Thread for random DIY-related questions

[crochambeau]

In that diagram the signal enters and exits the 4053, which is a CMOS circuit (not a physical connected switch) so that is not true bypass.

In my mind, anything that requires the power supply to be connected in order to complete a bypass signal path is not "true bypass".

That said, I use 4053 the switch signals and have no complaints. I would not sweat "true" anything - unless it's a client directive, these objects we make are ELECTRONICS.

I find each pass through a 4053 is roughly 100 ohms (it's a non-conventional rule of thumb I use, I've measured lower for sure) series resistance on the signal path, so the switches can get lossy if you're not planting any gain in the mix.

[crochambeau]

If you're goal is an A/B comparison fixture, I would say a relay would be a better switch source.

[FAP]

Yeah I figured as such I’m reluctant to try relay bypass just because I would have no idea which kind(s) to get; in theory it makes perfect sense but I’m typically wary of electro-mechanical solutions due to wear & tear… granted a toggle switch is also electro-mechanical: guess I can’t really win on that front

[crochambeau]

Relays should experience minimal wear and tear - unless you're switching something with DC on it - but yeah, it's basically just a switch. The real benefit is being able to throw multiples with a single command, or automate stuff (though relays switch slowly compared to CMOS) - CMOS can switch at audio rates.

[FAP]

Trying to make an “interface” for my Zoom recorder. Basically, I need something that will attenuate the signal going into each individual input on the Zoom (in 4CH mode, this means IN1, IN2, and the EXT. MIC input serving as an additional stereo input). I could reduce the input gain on the zoom, of course: I’m just not sure I trust it. Will this circuit work, or would it lead to signal bleed (even if each channel is separated/not mixed together)?

Viewed 3562 times

[crochambeau]

Trying to make an “interface” for my Zoom recorder. Basically, I need something that will attenuate the signal going into each individual input on the Zoom (in 4CH mode, this means IN1, IN2, and the EXT. MIC input serving as an additional stereo input). I could reduce the input gain on the zoom, of course: I’m just not sure I trust it. Will this circuit work, or would it lead to signal bleed (even if each channel is separated/not mixed together)?

That should work fine, the only thing that could cause bleed is if common/ground is resistive or floating - so if you do wind up with crosstalk that's where I'd look first.

[FAP]

Don’t have a schematic for this one at the moment so I’ll try my best to explain:
I hooked up four astable 555-based LED flasher circuits: none of them are connected together, and yet they will inevitably flash in sync with each other, seemingly based on which one is set at the highest frequency (e.g. if one is flashing at 16Hz, the others will as well).
Each circuit has identical components: the only variable is each one has a pot that adjusts output frequency (tl;dr less resistance = higher frequency).
My guess is this has something to do with the highest frequency effectively creating a path of least resistance, such that despite the fact that each LED comes from a different output (on a different 555), they are all ultimately connected to ground.

Can I do anything to avoid this? I feel like I was able to do this before without issue: why is it happening now? Would using a virtual ground/dual power supply help or is that not the issue?

[crochambeau]

They are probably coupling up/soft syncing through your power rail.

If you're using 555 you could try to add a little resistance (even a diode) and a filtering capacitor for each section. If you're using a 558 you're pretty fucked, as whichever section is doing the most work will put ripple on the power rail and everything else will kinda fall into step.

[FAP]

They are probably coupling up/soft syncing through your power rail.

If you're using 555 you could try to add a little resistance (even a diode) and a filtering capacitor for each section. If you're using a 558 you're pretty fucked, as whichever section is doing the most work will put ripple on the power rail and everything else will kinda fall into step.

Thanks. Where would the resistor/diode+filtering cap go? On the output?

[crochambeau]

Thanks. Where would the resistor/diode+filtering cap go? On the output?

Not on the output, on the DC voltage rail feeding the circuits. Here's a screen grab of the power rail decoupling I use on the Crustacean. I was running into issues in which one particular section would couple through the power rails and instigate feedback with the signal source. By branching the power rail and doing some "localized" power filtering I can manage to push problem sections harder than I can when they all drink from the same pool, so to speak.

Viewed 3160 times

The value of resistor will depend on how much power that stage draws. Larger resistors are more effective at centering any push and pull on the power rail onto the capacitor serving as DC reservoir/AC sink after that resistor, but if the power consumption on that branch is too high you wind up having to balance filtering with voltage drop. I have used inductors in higher draw sections, but the depicted build is low current enough that 100 ohms does not incur a substantial voltage drop across that resistor.

Using diodes in this position remains somewhat experimental to me, and was probably bad (unproven) advice on my part.