Thank you for your patience.
The only thing that I might say about the near indecipherable is that I would like to take a class in applied audio electronics engineering, even if it is online. The only problem is that is must be applied so that I can learn and remember. Certainly, a more traditional class would be even more useful, but I live in a forest with no such options in range.
Notes follow. . .
The modification of the Quasar K50 LM1875 kit, that I listed earlier in this thread, isn't speculation. There are many such modified K50 in service, most of which have replaced other amplifiers.
In an AC amplifier that's limited on maximum current, we can instead, increase output during transients with higher charge (being that the signal is AC, not a nonstop DC current). One way to do it is to cause the amplifier to slightly exaggerate the dynamics for a more lifelike sound. Its still the same signal, just expanded a bit, and who could say for certain that "more like real life" is a wrong thing? The other way for a more lifelike sound involves a preamp/predrive and a lot more current out of the power amp.
Oops! More current? We don't have that with LM1875; however, it can deliver more charge.
Thoughts:
The power caps are in series, thus the 220uF caps are a single 110uF resource. That 110uF is then, necessarily, the maximum size for NFB cap. We want the possibility, if not the fact, of flat to 20hz, and so the corresponding design eventually ends up with 47k~100k for feedback and input load (still leaving tolerances/allowances for end user to choose a variety of gain).
Going up to 100k, also with about 68uF NFB cap, in most cases, still allows for protection against the strong return force of the speaker hitting the ground plane of the amp, AND that protection IS that the amplifier board's power caps (a combined figure of 110uF, with our 220uF caps in series for split rail operation) is a larger figure than the NFB cap. Local power capacitance must be larger than NFB cap, else my example here isn't going to work as expected.
Since this is a small amplifier, in actuality, we want flat to 50hz, which is "down by 3 at 25hz" so if you aim your computerized filter calculator at flat to 25hz, giving the calculator the gain divider's figures, you'll have your NFB cap size (both the calculator and the cap will probably "fall short" thus really giving us flat to 50hz or somewhere in between). This cap size will be slightly smaller than AndrewT expects and much larger than everyone else expects. Many thanks to AndrewT for fixing my TDA7294 project's bass!
Too large or technically "just right" will make one-note Thud bass, wheras too small will make insufficient impact, loud mids!, boomy sound or absentee low bass.
Therefore, if your NFB cap is too small, you'd have a really terrible time fine-tuning your amp; however, if it is to large then you have the option of decreasing the low pitched thud via choosing a smaller size input filter cap.
And, if your NFB cap is just right for size, you'd have plenty of bass, plenty of dynamics, level response, and a much easier time with the rest of the amplifier.
The NFB cap along with the amp board's power caps IS your bass (rumble) versus drum (warm/Blur) balance.
Bigger capacitance at power supply board = more bass
Bigger capacitance at amplifier board = more drum
Bigger NFB cap = more bass, nicer mids
The amplifier board's smaller power caps have a great effect on the amplifier's sound, so chose carefully.
What you're checking for is if the amp can do a reasonable job with 1) female voice and 2) Cello (baritone voice) playback. She must not sound angry and he must not be absent.
These combination work out to:
Hi-gain LM1875, 120k feedback, 47uF nfb cap, 100k input load, -or- normal gain LM1875, 120k feedback, 68uF nfb cap, 100k input load.
Caveat:
And, if the normal gain example here operates with 47k (or slightly more) for feedback and 47k for input load, then of course, we'd double the NFB cap size. That ends up at approximately 150uF. We have now exceeded the protection range of our local power caps and need to change them to 330uF (because those in series for split rail operation are a single 165uF). If using easier to find cap sizes, the NFB cap becomes 220uF and the power caps, likewise become 470uF (a single resource of 235uF). There's no need to cut it so close, but I have merely listed minimum sizes for the amplifier board's power caps.
These examples apply when and if the speaker return force is directly wired to the amplifier board's 0v terminal or a very nearby speaker terminal.
Compare:
You do have the option to send the speaker return force into the 0v of the power supply board for a duller sound with less dynamics, and perhaps a more level response; but decreasing dynamics is probably not appropriate for a low power amplifier.
Sure, some of my comments would make no sense for larger amplifiers with larger resources; however, some hot-rodding is in order for smaller amplifiers to conserve their resources so that the AC output is strong enough to drive today's inefficient speakers. With LM1875 and a budget of 1/3rd the price of a cheap retail store amplifier, you should be able to exceed the performance of their most expensive lineup. LM1875 cannot do it on current alone, thus the mention of charge.
The only thing that I might say about the near indecipherable is that I would like to take a class in applied audio electronics engineering, even if it is online. The only problem is that is must be applied so that I can learn and remember. Certainly, a more traditional class would be even more useful, but I live in a forest with no such options in range.
Notes follow. . .
The modification of the Quasar K50 LM1875 kit, that I listed earlier in this thread, isn't speculation. There are many such modified K50 in service, most of which have replaced other amplifiers.
In an AC amplifier that's limited on maximum current, we can instead, increase output during transients with higher charge (being that the signal is AC, not a nonstop DC current). One way to do it is to cause the amplifier to slightly exaggerate the dynamics for a more lifelike sound. Its still the same signal, just expanded a bit, and who could say for certain that "more like real life" is a wrong thing? The other way for a more lifelike sound involves a preamp/predrive and a lot more current out of the power amp.
