My own best figures for my amplifier design.
Here some simulation to compare with your nice tests, wahab.
I think this is one of my best. Have been working with it last days.
The output stage is 2 parallel pairs MJL3281A/MJL1302A driven in a Diamond.
Idle is 100mA (50mA per device). So we have standard Class AB.
What lower the THD is the frontend.
Input and Feedback and correction of it all is by one OPA134 jfet Opamp.
The OPA134 sets the limit. And it is good.
One step further could be using AD797 or OPA627, which probably would improve a bit.
Supply is +/-45V. THD for 8 Ohm:
0.000046 ... 1 Watt
0.000061 ... 25 Watt (20Vp)
0.000106 ... 100 Watt (40Vp)
The blue curve in my image is the level you often have used.
I use 1kHz and 20 harmonics. It is enough for me.
Here some simulation to compare with your nice tests, wahab.
I think this is one of my best. Have been working with it last days.
The output stage is 2 parallel pairs MJL3281A/MJL1302A driven in a Diamond.
Idle is 100mA (50mA per device). So we have standard Class AB.
What lower the THD is the frontend.
Input and Feedback and correction of it all is by one OPA134 jfet Opamp.
The OPA134 sets the limit. And it is good.
One step further could be using AD797 or OPA627, which probably would improve a bit.
Supply is +/-45V. THD for 8 Ohm:
0.000046 ... 1 Watt
0.000061 ... 25 Watt (20Vp)
0.000106 ... 100 Watt (40Vp)
The blue curve in my image is the level you often have used.
I use 1kHz and 20 harmonics. It is enough for me.
Attachments
so if I get a carbon 3W/10R , my troubles are over ??
How about ceramic comp. devices ??
OS
Last edited:
Hi ostripper,
No, my test was only: What happens to the value of inductance if I stick different resistors into the middle of it. A static test without a signal going through the coil. The resistors on hand made little if any difference.
wahab is talking about the "changing" signal going through coil producing a magnetic field that induces eddy currents in the resistor. Sounds plausible but is it significant?
My best sounding amps have the resistor in the guts of the inductor, but I have never experimented changing it.
regards
Impressive ratios, indeed..
Just one thing : you should either increase the resolution
or use more cycles to compute the fourier analysis, or both.
If the base frequency is 1KHZ, we should see the curve
going deep between this frequency and its first harmonic, then
surge in direction of the 2khz harmonic level..
Also, it s odd that there s only odd(!) harmonics that
are at a significant level, the even ones being all absent..
Last edited:
The resistors s lead will be still metallic,
although the resistor body is less problematic
than with a metal resistor one...
Have you ever seen a japanese amp with the
coil wounded around the dumping resistor
of an output filter ?....
Hi wahab,
My own experience using the coil wrapping the resistor seems to point in the direction that there is very little, if any, effects from doing this. Even if one were to use a single copper wire for the damping resistor (there's that word again), the overall circuit runs through the speaker, power supply caps and output devices. The local circuit is made up of the coil and resistor. At best, it is a one half turn, lossy circuit. I don't know if there would be any more effective current running in the short loop. Sometime when I have something like this open on the bench, I'll use a current probe for my 'scope and measure between resistor in and resistor outside of the coil. I'll use an 8 ohm dummy load and excite it with an amplifier output at a single frequency. My intended resistor will be a 0.22R WW or helical cut type resistor. I'll use a switch between a resistor around the coil, and one inside. That ensures the same coil is in use for each test and also allows me to mount the current probe in one location without having to move it in any way. Those familiar with current probes know that position changes can affect the readings sometimes.
Hi OS,
I know your amps simulate extremely well. I still wish to build one stereo version.
Hi metal,
What other listed amplifiers have you built? Any commercial models? If so, what are some of your favorites? I'm trying to get a feel for what you like, or not. I'm not challenging you at all. I am genuinely interested and curious. Judging your preferences is not something I am planning to do either.
