Hi,
I'm looking for help in understanding the circuit design for the Haffler DH220.
First, a bit about me. I have a degree in physics and did some electronic labs in college, but I'm not an engineer and that was 20+ years ago. I do some electronics hobbying, mostly fixing broken stuff, but don't have any experience in audio projects, though that seems to be a hobby waiting to happen. I've good analytic and mechanical skills. I work as an IT Architect and restore motorcycles as a hobby. All this is to say that I think I'm up to the task of understanding this, but I could use a little help.
I have a working amp that I want to refurbish as it sounds a bit "dead". Not dead like it doesn't work, but just not very open, vibrant, etc. My thought is that I'll upgrade capacitors and replace components as needed based on testing them.
However, rather than just jump in, I'd like to have a solid understanding of the circuit first. From studying the schematic (posted below with my markup), reading literally hundreds of posts on this and other forums, reading appropriate sections electronics texts (The Art of Electronics, and Sloan's books High Power Audio Amplifier and Audiophile's Project Sourcebook), and reading through Marshall Leach's description of his amp design (very helpful), I have a pretty good high level understanding of the Haffler design. However, there are some sections of the circuit I don't understand.
What follows is my understanding (please correct any errors) as well as my questions.
The design follows Lin's 3 stage architecture. It is a complementary design, so in the following (and in my markup of the schematic), I'm just going to focus on 1/2 of the circuit
The input impedance is set by R1. The input is ac coupled through C1 and C2.
C21 and C22 are to remove any power supply ripple.
The input stage is a differential amplifier formed by Q3 & Q4 with a current source in the tail formed by Q2, D4, D5, D6 and R9.
The input stage is biased by the voltage divider that is adjusted by P1.
Output from the input stage is from the collector of Q3 and feeds the 2nd stage (the VAS). The 2nd stage is a Darlington Pair (Q7 & Q8). Diode D9 seems to work like a Baker Clamp to protect Q7 from saturating. If I understood my reading correctly, a Baker Clamp would have another diode, so I'm not sure if this actually is a Baker Clamp or not.
The output of stage two is from the collector of Q8 and feeds the base on the output driver Q12. Q12 is biased by the amplified diode formed with Q9, P2, and the associated resistors. Q12 is configured as an emitter-follower and the emitter output feeds the parallel MOSFETS that complete the output stage.
The output is run through some filters composed of L1& R36 along with R35 & C19 and R37 & C20 (I believe these two RC segments are part of a Zobel network).
Questions:
How is this 2nd stage biased? Is it through R9 and C4? If not, what is the purpose of R9 and C4?
What is the purpose of R34 and C18? My guess is that R34 is part of the feedback network and C18 provides a mean to bypass R18 for certain frequencies to adjust feedback.
The circuit near point A in the schematic is the feedback network, correct? If someone feels like explaining how that functions, that would be great, though just knowing it is the feedback network is enough for my general understanding.
What is the purpose of C23?
What is the purpose of C11, C12, and C13? More power supply filtering?
One of the output stage input drivers (lower part of diagram) has some capacitors in it's circuit segment (C10 and C17) that the other driver doesn't. What is the purpose of those and why on only one driver?
What are the diodes in the middle of the diagram for (D11, D12, D13, & D14)?
Finally, can someone explain the purpose of the Resistors and Capacitor (R401, R402, and C401) in the output stage? Are those there for biasing? I'm having a hard time wrapping my head around that detail.
Thanks in advance for your input on this. If I get a thorough (and accurate) understanding of the circuit, I'll write it up and repost it.
James
I'm looking for help in understanding the circuit design for the Haffler DH220.
First, a bit about me. I have a degree in physics and did some electronic labs in college, but I'm not an engineer and that was 20+ years ago. I do some electronics hobbying, mostly fixing broken stuff, but don't have any experience in audio projects, though that seems to be a hobby waiting to happen. I've good analytic and mechanical skills. I work as an IT Architect and restore motorcycles as a hobby. All this is to say that I think I'm up to the task of understanding this, but I could use a little help.
I have a working amp that I want to refurbish as it sounds a bit "dead". Not dead like it doesn't work, but just not very open, vibrant, etc. My thought is that I'll upgrade capacitors and replace components as needed based on testing them.
However, rather than just jump in, I'd like to have a solid understanding of the circuit first. From studying the schematic (posted below with my markup), reading literally hundreds of posts on this and other forums, reading appropriate sections electronics texts (The Art of Electronics, and Sloan's books High Power Audio Amplifier and Audiophile's Project Sourcebook), and reading through Marshall Leach's description of his amp design (very helpful), I have a pretty good high level understanding of the Haffler design. However, there are some sections of the circuit I don't understand.
What follows is my understanding (please correct any errors) as well as my questions.
The design follows Lin's 3 stage architecture. It is a complementary design, so in the following (and in my markup of the schematic), I'm just going to focus on 1/2 of the circuit
The input impedance is set by R1. The input is ac coupled through C1 and C2.
C21 and C22 are to remove any power supply ripple.
The input stage is a differential amplifier formed by Q3 & Q4 with a current source in the tail formed by Q2, D4, D5, D6 and R9.
The input stage is biased by the voltage divider that is adjusted by P1.
Output from the input stage is from the collector of Q3 and feeds the 2nd stage (the VAS). The 2nd stage is a Darlington Pair (Q7 & Q8). Diode D9 seems to work like a Baker Clamp to protect Q7 from saturating. If I understood my reading correctly, a Baker Clamp would have another diode, so I'm not sure if this actually is a Baker Clamp or not.
The output of stage two is from the collector of Q8 and feeds the base on the output driver Q12. Q12 is biased by the amplified diode formed with Q9, P2, and the associated resistors. Q12 is configured as an emitter-follower and the emitter output feeds the parallel MOSFETS that complete the output stage.
The output is run through some filters composed of L1& R36 along with R35 & C19 and R37 & C20 (I believe these two RC segments are part of a Zobel network).
Questions:
How is this 2nd stage biased? Is it through R9 and C4? If not, what is the purpose of R9 and C4?
What is the purpose of R34 and C18? My guess is that R34 is part of the feedback network and C18 provides a mean to bypass R18 for certain frequencies to adjust feedback.
The circuit near point A in the schematic is the feedback network, correct? If someone feels like explaining how that functions, that would be great, though just knowing it is the feedback network is enough for my general understanding.
What is the purpose of C23?
What is the purpose of C11, C12, and C13? More power supply filtering?
One of the output stage input drivers (lower part of diagram) has some capacitors in it's circuit segment (C10 and C17) that the other driver doesn't. What is the purpose of those and why on only one driver?
What are the diodes in the middle of the diagram for (D11, D12, D13, & D14)?
Finally, can someone explain the purpose of the Resistors and Capacitor (R401, R402, and C401) in the output stage? Are those there for biasing? I'm having a hard time wrapping my head around that detail.
Thanks in advance for your input on this. If I get a thorough (and accurate) understanding of the circuit, I'll write it up and repost it.
James
How is this 2nd stage biased? Is it through R9 and C4? If not, what is the purpose of R9 and C4?
R8/9 set the current source values, 1.2V/R9, for the diff amps. The C4/R12 and C5/R19 are compensation networks.
What is the purpose of R34 and C18?
R34/C18 operate when the fuse blows and keep the feedback loop connected.
The circuit near point A in the schematic is the feedback network, correct?
Yes, the gain is set by this network, and is unity at DC because of C8.
What is the purpose of C23?
It probably reduces noise in the ground system.
What is the purpose of C11, C12, and C13? More power supply filtering?
Yes, along with the diode/resistor feeding them. They also prevent the output stage from affecting the driver stage through the power supply.
One of the output stage input drivers (lower part of diagram) has some capacitors in it's circuit segment
(C10 and C17) that the other driver doesn't. What is the purpose of those and why on only one driver?
