Hi, LM1875 circuit:
The 1k/ 330p network is a low pass 482kHz I guess. Is it a good idea to tweak it to around 10kHz?
The datasheet states that the chip is stable at gain >= 10.
Is it ok to lower it to 6 or 4?
The preamp levels are high and I don't need gain of 10.
The circuit is intended to work in the range 2k5 - 10k. A HF driver for a bass guitar cabinet, 2k5 active crossover before the amp.
I want to minimize all hearable HF noise, introducing LPF and reducing the gain.
The 1k/ 330p network is a low pass 482kHz I guess. Is it a good idea to tweak it to around 10kHz?
The datasheet states that the chip is stable at gain >= 10.
Is it ok to lower it to 6 or 4?
The preamp levels are high and I don't need gain of 10.
The circuit is intended to work in the range 2k5 - 10k. A HF driver for a bass guitar cabinet, 2k5 active crossover before the amp.
I want to minimize all hearable HF noise, introducing LPF and reducing the gain.
To minimize gain and output noise, use a step-down output transformer. Seeing as it only needs to handle 2.5kHz and up, it can be a fairly compact ferrite cored trafo.
So it might be gain 11:
- swapping the feedback resistor to 10k || 1500p for ~10k LPF.
And I can do the feedback resistor cap on few other op-amp places in the preamp section, to get 2-nd and 3-rd order LPF.
- swapping the feedback resistor to 10k || 1500p for ~10k LPF.
And I can do the feedback resistor cap on few other op-amp places in the preamp section, to get 2-nd and 3-rd order LPF.
The 1kΩ and 330pF combination forms a low-pass filter with a cutoff frequency fc. fc=1/2PiRC=1/2pi*1K*330pF=482 KHz(approx.)
If your goal is to limit the bandwidth to approximately 10kHz (to avoid HF noise), you can tweak the RC network. For a 10kHz cutoff, adjust the capacitor value while keeping the resistor constant:
C=1/2*pi*R*Fc=1/2*pi*1k*10K=15.9pF(approx)
Replace the 30pF capacitor with a 15nF or 16nF capacitor.Avoid reducing the gain below 10 to maintain stability. Here is a stereo amplifier designed with LM1875. You can take a look for a reference. https://www.pcbway.com/project/shareproject/LM1875_TDA2030_Stereo_Amplifier_23042325.html
Ouch... Then the high-frequency gain becomes 1, while it is only guaranteed to be stable for gains of at least 10.
Every change you suggested is risky and may lead to instability.
Even if you want to use it from 2.5kHz to 10kHz, amplifier itself needs to be stable way beyond that.
Why noy just add passive capacitor infront of tweeter. Simple solutions are the best.
Its a bass guitar cabinet he is working on.
Instead of active crossover and separate amp for tweeter, he could just add capacitor and tweeter to bass section, and run it full.
Capacitor calculated for 2.5kHz. No need to limit above 10kHz. What would be the point for bass guitar to limit above 10kHz? But if he insists, he can select tweeter which drops in response after 10kHz.
Insted he is trying to degrade 30 years old working chip amp well tested schematics and make it oscillate.
Instead of active crossover and separate amp for tweeter, he could just add capacitor and tweeter to bass section, and run it full.
Capacitor calculated for 2.5kHz. No need to limit above 10kHz. What would be the point for bass guitar to limit above 10kHz? But if he insists, he can select tweeter which drops in response after 10kHz.
Insted he is trying to degrade 30 years old working chip amp well tested schematics and make it oscillate.
@emosms, you could run the amp inverting at whatever gain needed and compensate the amp with the so-called noise gain compensation to run at 10x or higher at high frequencies.
See for example https://www.apexanalog.com/resources/appnotes/an47u.pdf, section 3.2
See for example https://www.apexanalog.com/resources/appnotes/an47u.pdf, section 3.2
I see. One could also use an inductor in series with the tweeter to limit the noise bandwidth, but a crappy tweeter will also do the trick.
The gain of an op-amp or similar chip-amps with a minimum gain can be reduced with a resistor across the +/- inputs. This reduces the feedback without increasing the gain. However, the input common mode range may be limited to less than the output voltage, so you have to attenuate the input to fit within the common mode range and cancel the attenuation with gain.
Last edited: