This is about building a good amplifier where transformer is utilizing the commonly available step-down transformer (230V to 120V) that can be found in many devices (including old television). Most of them have a high VA rating and the biggest I have is a 2000VA toroid. For EI transformers, unwinding the secondary is of course out of question, hence this thread.
I want the best quality amp I can build. Preferably suitable for home use but if not possible then for PA use. The output device is of course a mosfet. Currently I'm using IRFP460. But I plan to use high power mosfets in ISOTOP case as I have plenty of them. I couldn't find Spice model for the ISOTOP mosfets but I planed to model the mosfet using paralleled mosfet with equal Yfs per Ciss (I hope that possible).
I have only tried one topology with output coupling capacitor, and I wonder what others might suggest regarding topology, or even better if there is a working design already available. My next candidate (for the topology) is the Legend Quasor by Boraomega (Dr Jagodic). Then the hybrid tube-mosfet based on John Broskie (but I'm worrying the high output impedance so not in the top list). I was also considering cascoded bipolar output ala SAE PA amplifiers but not a priority as I don't have many bipolar devices.
Please suggest a topology that you think will work best with this kind of rail voltage (150Vdc-160Vdc). Thanks.
I want the best quality amp I can build. Preferably suitable for home use but if not possible then for PA use. The output device is of course a mosfet. Currently I'm using IRFP460. But I plan to use high power mosfets in ISOTOP case as I have plenty of them. I couldn't find Spice model for the ISOTOP mosfets but I planed to model the mosfet using paralleled mosfet with equal Yfs per Ciss (I hope that possible).
I have only tried one topology with output coupling capacitor, and I wonder what others might suggest regarding topology, or even better if there is a working design already available. My next candidate (for the topology) is the Legend Quasor by Boraomega (Dr Jagodic). Then the hybrid tube-mosfet based on John Broskie (but I'm worrying the high output impedance so not in the top list). I was also considering cascoded bipolar output ala SAE PA amplifiers but not a priority as I don't have many bipolar devices.
Please suggest a topology that you think will work best with this kind of rail voltage (150Vdc-160Vdc). Thanks.
I have plenty of parts. I usually don't buy parts for a project. I have hundreds of transformers, including these step-down ones.
I think not many have interest in buying step-down transformer, especially for a high price
Why do you think it would be costly? I have plenty of big heatsinks also.
I think not many have interest in buying step-down transformer, especially for a high price
Why do you think it would be costly? I have plenty of big heatsinks also.
I would start with building a low power amplifier like 50-100 watts per channel into 4 ohms, plenty of schyematics to choose from on this forum.
Suitable rail voltages for these would be 2x35-45 volts DC.
Building a reliable amplifier in the kilowatt range is extremely hard due to the high rail voltages, it is also very expensive, one mistake and you can loose hundreds of dollars in output devices alone.
Suitable rail voltages for these would be 2x35-45 volts DC.
Building a reliable amplifier in the kilowatt range is extremely hard due to the high rail voltages, it is also very expensive, one mistake and you can loose hundreds of dollars in output devices alone.
If you mean the ones with "normal" rail voltage, I have built most of the best ones.
Hundreds of dollar?? I hope not due to electric shock
BTW, I'm trying to utilize transformers with 120V secondaries, not to build KW amps. Non-switching class-B is possible.
Here's one.
It does use split rails again though.
http://www.elektropage.com/images/circuits//1Kw_Rms_Mosfet_Amplifier_Shematic.jpg
It does use split rails again though.
http://www.elektropage.com/images/circuits//1Kw_Rms_Mosfet_Amplifier_Shematic.jpg
+/-95V is very common. 160V is not,mostly due to transistor SOA I guess. The original Legend Quasor uses 95V, and I think somebody has built it with very high voltage rail, but no schematic available.
I know this project is a bit crazy, that's why I posted it here, because I know there are many crazy people here
Thanks for the link. I like the circuit but I forget to mention that I will use N-channel device only, preferably mosfet in ISOTOP case (their relatively high Ciss makes it even more difficult to work with most circuits).
Attachments
Are those isotops STE47N50's ? I have twelve of those that i was planning to build a stereo amp with, i built a ugly prototype which can be seen here: NMOS amp in stereo - YouTube
The plan is to use six fets per channel, it'll need some hefty drive though, bit since this amp is not intended for hifi, i am not really worried about frequency response.
Schematic of the test amp in the video: http://i.imgur.com/z7i8h.png
The plan is to use six fets per channel, it'll need some hefty drive though, bit since this amp is not intended for hifi, i am not really worried about frequency response.
Schematic of the test amp in the video: http://i.imgur.com/z7i8h.png
I have at least six models. In front of me I have E110NA20. But I think I have also STE47N50.
Hi, I have downloaded the schematic the day you posted it, even if imho such topology will not give good sound. Even if these isotops have high Ciss, they can still sound good. Of course it needs relatively high current.
With around 6000pF of Ciss, I think the ISOTOP is comparable to 3 to 6 "ordinary" mosfets. So 3 pairs is a lot. One isotop is capable of very high current, so 3 pairs is feasible only if very big power is expected.
Good idea. Not perfect (efficient) but if nothing else it is a solution I can use, especially when the VA rating is too low for a working amplifier circuit.
This output transformer thing is new to me
Yes, TubeCad DotCom has such thing but with no gain. Actually I already have a solution but I think people might have better idea because imo this is a frequent common situation where people get in touch with step-down transformers. Well, of course not in US or Japan.
Something like the QSC or whoever it was topology might be worth a try, wheres the amp output is grounded and the speaker output is the center point between the two rail capacitors.
The all NMOS output as you say is not a very good topology, in fact it simulates really poorly no matter what, but some of the ppl doing mosfet amps seem to prefer NMOS output over complementary output as there is no good match between the N channel and P channel vertical mosfets.
The all NMOS output as you say is not a very good topology, in fact it simulates really poorly no matter what, but some of the ppl doing mosfet amps seem to prefer NMOS output over complementary output as there is no good match between the N channel and P channel vertical mosfets.
Thanks I will check the QSC. I forget the schematics but as far as I can remember it is not applicable. I think it was the complementary type.
Yes, I know. Quasi is chosen for economical reason only, not P-N matching reason. But for PA usage, the simulation performance is more than acceptable imo. Don't know with reliability as there is that trade-off.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.