On the post1 schematic, I checked it with a 110K feedback resistor made of a paralleled pair of 220K (the photos show this method of improving cheap resistors, or one could use a single high end resistor). That gain is 42X and it is definitely Not dull. This could be used with an MP3 player, phone, tablet or laptop and not need a preamp.
It can also be used for diagnosis: With the huge gain, the amplifier will increase everything, make everything really obvious and everything includes audio, dynamics, and the effect of power circuit quality.
Aha!
Now, if you've got the post1 schematic altered for high gain (as the one and only difference) and still didn't get rather pretty tone with excellent airy treble, then that tells us there is a fault with the power circuit supply or layout.
It can also be used for diagnosis: With the huge gain, the amplifier will increase everything, make everything really obvious and everything includes audio, dynamics, and the effect of power circuit quality.
Aha!
Now, if you've got the post1 schematic altered for high gain (as the one and only difference) and still didn't get rather pretty tone with excellent airy treble, then that tells us there is a fault with the power circuit supply or layout.
Last edited:
If one had installed a poor quality 220u NFB-Shunt cap, the HF loss would cut the gain on treble audio. Oops! An easy fix is to parallel that 220u with a 1u or smaller electrolytic cap, such as Nichicon ES 0.47u. An alternative fix is to replace the 220u with a paralleled pair of identical models of 100u (makes a low loss 200u) or a paralleled pair of identical models of 150u (makes a low loss 300u).
Low loss caps provide louder clearer high treble and you'll probably get airy treble.
About 10n polyester cap in parallel with your 1u input cap, can also help treble.
Low loss caps provide louder clearer high treble and you'll probably get airy treble.
About 10n polyester cap in parallel with your 1u input cap, can also help treble.
Well, after you try that, you'll be afraid not to do it.
0.47uf for electro or 10n for polyester. Nearby figures work too. But, 0.47uf polyester probably doesn't work. It is needful to choose your bypass cap that best blends in with a given model of larger electrolytic. So, the optimal is different every time. Yes, it is time consuming, but can sound really good (with some selection effort). Sound quality has many different aspects. In the case of high gain, perceived soundfield size will be smaller, sonic signature will be slightly larger, treble harmonics will be louder and dynamics will be much greater. It will be very punchy and lively. And, a phone or MP3 player will be able to direct drive a 65w power amp.
The issue is not black and white, on or off. But really you just need to choose which features you want most. The first one I made did have 42x gain, and minimal feedback current, so it is a minimum negative feedback example. It is possible that you might like that, and finding out is really easy. It is probably suitable for a party.
Nope.
The only dramatic difference is that its a lot louder.
The other differences aren't vast.
Have you tried it yet?
The only way to thoroughly answer the sound-related question is do it and hear it. Either you like the big wide sound of a reserved mature low gain amp, Or you like a dynamics amp's gleeful party voice. It is a matter of personal preference. All of audio has compromises. Therefore, there are compromises if you do high gain and there are different compromises if you don't do it. "No free lunch," applies. The amplifier has many possibilities for different benefits and all of them come with a different sounding compromise. I think that you need to do some exploring to find out which you like. The soldering iron can help you answer this question fast. Unfortunately no amount of typing could be sufficient.
The only dramatic difference is that its a lot louder.
The other differences aren't vast.
Have you tried it yet?
The only way to thoroughly answer the sound-related question is do it and hear it. Either you like the big wide sound of a reserved mature low gain amp, Or you like a dynamics amp's gleeful party voice. It is a matter of personal preference. All of audio has compromises. Therefore, there are compromises if you do high gain and there are different compromises if you don't do it. "No free lunch," applies. The amplifier has many possibilities for different benefits and all of them come with a different sounding compromise. I think that you need to do some exploring to find out which you like. The soldering iron can help you answer this question fast. Unfortunately no amount of typing could be sufficient.
