench top meter. Or you could use an oscilloscope. The drawback of using the o'scope is that they tend to have pretty low sensitivity. The better ones go down to 1-2 mV/div or so. You could work around that by configuring the amp for some gain. If the amp has a gain of 40 dB you shave off 40 dB from the loop gain so the voltage between the amp's inputs gets 40 dB higher. Also, it's usually only maybe 1-2 decades before and about one decade after the UGBW that are relevant for stability. You should find enough signal there to pick it up with a scope.
Measuring the error voltage (Ve = Vin+ - Vin-) at an inverting gain of -1 V/V as described above will give you 1/a, where a is the loop gain. If you plot the measured voltage in dB relative to the input voltage all you need to do is to drop the minus sign (or flip the curve on the horizontal axis graphically) to get the AVOL.
You can also follow the approach here (fig 5): https://www.analog.com/en/analog-dialogue/articles/simple-op-amp-measurements.html
Tom
The easiest way to measure AVOL is to configure the chip as an inverting amplifier. Then measure the AC voltage between its inputs as function of frequency.
Personally I would use a network analyzer for this or (maybe) an audio analyzer, but nothing stops you from measuring with an AC voltmeter, though you need one with fairly high bandwidth, which usually means a bench top meter. Or you could use an oscilloscope. The drawback of using the o'scope is that they tend to have pretty low sensitivity. The better ones go down to 1-2 mV/div or so. You could work around that by configuring the amp for some gain. If the amp has a gain of 40 dB you shave off 40 dB from the loop gain so the voltage between the amp's inputs gets 40 dB higher. Also, it's usually only maybe 1-2 decades before and about one decade after the UGBW that are relevant for stability. You should find enough signal there to pick it up with a scope.
Measuring the error voltage (Ve = Vin+ - Vin-) at an inverting gain of -1 V/V as described above will give you 1/a, where a is the loop gain. If you plot the measured voltage in dB relative to the input voltage all you need to do is to drop the minus sign (or flip the curve on the horizontal axis graphically) to get the AVOL.
You can also follow the approach here (fig 5): https://www.analog.com/en/analog-dialogue/articles/simple-op-amp-measurements.html
That hasn't been my experience with TI and ADI parts. A power opamp without an AVOL curve in the data sheet is quite useless as you have no idea what the stability will be like. But whatever. Lack of date doesn't have to be a show stopper. It's just one of those things that make me avoid a chip.
Tom
I doubt ST was ever worried about a private person's opinion on a chip they sold by the millions to many big companies...
I highly doubt that as well. 🙂 They should be worried about those looking to buy 100 kU/year.
One could argue that they should just add the curve to the data sheet, but that's actually a major pain after the part has been released. Sometimes that'll trigger a re-qualification round with the end customer. That gets expensive in a hurry, hence, is a rather big deal. Then suddenly the customer becomes more motivated to shift their 100 kU/year to a different manufacturer... 🙂
Tom
One could argue that they should just add the curve to the data sheet, but that's actually a major pain after the part has been released. Sometimes that'll trigger a re-qualification round with the end customer. That gets expensive in a hurry, hence, is a rather big deal. Then suddenly the customer becomes more motivated to shift their 100 kU/year to a different manufacturer... 🙂
Tom
They probably don't care at this moment...all they sell as tda chips might have been produces a decade ago...their gross production is probably oriented on class d chips , microcontrollers , dsp's , motor drivers and regulators...
I would think that current production prioritizes anything automotive qualified. That'll remain the case until the current supply chain stuff flushes out.
Tom
Tom