I have not heard the DX, but it's schematic suggests higher levels of H2, which some designers purposely add in. In fact, there are designers who tailor the harmonic content in order to maintain a certain sound signature. They could make a lower distortion amplifier, but they choose a particular sound and therefore will have some fans of that sound. But that is an entirely different discussion as I mentioned above.
-Chris
My own experience using the coil wrapping the resistor seems to point in the direction that there is very little, if any, effects from doing this. Even if one were to use a single copper wire for the damping resistor (there's that word again), the overall circuit runs through the speaker, power supply caps and output devices. The local circuit is made up of the coil and resistor. At best, it is a one half turn, lossy circuit. I don't know if there would be any more effective current running in the short loop. Sometime when I have something like this open on the bench, I'll use a current probe for my 'scope and measure between resistor in and resistor outside of the coil. I'll use an 8 ohm dummy load and excite it with an amplifier output at a single frequency. My intended resistor will be a 0.22R WW or helical cut type resistor. I'll use a switch between a resistor around the coil, and one inside. That ensures the same coil is in use for each test and also allows me to mount the current probe in one location without having to move it in any way. Those familiar with current probes know that position changes can affect the readings sometimes.
Hi OS,
I know your amps simulate extremely well. I still wish to build one stereo version.
Absolutely! Not unless the flux is high enough to crowd the electron average pathway. Otherwise, carbon would be very similar to electrons flowing in air. Only the coil will exert a magnetic field.
Hi metal,
Cool.
What other listed amplifiers have you built? Any commercial models? If so, what are some of your favorites? I'm trying to get a feel for what you like, or not. I'm not challenging you at all. I am genuinely interested and curious. Judging your preferences is not something I am planning to do either.
I have not heard the DX, but it's schematic suggests higher levels of H2, which some designers purposely add in. In fact, there are designers who tailor the harmonic content in order to maintain a certain sound signature. They could make a lower distortion amplifier, but they choose a particular sound and therefore will have some fans of that sound. But that is an entirely different discussion as I mentioned above.
-Chris
Hi wahab,
Circuit PCB traces count as a metallic conductor as well. What effect might they have if they form a partial loop in this circuit even if the resistor is located away from the coil?
-Chris
Edit:
This is fairly common practice. Only component insertion costs have changed that practice. A resistor and coil are easier installed by machine separately, rather than one assembly. Not unless someone comes out with a cheaper combination part. the choice is driven by economics, not by science most often.
True, but the circuit is not continuous. Only an experiment would really show how this might be affected by metallic leads.
Circuit PCB traces count as a metallic conductor as well. What effect might they have if they form a partial loop in this circuit even if the resistor is located away from the coil?
-Chris
Edit:
Yes, thousands believe it or not.
This is fairly common practice. Only component insertion costs have changed that practice. A resistor and coil are easier installed by machine separately, rather than one assembly. Not unless someone comes out with a cheaper combination part. the choice is driven by economics, not by science most often.
Last edited:
Fidelix LB-4, 1980, Mr. Shin Nakagawa
Even expensive critters for merely 25W amps.
(just for fun : http://www.fidelix.jp/products/LZ-12.pdf )
Attachments
Yes, you have been talked into a test, Wahab. Now I am afraid you may have to compare HD spectra or just THD with a lash-up inductor. Could I suggest a significant 3-7 uH winding with damper resistors of 12 ohms WW and carbon 1W? These are typical bullet-proofing values used for many years in commercial amps.
Drive nominally capacitive and plain resistive loads at say 50W or half power, 10 kHz?
Of course, we can all eat our (paper) hats if your suspicions are right!
Drive nominally capacitive and plain resistive loads at say 50W or half power, 10 kHz?
Of course, we can all eat our (paper) hats if your suspicions are right!