Last edited:
NFB current and Compensation
Compensation:
Voltage Gain is one way to do external compensation and higher gain is more likely to be stable. This another reason I suggested that you try it. See if your power supply lights take a few seconds to go out rather than go out immediately. See if there's a change in the difference between heatsink temp and ambient temp.
Questions:
Do you like 110k with 2.7k+220u or do you like 62k with 1.5k+440u? They're very similar for voltage, so the gain is very similar (stability is similar, loudness is similar); however, the latter example with smaller figures is higher negative feedback current. Due to the different current, both "perceived soundfield size" (that phrase means imaging in mono--per each individual speaker) and tone will be different.
And now, the rest of the story. . .
The feedback divider can be used for tone, since a 100k pot has better treble and worse noise, a 20k pot has dull treble and low noise, and a 50k pot is a generally excellent compromise between the two extremes. This is true for both an input pot and also true for the feedback divider in most non-inverting amplifiers.
But, I theorize that we should not use the feedback divider for tone because:
Using the feedback divider for tone compromises the soundfield size. Instead, these two resistors are best used to get the soundfield size nice and big. Opportunities to install that are limited, but opportunities for tone are not limited, so. . . then work out the tone needs elsewhere, such as with little bypass caps and well done power circuit.
The real questions:
How much gain do you need for stability?
How much can you level the harmonic tone with all other parts of the amp (especially including small signal bypass caps--treble helpers) before having to resort to using the feedback divider settings to improve the tone?
A third option to explore:
Average values tend to land somewhere in the right ballpark, but that is not especially informative except as a repair method (when everything is "in the ballpark" you can more easily discover and fix an error).
SO, 68K with 2.2K+(330u||1u both electrolytic). This is an approximately averaged value, just barely edged towards clearer bass and treble. The gain is a useful albeit not thrillingly high 32X. This is enough to get 30 watts of reasonably clean power from a computer source, prior to the source straining itself, which is not fun to amplify. So, if you want to rock the house or if you have a digiplayer source, you'd need a preamp; however, if you want to use the extra for headroom, then you don't need a preamp.
If you made the original build, with 60k trapped inside, a similar option to explore is 60k with 1.8k+(440u||10n), which is 34X, good feedback current for big soundfield, and the bass should be high quality; however, the treble is worked out via little bypass caps as needed, which can be excellent depending on what you do.
Exploring higher and lower from centerpoint ("edging" textbook values) is not as quickly informative as trying low and high first. Either way can work. The slower textbook method has the advantage of discovering if a trouble might be located elsewhere in the circuit, so we can avoid fixing the wrong spot. The main point is that if you're displeased with the amp, then explore its options and set it to work better for your needs and your application.
EDIT: If this doesn't do it, we'd need to consider speaker repair and/or active boosters.
Even more people distance from negative feedback, and higher gain with high feedback resistor value = minimal negative feedback. The main cost is needing clean power. I liked it and left one that way for 5 years.
Compensation:
Voltage Gain is one way to do external compensation and higher gain is more likely to be stable. This another reason I suggested that you try it. See if your power supply lights take a few seconds to go out rather than go out immediately. See if there's a change in the difference between heatsink temp and ambient temp.
Questions:
Do you like 110k with 2.7k+220u or do you like 62k with 1.5k+440u? They're very similar for voltage, so the gain is very similar (stability is similar, loudness is similar); however, the latter example with smaller figures is higher negative feedback current. Due to the different current, both "perceived soundfield size" (that phrase means imaging in mono--per each individual speaker) and tone will be different.
And now, the rest of the story. . .
The feedback divider can be used for tone, since a 100k pot has better treble and worse noise, a 20k pot has dull treble and low noise, and a 50k pot is a generally excellent compromise between the two extremes. This is true for both an input pot and also true for the feedback divider in most non-inverting amplifiers.
