In another thread, there was an implicit challenge to build an amplifier that was totally unlistenable with the following specifications:
The goal is to create an amplifier that can clearly be picked out in a double blind test, while having good paper specifications.
How would you go about creating such an amp?
My first Ideas were:
Near Class B operation.
Voltage doubler power supply. +-48V to prevent clipping.
Large Cgs input FET.
Low current driver
High feedback, about 60 db.
Additional series resistance in the power supply to simulate High ESR capacitors.
If this topic offends you, please take your input offline.
I await your creativity.
Doug
To add to the requirements list, I want it stable enough to not destroy speakers and use Power transformers that can be reused in an F5.high input impedance, low output impedance, flat response, low distortion, and low noise
The goal is to create an amplifier that can clearly be picked out in a double blind test, while having good paper specifications.
How would you go about creating such an amp?
My first Ideas were:
Near Class B operation.
Voltage doubler power supply. +-48V to prevent clipping.
Large Cgs input FET.
Low current driver
High feedback, about 60 db.
Additional series resistance in the power supply to simulate High ESR capacitors.
If this topic offends you, please take your input offline.
I await your creativity.
Doug
(This does not offend me, but it may be better placed in the Solid State section).
1. Make sure it clips just above rated max V swing
2. Agree with your 'near Class B' setting
3. Arrange it so that THD increases disproportionately with frequency
4. No zobel network (because you're going to test it on an ideal test load and then use it on a non-ideal real load)
5. Nested feedback? [I'm guessing]
6. Low PSRR?
Have fun!
1. Make sure it clips just above rated max V swing
2. Agree with your 'near Class B' setting
3. Arrange it so that THD increases disproportionately with frequency
4. No zobel network (because you're going to test it on an ideal test load and then use it on a non-ideal real load)
5. Nested feedback? [I'm guessing]
6. Low PSRR?
Have fun!
(This does not offend me, but it may be better placed in the Solid State section).
1. Make sure it clips just above rated max V swing
2. Agree with your 'near Class B' setting
3. Arrange it so that THD increases disproportionately with frequency
4. No zobel network (because you're going to test it on an ideal test load and then use it on a non-ideal real load)
5. Nested feedback? [I'm guessing]
6. Low PSRR?
Have fun!
Thanks,
I thought of 741 op amps, crazy power supply tricks.
Your list is good too. Its a silly thing, but it might swing a debate or 3.
Doug
LMAO
I really like this idea. Go for it.
You want at least 6 gain stages, why not make it 10 (there is my 2c). 10 is a nice round number and it sounds impressive. You could even market the amp - as the amp with the most gain stages in the world or even the galaxy (wow now that sounds impressive), the universe might be pushing the limit of credibility.
In the end it would be cool if you were to take it to the burning amp festival, and put it up against one of Nelson's early Zen designs.
I really like this idea. Go for it.
You want at least 6 gain stages, why not make it 10 (there is my 2c). 10 is a nice round number and it sounds impressive. You could even market the amp - as the amp with the most gain stages in the world or even the galaxy (wow now that sounds impressive), the universe might be pushing the limit of credibility.
In the end it would be cool if you were to take it to the burning amp festival, and put it up against one of Nelson's early Zen designs.
Last edited:
According to the Hitch Hikers Guide To The Galaxy, the answer to life the universe and everything = 42.
So I would like to edit my last post from 10 gain stages to 42 gain stages.
So I would like to edit my last post from 10 gain stages to 42 gain stages.
According to the Hitch Hikers Guide To The Galaxy, the answer to life the universe and everything = 42.
So I would like to edit my last post from 10 gain stages to 42 gain stages.
Marvin once built an amp just like this and it really bummed him out.
So I would like to edit my last post from 10 gain stages to 42 gain stages.
If you make it a Class D amp it would need 101010 stages?!?!
Just one more thing.
If the amp is going to be rated at 50W/channel into 8Ohms, then you don't want to put anything bigger than a 100VA (or 50VA per channel) transformer in their.
Also, put the crappiest chinese made caps (85 degree C) you can find in the power supply.
I think around the year 2000 there were some really "special" chinese caps available. If you can find some of those you couldn't go wrong really. Check out ebay.
If the amp is going to be rated at 50W/channel into 8Ohms, then you don't want to put anything bigger than a 100VA (or 50VA per channel) transformer in their.
Also, put the crappiest chinese made caps (85 degree C) you can find in the power supply.
I think around the year 2000 there were some really "special" chinese caps available. If you can find some of those you couldn't go wrong really. Check out ebay.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Pass Labs
- "Reference" amplifier