Oops! More current? We don't have that with LM1875; however, it can deliver more charge.
Thoughts:
The power caps are in series, thus the 220uF caps are a single 110uF resource. That 110uF is then, necessarily, the maximum size for NFB cap. We want the possibility, if not the fact, of flat to 20hz, and so the corresponding design eventually ends up with 47k~100k for feedback and input load (still leaving tolerances/allowances for end user to choose a variety of gain).
Going up to 100k, also with about 68uF NFB cap, in most cases, still allows for protection against the strong return force of the speaker hitting the ground plane of the amp, AND that protection IS that the amplifier board's power caps (a combined figure of 110uF, with our 220uF caps in series for split rail operation) is a larger figure than the NFB cap. Local power capacitance must be larger than NFB cap, else my example here isn't going to work as expected.
Since this is a small amplifier, in actuality, we want flat to 50hz, which is "down by 3 at 25hz" so if you aim your computerized filter calculator at flat to 25hz, giving the calculator the gain divider's figures, you'll have your NFB cap size (both the calculator and the cap will probably "fall short" thus really giving us flat to 50hz or somewhere in between). This cap size will be slightly smaller than AndrewT expects and much larger than everyone else expects. Many thanks to AndrewT for fixing my TDA7294 project's bass!
Too large or technically "just right" will make one-note Thud bass, wheras too small will make insufficient impact, loud mids!, boomy sound or absentee low bass.
Therefore, if your NFB cap is too small, you'd have a really terrible time fine-tuning your amp; however, if it is to large then you have the option of decreasing the low pitched thud via choosing a smaller size input filter cap.
And, if your NFB cap is just right for size, you'd have plenty of bass, plenty of dynamics, level response, and a much easier time with the rest of the amplifier.
The NFB cap along with the amp board's power caps IS your bass (rumble) versus drum (warm/Blur) balance.
Bigger capacitance at power supply board = more bass
Bigger capacitance at amplifier board = more drum
Bigger NFB cap = more bass, nicer mids
The amplifier board's smaller power caps have a great effect on the amplifier's sound, so chose carefully.
What you're checking for is if the amp can do a reasonable job with 1) female voice and 2) Cello (baritone voice) playback. She must not sound angry and he must not be absent.
These combination work out to:
Hi-gain LM1875, 120k feedback, 47uF nfb cap, 100k input load, -or- normal gain LM1875, 120k feedback, 68uF nfb cap, 100k input load.
Caveat:
And, if the normal gain example here operates with 47k (or slightly more) for feedback and 47k for input load, then of course, we'd double the NFB cap size. That ends up at approximately 150uF. We have now exceeded the protection range of our local power caps and need to change them to 330uF (because those in series for split rail operation are a single 165uF). If using easier to find cap sizes, the NFB cap becomes 220uF and the power caps, likewise become 470uF (a single resource of 235uF). There's no need to cut it so close, but I have merely listed minimum sizes for the amplifier board's power caps.
These examples apply when and if the speaker return force is directly wired to the amplifier board's 0v terminal or a very nearby speaker terminal.
Compare:
You do have the option to send the speaker return force into the 0v of the power supply board for a duller sound with less dynamics, and perhaps a more level response; but decreasing dynamics is probably not appropriate for a low power amplifier.
Sure, some of my comments would make no sense for larger amplifiers with larger resources; however, some hot-rodding is in order for smaller amplifiers to conserve their resources so that the AC output is strong enough to drive today's inefficient speakers. With LM1875 and a budget of 1/3rd the price of a cheap retail store amplifier, you should be able to exceed the performance of their most expensive lineup. LM1875 cannot do it on current alone, thus the mention of charge.
The LM1875's can also be entertaining on single rail power with buffer, input cap, NFB cap, dual feedback, and output cap. In this case, the buffer is not totally at unity, but you'll need a unity stable chip as would be used in a buffer. The buffer chip and LM1875 have their own feedback, with the LM1875 set at max useful gain, which is 45. The buffer chip (used as a low gain preamp) has an additional feedback resistor connecting to the LM1875's output cap. In this way, only the buffer chip's own signal caps will affect fidelity and the amplifier's signature will be that of the buffer chip itself. Here we have an example of a high-gain power amp able to boost the slightest hint of dynamics up to massive scale. In practice it takes four such examples in a parallel + bridged arrangement to make a satisfactory amplifier for home theater use.
In this example, the preamp (a unity stable op-amp with capacity for buffer use--Very stable) can be run from its own, small, clean, regulated, power supply. Since the preamp controls the global feedback, the clean sound applies globally (to the entire amplifier).
This is an example of much less time spent fine tuning, with much more time spent soldering.
In this example, the preamp (a unity stable op-amp with capacity for buffer use--Very stable) can be run from its own, small, clean, regulated, power supply. Since the preamp controls the global feedback, the clean sound applies globally (to the entire amplifier).
This is an example of much less time spent fine tuning, with much more time spent soldering.
Last edited:
Daniel, applied audio electronics engineering classes are few and far between. What you should do is go to your local two year Jr. collage and take the basic AC/DC electronics engineering course. Most of what you need to know will be in this course (and you need to know this stuff before you take an advanced course). Maybe you can get your employer to pay for part of the class.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.