I will have the champ, which is NOT the AX .... the GX , done soon. I am very methodical with a new prototype. Below I am testing the ripple fed from another amps PS (playing loud techno-nice fat ripple) through my capacitance multipliers. my simulator can not reproduce the entropy of this test. final tests confirm 2-5mV ripple. Almost matches the performance of a separate PS for the voltage stage (you can do that as well on my PB-250 -arrows).
The AX works perfectly , same sound as my first blameless.perfectly balanced LTP and under 1mv offset. I must confess one must stray from a simulation slightly , as the real components are not just models. It was VERY close to the sim's , but not absolute.
These voltage stages will be "sippin'" on some very clean power!!
OS
Attachments
Hi OS,
We'll have to talk via messaging I think. Your ideas on how an amplifier is designed are close to my own. There is a comfort there. For me, I'm not yet able to create PCBs, I'll need some help in that direction. The Symasym amplifiers turned out well, certainly better than the cheap and cheerful utility amps I had thought they were.
Every amplifier I have designed runs the voltage amp from a regulated power supply. A commercial amplifier design that does the same thing is the Marantz 300DC. One of those is worth acquiring and rebuilding. They are very pricey though, which sunk my hopes for another one of those. Clean power in the voltage amp stage matters. That's all I'm going to say.
-Chris
We'll have to talk via messaging I think. Your ideas on how an amplifier is designed are close to my own. There is a comfort there. For me, I'm not yet able to create PCBs, I'll need some help in that direction. The Symasym amplifiers turned out well, certainly better than the cheap and cheerful utility amps I had thought they were.
Every amplifier I have designed runs the voltage amp from a regulated power supply. A commercial amplifier design that does the same thing is the Marantz 300DC. One of those is worth acquiring and rebuilding. They are very pricey though, which sunk my hopes for another one of those. Clean power in the voltage amp stage matters. That's all I'm going to say.
-Chris
Hi OS,
We'll have to talk via messaging I think. Your ideas on how an amplifier is designed are close to my own. There is a comfort there. For me, I'm not yet able to create PCBs, I'll need some help in that direction. The Symasym amplifiers turned out well, certainly better than the cheap and cheerful utility amps I had thought they were.
Every amplifier I have designed runs the voltage amp from a regulated power supply. A commercial amplifier design that does the same thing is the Marantz 300DC. One of those is worth acquiring and rebuilding. They are very pricey though, which sunk my hopes for another one of those. Clean power in the voltage amp stage matters. That's all I'm going to say.
-Chris
I like "comfort" , too. Comfort that I have NO oscillation , no feature that I left out just for a few cents , and the comfort that it will last for decades. The clean power definitely pays off , these(the GX's) are noticeably cleaner and more transparent than my "syms". The AX is also on a different level than my "blameless frugalamps" , too. If you need PCB's.. chris, not only can I now clone just about anything but can make a "pretty" amp , too .... that is ROCK stable!! (below) Your my favorite mod , so I'll "hook you up".
To be on topic , the difference between LT's simulations and the real products I am now churning out is VERY minor. Gain is within a decibel , all voltage readings are within millivolts. The bode plots predicting stability were correct.. not even a warm Zobel network under any condition, and the biasing was predicted with in a millivolt (I knew where to set my trimmers before I even powered up.) In some regards , the real amp exceeded the simulation , balance between the input pair semi's were nearly perfect at ANY frequency without adjustment as seen on my Oscope.
BTW , the cap multiplier with high gain KSA/C's is VERY good at cleaning up the ripple , 2 to 3 mv's is not bad while "cranking" the amp. Saves on trafo's ,bridges and FR-4 as well.
OS
Attachments
He measured a tiny difference, but with metalic parts
having little influence since they were not strictly inside
the coil...
A very good idea which i also advocate..
Unfortunately, this increase drastically the cost/complexity
of the whole amp, unless it s a low power one, in which case
not only regulation becomes cheap, but power savings turn
also substancial, not counting the much easier thermal design...
So much advantages, yet, power hogs are the most revered...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.