It seems that if you bias up the output transistors of an ab amp, then until some output power the amp is operating in class A. To me, most of my critical listening probably all happens below 100mW max(besides my bass amp, I'm running a minidsp) so is there actually any difference between my class ab amp running at low power compared to a class A amp? (assuming similar build and design quality)
And also is there an easy way to tell at what powe output the amp transitions from class A to class AB? My full ranges are running on a Kenwood KA-7300 and my bass on a McIntosh MC2205.
And also is there an easy way to tell at what powe output the amp transitions from class A to class AB? My full ranges are running on a Kenwood KA-7300 and my bass on a McIntosh MC2205.
Yes, there is a difference. For an example with Mosfet PP followers, see my
article "Leaving Class A" on FIRST WATT
😎
article "Leaving Class A" on FIRST WATT
😎
Nelson,
Thank you for posting a link to that article, it was very informative. I always love reading the articles you have posted.
I see that the higher bias point of class A and the single ended topology avoids the mismatch of output transistors, and together those both reduce the distortion at all power levels.
1 side note question I have is do you have any blind tests that have been done with your amps comparing a good single ended class A to a good higher power class AB amp with a lower bias point to see if the distortion reduction is audible? I find it hard to find any blind tests or ABX tests that compare a good class AB to a good class A, and don't have the resources personally to run any kind of test of the sort.
Thank you for posting a link to that article, it was very informative. I always love reading the articles you have posted.
I see that the higher bias point of class A and the single ended topology avoids the mismatch of output transistors, and together those both reduce the distortion at all power levels.
1 side note question I have is do you have any blind tests that have been done with your amps comparing a good single ended class A to a good higher power class AB amp with a lower bias point to see if the distortion reduction is audible? I find it hard to find any blind tests or ABX tests that compare a good class AB to a good class A, and don't have the resources personally to run any kind of test of the sort.
ClassAB does operate in ClassA upto the limit of ClassA current.
There is an enormous range of this ClassA output from a ClassAB amplifier.
Take a typical ClassAB amplifier with a 1pair output stage biased to 26mVre across a 0r47 emitter resistor.
The output bias will be 65mA
The maximum ClassA current will be ~130mA, or just a fraction under.
This ClassAB amplifier will remain in ClassA for all outputs below 130mApk.
Power does not enter any of those controlling criteria.
But, if you need a power reference, then into a resistive load, 130mApk is equivalent to 67.6mW into 8r0 (8ohms resistive) That would be -28.7dB ref 50W into 8ohms.
If your PEAK current levels stay below 130mApk, then your maximum listening level would be -28.7dB below your amplifier's 50W maximum.
But change the output stage to 3pr with 0r1 emitter resistors and with 20mVre for the bias, the maximum ClassA becomes 1.2Apk and that would be -9.4dB ref the same 50W into 8ohms.
You could listen with 19dB more SPL and still stay in ClassA.
Now you need to look at what you measure as average levels at your speakers. Compare this average to your peak levels for your typical music/audio styles. This average to peak ratio can be as low as 6db, or as high as 30db.
A measured average level of 100mVac could have maximum output levels of 316mVac (447mVpk) to 3.16Vac (4.47Vpk). This average level of 100mVac is equivalent to 1.25mW and the range of maximum levels could be anywhere between 12.4mW and 1.24W
The first is less that the ClassA limit of the 1pr example and the second is also below the ClassA limit of the 3pr example.
Clearly at low level listening much of the output can be in ClassA and avoids all the crossover distortion of ClassAB.
There is an enormous range of this ClassA output from a ClassAB amplifier.
Take a typical ClassAB amplifier with a 1pair output stage biased to 26mVre across a 0r47 emitter resistor.
The output bias will be 65mA
The maximum ClassA current will be ~130mA, or just a fraction under.
This ClassAB amplifier will remain in ClassA for all outputs below 130mApk.
Power does not enter any of those controlling criteria.
But, if you need a power reference, then into a resistive load, 130mApk is equivalent to 67.6mW into 8r0 (8ohms resistive) That would be -28.7dB ref 50W into 8ohms.
If your PEAK current levels stay below 130mApk, then your maximum listening level would be -28.7dB below your amplifier's 50W maximum.
But change the output stage to 3pr with 0r1 emitter resistors and with 20mVre for the bias, the maximum ClassA becomes 1.2Apk and that would be -9.4dB ref the same 50W into 8ohms.
You could listen with 19dB more SPL and still stay in ClassA.
Now you need to look at what you measure as average levels at your speakers. Compare this average to your peak levels for your typical music/audio styles. This average to peak ratio can be as low as 6db, or as high as 30db.
A measured average level of 100mVac could have maximum output levels of 316mVac (447mVpk) to 3.16Vac (4.47Vpk). This average level of 100mVac is equivalent to 1.25mW and the range of maximum levels could be anywhere between 12.4mW and 1.24W
The first is less that the ClassA limit of the 1pr example and the second is also below the ClassA limit of the 3pr example.
Clearly at low level listening much of the output can be in ClassA and avoids all the crossover distortion of ClassAB.
Last edited:
- Status
- Not open for further replies.