Hi Everyone,
I redesigned a LM1875-based guitar amp that I'd previously made with a few changes as it popped on turn-on due a small issue.
I redesigned the circuit with a few changes and had them made. I've populated the circuit, put it inside a new enclosure and used the same SMPS (Meanwell LRS-36-100) as I did for the original. When the amp if turned on, I'm met with a lot of noise and barely any output from the amp itself. Whatever I do get out of the amp seems to distort extremely easily. Below are a few issues I've noticed -
a few things I've tried -
Does anyone have an idea what it could be and what my next steps could be? I'd love to probe the power supply, but I only have a DSO Nano V3 which isn't great for measuring ripple.
Ive attached a picture inside the chassis and the schematics.
Thanks
I redesigned a LM1875-based guitar amp that I'd previously made with a few changes as it popped on turn-on due a small issue.
I redesigned the circuit with a few changes and had them made. I've populated the circuit, put it inside a new enclosure and used the same SMPS (Meanwell LRS-36-100) as I did for the original. When the amp if turned on, I'm met with a lot of noise and barely any output from the amp itself. Whatever I do get out of the amp seems to distort extremely easily. Below are a few issues I've noticed -
- Noise / Hum (White Noise) without input connected. Input / Output Jacks are Cliff and the input is shorted without connection.
- Increased noise with an input connected (This is due to a ground loop, but I'm using the same setup as was previously used without any noise)
- Turning the volume pot changes the amount of noise, but also at times causes the noise to increase until it gets unbearable.
- Probably unrelated to the noise, but the mute/delay works for power up, but the output cap's supply collapses before it de energiese and causes turnoff pop.
a few things I've tried -
- Changed the PSU board to the old one that had a regulator for both the opamps and the amplifier.
- Soldered a second board up to make sure it wasn't an issue with a component.
- Tried different power socket.
Does anyone have an idea what it could be and what my next steps could be? I'd love to probe the power supply, but I only have a DSO Nano V3 which isn't great for measuring ripple.
Ive attached a picture inside the chassis and the schematics.
Thanks
As far as I can tell by looking at the schematics, there doesn't seem to be any local decoupling of the 12 V supply for the TL072. It might be oscillating due to that. Does a 100 nF capacitor from the TL072 supply pin to ground change anything?
One question: do you hear hum (100Hz) or hiss (White noise)? If it is hum, it’s likely due to a ground loop rather than an issue with the circuit itself.
You could try to :
On Amp side
On PSU side
You could try to :
On Amp side
- Install the 100nF as recommended by Marcel
- Reduce R1 & R2 to 100k
- R3 could be shorted, but you’ll have to change R1 & R2 on supply side to 68k
- You need a cap between RV1 and positive input of AmpOp, otherwise it’s biasing is related to RV1 position, which may lead to unexpected distortion.
- with R1=100k ( and not 1M) a 100nF should be alright
- add a cap on input to avoid catching HF signals
On PSU side
- Install a 10uF ecaps to ground on RV1, on R4 and on R1
- Replace C2 by a ceramic rather than a ecap
Thanks guys. I added the 100nf at the input and the ceramic on the tl072 and it is markedly improved. There still a rather loud hiss (white noise / raised noise floor).
The input coming through is still distorted, and when a loud note is played all audio stops and there seems to be a slowly increasing noise, which might be oscillation?
Stupidly overlooked the parallel resistors, which is probably screwing up the bias of the second opamp. Does this sound about correct?
The input coming through is still distorted, and when a loud note is played all audio stops and there seems to be a slowly increasing noise, which might be oscillation?
Stupidly overlooked the parallel resistors, which is probably screwing up the bias of the second opamp. Does this sound about correct?
I'll try that tomorrow. Thank you both for the help so far. As for the output pop, is there anything you can see there?
Ok, is this because guitar microphone are hi impedance? I now remember that my Focusrite input impedance when set to 'instrument' is 1,5M...
Yes.
A guitar pick-up typically has an inductance of about 2 H, or even higher when it is a so-called humdinger. There is usually a 250 kohm to 500 kohm potmeter in the guitar between the pick-up and the cable to the amplifier, and another potmeter connected as a variable resistor with a small series capacitor shunting the pick-up. Or at least that's what I've been told in other electric guitar threads.
A guitar pick-up typically has an inductance of about 2 H, or even higher when it is a so-called humdinger. There is usually a 250 kohm to 500 kohm potmeter in the guitar between the pick-up and the cable to the amplifier, and another potmeter connected as a variable resistor with a small series capacitor shunting the pick-up. Or at least that's what I've been told in other electric guitar threads.
I will have to redo the PCB. You were correct. I checked the voltage into the second opamp and indeed it's sitting at 0.5v which correlates with the voltage divider created by the 1meg and the 7.5k.
It's all learning I guess! Thanks for the help
It's all learning I guess! Thanks for the help
If you redo the pcb, I encourage you to locate the biasing resistors divider and their filtering cap on the Amp board, not the psu, and close the input of the opamps and LM1875.
In addition, those 4 GND connection here under should be connected to GND Power, not GND Audio. GND Audio and Power should be connected together on a single point. It seems you are using Kicad; use a special component called NetTie. It is same as a zero Ohm resistor. As it has two nets, Kicad helps you to route them separately, thus avoiding high current of power ground flowing into signal ground and inducing distortion.
You could also increase C2, 1000u won't harm.
A NetTie :
In addition, those 4 GND connection here under should be connected to GND Power, not GND Audio. GND Audio and Power should be connected together on a single point. It seems you are using Kicad; use a special component called NetTie. It is same as a zero Ohm resistor. As it has two nets, Kicad helps you to route them separately, thus avoiding high current of power ground flowing into signal ground and inducing distortion.
You could also increase C2, 1000u won't harm.
A NetTie :
Thank you @AIM65.
I see why the Zobel would get connected to power, but why would the others? Not disputing, Just wondering.
I thought anything in the Audio Path would require them to go to signal ground.
Would it also be wise to remove that 100 ohm resistor as the cap is already tied to the ground with that relay config?
I see why the Zobel would get connected to power, but why would the others? Not disputing, Just wondering.
I thought anything in the Audio Path would require them to go to signal ground.
Would it also be wise to remove that 100 ohm resistor as the cap is already tied to the ground with that relay config?