But, I theorize that we should not use the feedback divider for tone because:
Using the feedback divider for tone compromises the soundfield size. Instead, these two resistors are best used to get the soundfield size nice and big. Opportunities to install that are limited, but opportunities for tone are not limited, so. . . then work out the tone needs elsewhere, such as with little bypass caps and well done power circuit.
The real questions:
How much gain do you need for stability?
How much can you level the harmonic tone with all other parts of the amp (especially including small signal bypass caps--treble helpers) before having to resort to using the feedback divider settings to improve the tone?
A third option to explore:
Average values tend to land somewhere in the right ballpark, but that is not especially informative except as a repair method (when everything is "in the ballpark" you can more easily discover and fix an error).
SO, 68K with 2.2K+(330u||1u both electrolytic). This is an approximately averaged value, just barely edged towards clearer bass and treble. The gain is a useful albeit not thrillingly high 32X. This is enough to get 30 watts of reasonably clean power from a computer source, prior to the source straining itself, which is not fun to amplify. So, if you want to rock the house or if you have a digiplayer source, you'd need a preamp; however, if you want to use the extra for headroom, then you don't need a preamp.
If you made the original build, with 60k trapped inside, a similar option to explore is 60k with 1.8k+(440u||10n), which is 34X, good feedback current for big soundfield, and the bass should be high quality; however, the treble is worked out via little bypass caps as needed, which can be excellent depending on what you do.
Exploring higher and lower from centerpoint ("edging" textbook values) is not as quickly informative as trying low and high first. Either way can work. The slower textbook method has the advantage of discovering if a trouble might be located elsewhere in the circuit, so we can avoid fixing the wrong spot. The main point is that if you're displeased with the amp, then explore its options and set it to work better for your needs and your application.
EDIT: If this doesn't do it, we'd need to consider speaker repair and/or active boosters.
Last edited:
Fault detected!
Check amplifier for DC offset.
Check resistor values for standby and mute.
If no errors found at amp, then. . .
Check drainer resistors to see if one broke.
Transformer or bridge rectifier may have a fault.
(fault could be lopsided load or lopsided ac or bad bridge rectifier)
p.s.
It is also possible to have an "almost" failed shorted cap in the power supply board and/or amp power cap and in this case we'd have to swap caps one at a time until the bad cap is found. The very resistant, partially failed cap acts a lot like a backwards cap. We'd begin repairs by changing out whichever caps are the warmest, since erroneously shunting the power supply makes a little bit more heat.
Last edited:
They should not go off instantly except for when the amplifier is playing loud (high workload) at the time mains are disconnected.
More importantly, the LED's should turn off at approximately the same time.
You can create a 50/50 voltage divider with identical pair of resistors in series and that connects from V+ to V-, whereby centerpoint is virtual ground (perfectly half of the total) and then measure from virtual ground with your voltmeter to 0v to measure how many volts of asymmetry error.
SO, this information is that the asymmetry happens when there is a load attached to the amp, but not when the load is removed.
Asymmetry when loaded:
The bridge rectifiers and transformer need checked.
Also check the PI filters to see that the resistors are healthy and their bypass diodes are in useful orientation.
When there is load there is just 0.1 or .2 V difference.The diode is opposite to each other.I think everything is fine but..
My amp gives 12, or 11mv dc offset.
Interesting. I wonder why the front left speaker would resist far more in one direction than the other (asymmetric load). Perhaps it has a bad cap in the crossover? I do not know. But, you can find out by swapping parts from the left speaker to the right speaker until you find the one part that also relocates the problem, and then that part needs replaced.
Last edited:
I think my drivers are old needs replacing.I have remove the tweeter cap ,,but still the same problem..I will try change a speaker in the same cable,,,or i will change the cable..or swaping of the two drivers...every trick...but I think the speaker ,woofer is the trouble.
Anyway Im getting adapted to this tda7294 sound.I think i love the sound.
Edit; Now I know the uses of the LED
Last edited: