John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
I usually avoid the resistor-inductor combination in my power amps. This takes a bit more internal compensation to keep the amp stable.

This is perfect "wire with gain". :)

Inductor, zobels are gadgets for mass-production amplifiers, to achieve some kind of stability, to make good results in some typical capacitive load measurements, (maybe) to load cheap speakers lack of zobel in crossovers...or without impedance corrections...
It is funny to see random microH coil in front off 1H or similar coil in crossovers..
Or ubiquitous 10ohm -100nF combination for all purposes.
 
Surely there is a question of how far are you prepared to push the amp into difficult loads and for how long before the protection kicks in. Some very expensive and highly regarded amps define impulses as lasting all of 1 mS, a period I consider to be uselessly short. Remember, IEC standards require a period of 20 mS. In times of stress, the output stage may well find itself called upon to dissipate over 5A per output device, depending on te load.

This will tax ANY compensation network inside the amp and may well demand a series L/R circuit. Whoever disbelieves is advised to use Nelson's simulated speaker for a test, just watch the required output current jump from nominal 2.4A to 5.4 A. then observe how it goes from 1 to 20 to 50 mS intervals. The most obvious cure is to add new output devices, offloading each one of them by sharing the workload, but that costs money and requires more interior real estate both on the PCB(s) and heat sinks. Spend enough and you will be just fine, but remember that even using ON Semi's 230W NJL 4281/4302 plastic pack moneters (also available in a full metal TO-3 package) in four pairs will not be enough for a full peace of mind. Mostly because of heating problems.

Thus, the power output required and the difficulty level of the load will greatly influence the design of the entire output stage and may well dictate what else you miight need - or not.
 
Actually impedance-corrected speakers are in general more in need of the series R//L because their Zin is more capacitive.

Jan

But until this becomes a hard rule, it's the amps who will have to make do as best they can.

BTW, my own speakers have it built in, how else could we keep the impedance modulus 6.5 worst case to 12 Ohms mximum, and in a 3 way system?
 
Member
Joined 2011
Paid Member
The reason for the output inductor is not well understood here. The major component is the added output resistor (1-10 ohms) that buffers the load capacitance from the output. The inductor is simply added to bypass the series resistor at low frequencies to keep the output impedance at audio frequencies low. The output inductance is usually between 0.5 and 5uH, 2uH being typical
What exactly is wrong with an output inductor?
Plot the magnitude of impedance of (10R || 2uH) vs frequency. Compare to a "4 ohm" loudspeaker. Calculate damping factor (Zspkr/Zamplifier). Are you pleased?

{R=10 ohms, from D. Self's schematics of both "Trimodal Amplifier" and "Load Invariant Power Amplifier"}
 
Plot the magnitude of impedance of (10R || 2uH) vs frequency. Compare to a "4 ohm" loudspeaker. Calculate damping factor (Zspkr/Zamplifier). Are you pleased?

{R=10 ohms, from D. Self's schematics of both "Trimodal Amplifier" and "Load Invariant Power Amplifier"}

This question just opens up another can of worms.

To JC. Isn't the 10 ohms or so of resistance there to damp the inductor.
 
Plot the magnitude of impedance of (10R || 2uH) vs frequency. Compare to a "4 ohm" loudspeaker. Calculate damping factor (Zspkr/Zamplifier). Are you pleased?

{R=10 ohms, from D. Self's schematics of both "Trimodal Amplifier" and "Load Invariant Power Amplifier"}

Let's ballpark that:

2 uH inductor is roughly 20 turns around 12mm ID form (single layer). That's (roughly) 75 cm of wire. Using 16 awg wire (not too crazy nor too stingy), that's 12 mOhms. Needless to say, I'd be much more worried about stray fields from the solenoid.
 
Last edited:
www.hifisonix.com
Joined 2003
Paid Member
Nice little bit of simulation. Yes, looks like a little output inductance as insurance against pathological cables buys you more stability than lowering ULGF, too. :)

Yes - absolutely.

Re the rise time comment posted earlier: I don't think 1-2uH on the amplifier output will decrease the post output stage system rise time - you have as much or more in cable inductance plus your crossover and drive units which completely swamp that.
 
Output inductors can change the sound quality. I have heard the difference.
As far as your analysis, please realize that we have been looking closely at this addition to the output for many decades. I first analyzed it carefully 40 years ago while making a balanced bridge output amp for GALE Electronics in London. Yes, I used an output coil in my designs for more than 15 years, until I was shown that they do make a sonic difference. It would seem to do little except to add a little phase shift and lower the damping factor at higher frequencies, but it appears to do more, subjectively. Many noted designers have removed the output coil in their designs, including Julian V. of NAIM, Nelson Pass, Charles Hansen, and me. It is difficult to maintain complete stability with every possible capacitive load, so be careful if you try this in your next design. For example, some loudspeaker cables may just not work and will make the amp oscillate. You might have to increase the base resistor value in each of your output devices to as much as 10 ohms to maintain reasonable stability.
 
Output inductors can change the sound quality. I have heard the difference.
As far as your analysis, please realize that we have been looking closely at this addition to the output for many decades. I first analyzed it carefully 40 years ago while making a balanced bridge output amp for GALE Electronics in London. Yes, I used an output coil in my designs for more than 15 years, until I was shown that they do make a sonic difference. It would seem to do little except to add a little phase shift and lower the damping factor at higher frequencies, but it appears to do more, subjectively. Many noted designers have removed the output coil in their designs, including Julian V. of NAIM, Nelson Pass, Charles Hansen, and me. It is difficult to maintain complete stability with every possible capacitive load, so be careful if you try this in your next design. For example, some loudspeaker cables may just not work and will make the amp oscillate. You might have to increase the base resistor value in each of your output devices to as much as 10 ohms to maintain reasonable stability.


Okay Sure. But why does it make a sonic difference?
 
Member
Joined 2011
Paid Member
Okay Sure. But why does it make a sonic difference?
Surely you must mean "What is the largest value of inductance that DOES NOT make a sonic difference? 1 picohenry? 1 nanohenry? 10 nanohenrys? 100 nanohenrys?"

Surely you are asking "Where is the dividing line between audible and inaudible amounts of inductance?" and surely you don't believe it is smaller than 5 nanohenries.
 
Surely you must mean "What is the largest value of inductance that DOES NOT make a sonic difference? 1 picohenry? 1 nanohenry? 10 nanohenrys? 100 nanohenrys?"

Surely you are asking "Where is the dividing line between audible and inaudible amounts of inductance?" and surely you don't believe it is smaller than 5 nanohenries.

Some typical values used where mentioned. 1uH to 5uH.

But John and company say no inductor is unacceptable.
 
Status
Not open for further replies.