The different polarity output devices have differing parasitic capacitances. This equalizes that, and/or improves stability.
What are the diodes in the middle of the diagram for (D11, D12, D13, & D14)?
These are bilateral clamps to limit the voltage.
Finally, can someone explain the purpose of the Resistors and Capacitor (R401, R402, and C401) in the output stage?
These gate resistors are essential for stable operation and prevent oscillation. They limit hf response.
R8/9 set the current source values, 1.2V/R9, for the diff amps. The C4/R12 and C5/R19 are compensation networks.
What is the purpose of R34 and C18?
R34/C18 operate when the fuse blows and keep the feedback loop connected.
The circuit near point A in the schematic is the feedback network, correct?
Yes, the gain is set by this network, and is unity at DC because of C8.
What is the purpose of C23?
It probably reduces noise in the ground system.
What is the purpose of C11, C12, and C13? More power supply filtering?
Yes, along with the diode/resistor feeding them. They also prevent the output stage from affecting the driver stage through the power supply.
One of the output stage input drivers (lower part of diagram) has some capacitors in it's circuit segment
(C10 and C17) that the other driver doesn't. What is the purpose of those and why on only one driver?
The different polarity output devices have differing parasitic capacitances. This equalizes that, and/or improves stability.
What are the diodes in the middle of the diagram for (D11, D12, D13, & D14)?
These are bilateral clamps to limit the voltage.
Finally, can someone explain the purpose of the Resistors and Capacitor (R401, R402, and C401) in the output stage?
These gate resistors are essential for stable operation and prevent oscillation. They limit hf response.
Last edited:
you should look for articles "servo 100" and "servo 50" by erno borbely, (who also designed these amps for hafler).
they might be available via the wayback machine since borbely closed up shop some years ago.
wow, all these years with this schematic and i never noticed no mosfet source resistors.
yikes ...
mlloyd1
they might be available via the wayback machine since borbely closed up shop some years ago.
wow, all these years with this schematic and i never noticed no mosfet source resistors.
yikes ...
mlloyd1
R12/19 - C4/5 + C23 are lead/lag compensation.
Mosfet designers seem to favor this method instead
of miller like in BJT EF3 designs (leach amp uses miller and lead comp.).
OS
Mosfet designers seem to favor this method instead
of miller like in BJT EF3 designs (leach amp uses miller and lead comp.).
OS
There is also Issue 2/1982 Audio Amateur, "A 60W MOSFET Power Amplifier" by Erno Borbely with circuit info and using a single pair of outputs.
Wow, thanks for the responses. Rayma, your answers helped me understand this quite a bit.
I realize I wrote this question incorrectly:
"How is this 2nd stage biased? Is it through R9 and C4? If not, what is the purpose of R9 and C4?"
What I meant to suggest is that the biasing is through R12 and C4. You indicated those are for compensation, and that biasing is done through R8/9, but that is for the input stage only, right?
How does the Voltage Amp Stage (Q7) get biased?
Thanks again,
James
I realize I wrote this question incorrectly:
"How is this 2nd stage biased? Is it through R9 and C4? If not, what is the purpose of R9 and C4?"
What I meant to suggest is that the biasing is through R12 and C4. You indicated those are for compensation, and that biasing is done through R8/9, but that is for the input stage only, right?
How does the Voltage Amp Stage (Q7) get biased?
Thanks again,
James
R8/9 set the value of the currents through each input diff amp. They are each part of a current source for the two "tails."
Buffers Q7/10 and amplifiers Q8/11 are directly coupled to the diff amps, and R21/30 set the current. These drive the bias circuit for the output stage.
Last edited:
Hi guys, need your advice on the project I'm doing now. I need to do my physics homework using an old amplifier, I've both: Hafler DH-200 andHafler DH-220 (bought by my dad years ago). I planned to use two sets of matched pairs of 2N5550 and 2N5401 - is it possible to get pre matched sets of these without paying too much? Has anyone modified/repaired Hafler DH-220 and could you please share the results? Thanks.