Hello,
I am experimenting with a Quasi Complementary amp design that uses TIP41C output transistors. The DC rail voltages are 45 Volts, and the amp is only rated 25 Watts. I think that is a little low, and I should be able to get more power out of the circuit.
Maybe swapping transistors would help. Does anyone know of a transistor upgrade to a TIP41C that will boost output power? Maybe I have to go to a TO-3 package instead of TO-220.
Thanks,
Woodstock
I am experimenting with a Quasi Complementary amp design that uses TIP41C output transistors. The DC rail voltages are 45 Volts, and the amp is only rated 25 Watts. I think that is a little low, and I should be able to get more power out of the circuit.
Maybe swapping transistors would help. Does anyone know of a transistor upgrade to a TIP41C that will boost output power? Maybe I have to go to a TO-3 package instead of TO-220.
Thanks,
Woodstock
TIP41 substitutes
I assume you have a single 45V rail and ground as 90V across what is fundamentally still only a 60V transistor will run it outside its Safe Operating Area. With +/-45V rails, you should be thinking big stuff, even with 8R loads.
These days, we use TO247, TO264, TO3P flat plastic packs like large T0220s for more dissipation. Actually, TO220 packages have been handling around 25 watts since the 1970s so it's not necessarily a problem. The TIP41 however, is not a modern device and is not as linear as the brilliant MJE15028-35 series transistors which can handle quite bit of thrashing with considerably lower distortion within the package and SOA limits.
Either MJE15028,30,32,34 would be fine. If you need more power for 4R loads, 2 pairs would also give better performance again.
Price can be important in parts choice and the very popular, cheap and capable 2SC5200 are a good choice for Quasi designs. I've built several sets of Naim clones using them, without problems. Though cheap, they are much larger parts than TO220. (sim. to TO264). Fairchild manufacture a series of good clones of these Toshiba parts and also their smaller, TO247 cousins.
The lowest distortion should be easiest using On-semi MJL3281 series parts but as Quasi designs go, I don't think there is a real benefit there - at least I haven't heard any improvement as you might discern in complementary designs.
There are other good parts, if a bit slow, like MJL21194 and the older TO3 types. You'll find most of them in the On-semi catalogue of Audio power transistors. There are many variations by family group number and prefix which denote Vceo max, package type and features like diodes for thermal sensing. Take a close look at the specs. - there actually aren't many fundamental differences within each series at all. Audio Transistors
I assume you have a single 45V rail and ground as 90V across what is fundamentally still only a 60V transistor will run it outside its Safe Operating Area. With +/-45V rails, you should be thinking big stuff, even with 8R loads.
These days, we use TO247, TO264, TO3P flat plastic packs like large T0220s for more dissipation. Actually, TO220 packages have been handling around 25 watts since the 1970s so it's not necessarily a problem. The TIP41 however, is not a modern device and is not as linear as the brilliant MJE15028-35 series transistors which can handle quite bit of thrashing with considerably lower distortion within the package and SOA limits.
Either MJE15028,30,32,34 would be fine. If you need more power for 4R loads, 2 pairs would also give better performance again.
Price can be important in parts choice and the very popular, cheap and capable 2SC5200 are a good choice for Quasi designs. I've built several sets of Naim clones using them, without problems. Though cheap, they are much larger parts than TO220. (sim. to TO264). Fairchild manufacture a series of good clones of these Toshiba parts and also their smaller, TO247 cousins.
The lowest distortion should be easiest using On-semi MJL3281 series parts but as Quasi designs go, I don't think there is a real benefit there - at least I haven't heard any improvement as you might discern in complementary designs.
There are other good parts, if a bit slow, like MJL21194 and the older TO3 types. You'll find most of them in the On-semi catalogue of Audio power transistors. There are many variations by family group number and prefix which denote Vceo max, package type and features like diodes for thermal sensing. Take a close look at the specs. - there actually aren't many fundamental differences within each series at all. Audio Transistors
Upgrade from TIP41C
Ian,
Thanks for your reply, pointing me to the On Semi audio site, and your suggestions to look at MJE15028, 30,32, 34, 2SC5200, MJL3281.
