EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: ledtester on August 09, 2021, 01:45:50 am
-
I have a MidiMan MiniMixer - a simple TL062-based mixer powered by a 9V battery.
When you turn it on and off it creates a popping sound on the output. Is there a modification to the power control that would prevent these pops? Maybe something that ramped the power supply voltage up/down?
A pic of the innards:
[attach=1]
-
This can be caused in a few ways. Ramping up the power can help. Disconnecting the output until after the pop may be another way. If you have a circuit diagram it might help to see the reason and thus eliminate the problem.
-
You are going to have to add circuitry to fix this problem. Turn-on pop can be eliminated with a mosfet + RC timer (to slowly turn on the power supply OR slowly turn on the speaker output) + diode (to make it reset faster). Turn-off pop is a bit harder and will probably require a comparator (you have to detect the power supply voltage starting to drop & act fast).
-
Popping is always a problem ito a lesser or greater degree in any unprotected audio amplifier.
I think that because your power supply is single ended, that makes it worse.
With the battery removed, check continuity between the black battery lead and the "ground" contact of all the input and output connectors.
If it's all zero ohms (which I expect) then we know what we're dealing with.
If this is true, the main output of the op amp floats at 4.5 V and is connected to the output through a big cap.
This is going to give you a tremendous pop.
I personally would try to re-engineer the circuit.
Use the "rail splitter" as "ground" for the inputs and outputs.
Make sure that the rail splitter has two caps, going low and high.
Yes, you can add mute circuits too.
-
simplest option is to not to power your amps until your mixer is powered and stable,and power down your amp before killing your mixer,in days of old i was taught to power up in the direction of the signal flow,so input devices -mixer-amps and power off in reverse.
-
simplest option is to not to power your amps until your mixer is powered and stable,and power down your amp before killing your mixer,in days of old i was taught to power up in the direction of the signal flow,so input devices -mixer-amps and power off in reverse.
N.B. that only works if your amps have turn on/off anti-pop themselves.
-
one solution is to put a small relay on mixer OUT, relay designed for audio signal or why not normal one, and delay his NO contact. that's gonna eat your battery in 5 mins |O
but this can work with power supply instead of battery...
-
Any schematic?
Scope trace of the turn-on transient?
-
...
With the battery removed, check continuity between the black battery lead and the "ground" contact of all the input and output connectors.
If it's all zero ohms (which I expect) then we know what we're dealing with.
...
Yes, the battery negative terminal is connected to the sleeve of both inputs and outputs.
Any schematic?
It's hard to extract the PCB from the case... it doesn't seem to have been designed for easy removal:
[attach=1][attach=2]
I think the PCB was inserted at an angle (the switch side first) and then the jack side was snapped into position.
I did find this verbal description at https://groupdiy.com/threads/midiman-mini-mixer-mod-advice-needed.44685/
I have not a schematic, sorry, but I try to make it in simple words:
- input jacks to opamp inverting inputs via 100K res. and mylar .22 cap. in series
- feedback res. 100K (unity gain)
- non-inverting inputs biased to 1/2 V with couple 100K res. in series + a .22 mylar filter cap between 1/2 V and ground
- opamp output to out jacks via 100R and 5uF bipolar cap. (two back-to-back 10uF electros), plus 100K to ground in parallel to output jacks
- internal 9V battery, no power filtering, no external power socket.
A response on that forum page suggests just using a passive resistive mixer -- like a 22K for each input. My speakers are powered -- they are Yamaha HS7's -- and all of my inputs have volume controls.
-
I'd still go with my suggestion, although modifying everything could be a bit tricky.
The basic problem is the unipolar power supply.
You'd need to disconnect the two +IN of the opamps from the split voltage and connect them to the same ground as all the connectors.
You'd need to disconnect the V- power of the opamps and have it run isolated to the battery minus.
If you want to still run it off a single 9V battery, then you:
Need to add a more powerful splitter that goes between switched battery + and - and the middle connects to the connector grounds.
Maybe two 10µF caps and two 2.2k resistors?
-
Renate's solution may work, but it requires modifications to the circuit and the more powerful rail splitter needed to drive the output may increase power consumption unless you use something "smart" like the famous TLE2426 or a DIY equivalent.
Here's what happens right now: as the power is turned on, the 2×100kΩ divider acts as a 50kΩ resistor biased to 4.5V (Thevenin equivalence) and charges the 220nF capacitor at IN+. Time constant is 11ms so the process takes a fraction of a second to complete and the initial peak rate of rise of IN+ voltage is 4.5V/50kΩ/220nF or some 400V/s. The opamp wants IN- to follow IN+ and therefore it makes the output to also rise at 400V/s (actually, OUT is even slightly higher than IN- because the input caps charge up towards 4.5V and "steal" some current from the 100kΩ feedback resistor). Anyway, assuming peak 400V/s rise at OUT, we get peak 2mA current through the 5µF output coupling caps which can produce a bit of voltage across high impedance inputs of dowstream gear.
A further complication is the TL062 opamp, which has phase reversal: as long as IN+ stays below some threshold (about 2V?) the output is driven hard to the positive rail. I suspect that this contributes greatly to the amplitude of of the turn-on transient and makes it sharp, fast, and loud rather than a dull "thump".
If not for phase reversal, you could simply increase the divider resistors or the IN+ cap and the turn-on process would slow down, reducing the amplitude and frequency content of the transient. But because of phase reversal, such mod might only make things worse. The correct solution would be to modify the circuit so that all capacitors are referenced to the positive terminal of the battery, which is what the original designers should have done if they chose such opamp |O
But you also came up with slowing down the rate of rise of the supply voltage. I think this may work. It achieves two things: the rate of rise of IN+ voltage is slowed down of course, and the rate of rise of OUT voltage during phase reversal is slowed down too. I suspect that increasing the supply ramp-up time to 1 second would mostly eliminate transients while still not being much time to wait.