Member
Joined 2009
Paid Member
Wavebourn - I love this soft clipping, I gotta figure how to 'proudly borrow' this into my Harry 77 amp design - where my goal is to make a guitar-style amp for hi-fi 😀
I don't have that diode model. I tried other types of Silicon and
Schottky in that position with negligible effect. Series resistance
seems to be dominant in that branch.
However, I do see what you mean about current limiting. Hardly
any difference in current seen between 2 ohm load and 1ohm....
Your amp tops out a little over 4 amps and refuses to self destruct.
Mine (s19.asc) tries to deliver 14 Amps into same 1ohm abuse.
Both tested with same 23V input...
Schottky in that position with negligible effect. Series resistance
seems to be dominant in that branch.
However, I do see what you mean about current limiting. Hardly
any difference in current seen between 2 ohm load and 1ohm....
Your amp tops out a little over 4 amps and refuses to self destruct.
Mine (s19.asc) tries to deliver 14 Amps into same 1ohm abuse.
Both tested with same 23V input...
Member
Joined 2009
Paid Member
perfect !
I don't want to hijack this thread, so I started another one [http://www.diyaudio.com/forums/showthread.php?s=&threadid=146015&highlight=] . There's no hurry, I'm not going to be able to do much on it yet and I'm quite interested to follow this hybrid design.
I don't want to hijack this thread, so I started another one [http://www.diyaudio.com/forums/showthread.php?s=&threadid=146015&highlight=] . There's no hurry, I'm not going to be able to do much on it yet and I'm quite interested to follow this hybrid design.
kenpeter said:
Exactly! 😉
I also made such amps long time ago (in 1980'th) for cars, owners liked to roll-over speakers shorting wires, and Japanese amps did not survive. Mine never destructed themselves. 🙂
Edit: I called it Aikido, but returning back from software to audio design I discovered that the name had been taken by Tubecad sticking forever to SRPP.
gaetan8888 said:
About 0.5 milliamperes minimum, but I recommend about 100 times more idle current in your tube stage for better linearity. Don't forget about a coupling capacitor and couple of diodes to 25V rails to protect it's input transistors from overvoltages. You may use one triode strapped EL84 tube with 6 LEDs (2 parallel strings of 3 in series) in cathode, loaded on SVCS, like this one:
http://wavebourn.com/forum/download.php?id=125&f=7
You may remove D5 Zener in the SVCS (short it), decrease R7 to 24K, and power it from 300V instead of 500V. Tune R3 according to your taste.
Wavebourn said:
My experience is to use more current, and lower driving impeadance is better.
Another thing: I have no good experience with CFP output, and tube drivers. I don't know what is the problem with it, but it never sounds good to me. (I found something in Self's measurements, which reported wrong input impedance response for CFP output)
Sajti
kenpeter said:
At audio frequencies. The IRF510/9510 have low Cdg (some 10pF), it even less than the commonly used driver transistors (MJE1503x). The Ugs change is very low, due the driver function
Sajti
sajti said:
We were speaking about my bulletproof Aikido output follower (not SRPP, it is a complementary 2-stage opamp with unity gain). Contrary to conventional output follower it's input resistance goes down when it shifts from class A to B.
Hi Sajti,
would you have a screenshot in something like 10us/div, rather than 200us/div?
Thanks,
would you have a screenshot in something like 10us/div, rather than 200us/div?
Thanks,
I designed my own hybrid amplifier and simply bolted on an SRPP front end to a standard MOSFET class AB amp. Because the valve had a gain of 11 I had to reduce the gain of the MOSFET amp by that factor otherwise it was straightforward.
Last edited:
The only question is the bias, because there is no possibility to do any settings.
Sajti
The bias is adjusted by changing the values of resistors in the emitters of the first two transistors.
Do you think there is a danger of oscillations?
Do you think there is a danger of oscillations?
That setting is possible, but it's not easy to set the bias by that way...
The oscillation is not a big issue without feedback. But some base stopper resistor can be good
Sajti
What type of chain I can add in scheme to easily tune the bias?
Where I can add base stopper resistors and what are their values?
Where I can add base stopper resistors and what are their values?
- Status
- Not open for further replies.