TIP41A is 60 V, TIP41B is 80 V, TIP41C is 100 V. TIP 41C works reliably, but with +/-45V rails, I should be able to get more power.
I realize that I need to go to a bigger package. I was looking at the MJL21194 in the TO-247 package, but as you mentioned, it is a bit slow. [Ft only goes from TIP41C's 3 to 4. I won't be able to hear that difference. Maybe I need a transistor that was designed this century]. Also, I don't know if it will be that big of an improvement. From 25 Watts to 50 Watts is 3dB. Would I even get that much from the MJL21194?
I will try it since I can get it locally while I am waiting for the real parts.
I like the 2SC5200 and MJL3281 a lot. I will order both and see which works best.
Thanks again.
Ian,
Thanks for your reply, pointing me to the On Semi audio site, and your suggestions to look at MJE15028, 30,32, 34, 2SC5200, MJL3281.
TIP41A is 60 V, TIP41B is 80 V, TIP41C is 100 V. TIP 41C works reliably, but with +/-45V rails, I should be able to get more power.
I realize that I need to go to a bigger package. I was looking at the MJL21194 in the TO-247 package, but as you mentioned, it is a bit slow. [Ft only goes from TIP41C's 3 to 4. I won't be able to hear that difference. Maybe I need a transistor that was designed this century]. Also, I don't know if it will be that big of an improvement. From 25 Watts to 50 Watts is 3dB. Would I even get that much from the MJL21194?
I will try it since I can get it locally while I am waiting for the real parts. I like the 2SC5200 and MJL3281 a lot. I will order both and see which works best.
Thanks again.
With stiff +/-45V rails, you should be able to reliably get >80W in class AB,
The point to be aware of with varieties of Vceo ranges in TIP31/2 TIP41/2 A-C ranges and even the MJE15028-35 range is that the chips have not increased in size or dissipation to match the greater potential of higher applied rail voltage. The only difference between them is chip Vceo & Vcbo, not power.
IOW, we must raise load impedance to counter the increased dissipation, otherwise the higher supply and drive potential will now cause it to exceed Safe Operating Area rating. (see the SOA graphs on the manufacturer's data sheets) As far as domestic audio applications are concerned, it is still only a 25W/pr transistor whatever it is graded so rails should be kept low as appropriate to a 25W output stage driving 4R loads. That is only about 25V at full power for a single pair. That's also why we usually give smaller packages and small SOA chips the flick when the power requirements start to get serious - dissipation.
There are a few angles on the listener benefits of small power increases and to be sure, in that critical 1kHz band, our ears are said to require even 10dB difference for a perceived doubling of volume. However, down in the bass region, there is a much more linear relationship with the upshot that the larger amps will usually be capable of stronger apparent bass - no small issue with modest speaker systems.
At least, I think it's worth having 80W rather than even 50W in that situation, as the cost difference will not be great.
The point to be aware of with varieties of Vceo ranges in TIP31/2 TIP41/2 A-C ranges and even the MJE15028-35 range is that the chips have not increased in size or dissipation to match the greater potential of higher applied rail voltage. The only difference between them is chip Vceo & Vcbo, not power.
IOW, we must raise load impedance to counter the increased dissipation, otherwise the higher supply and drive potential will now cause it to exceed Safe Operating Area rating. (see the SOA graphs on the manufacturer's data sheets) As far as domestic audio applications are concerned, it is still only a 25W/pr transistor whatever it is graded so rails should be kept low as appropriate to a 25W output stage driving 4R loads. That is only about 25V at full power for a single pair. That's also why we usually give smaller packages and small SOA chips the flick when the power requirements start to get serious - dissipation.
There are a few angles on the listener benefits of small power increases and to be sure, in that critical 1kHz band, our ears are said to require even 10dB difference for a perceived doubling of volume. However, down in the bass region, there is a much more linear relationship with the upshot that the larger amps will usually be capable of stronger apparent bass - no small issue with modest speaker systems.
At least, I think it's worth having 80W rather than even 50W in that situation, as the cost difference will not be great.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.