Ok, this is maybe a bit radical, but I ordered opamps with 3 different suppliers at aliexpress to spread the risk of getting fake ones (and not having to wait for another few weeks for new ones to arrive).
6 Likes
My first ever self-designed PCBs arrived today!
It’s a carrier board for the Daisy Patch submodule (as pictured) which has taken inspiration from a couple of sources. Looks OK, everything fits, so I’ll solder up a minimal one to test before going all out.
My second ever self-designed PCBs are en route as well…
14 Likes
My second ever self-designed PCBs!
These are simple buffer boards to make building things easier. The standard two-inverting-amplifiers-in-series, so two channels from a TL074. Between the first and second PCBs I learnt a lot about Kicad!
I wish I’d made the resistor footprints a little bigger, that’s the only thing. The dimensions seemed sound on paper but when I started filling my Daisy board it turned out to be a bit too tight for some of my stock.
5 Likes
Why inverting?
If the capacitors are for stabilization then the appropriate value will depend on the resistor values. In principle bypass capacitors would be a good idea too.
I used to use R_Axial_DIN0207_L6.3mm_D2.5mm_P7.62mm_Horizontal (for the standard size Royal Ohm 1/4 W resistors) but bending over the leads that close to the body isn’t good practice. I switched to R_Axial_DIN0207_L6.3mm_D2.5mm_P10.16mm_Horizontal.
1 Like
I’ll do better with the capacitors next time! I’m still at a stage where I’m throwing things around and seeing what sticks, to be honest.
Why inverting? I’d seen a few schematics for more mature projects (i.e. made by people with a better idea than me) with this arrangement and it stuck in my mind. I also wanted to be able to use these for active attenuators as well, and of course non-inverting arrangements have a minimum gain of 1. (I put the note about the gain setting so that Future Steve wouldn’t forget this!)
Inverting keeps the inputs at ground, and therefore, safely far away from the TL0xx series problem with phase inversion when the input gets below about 4 volts above the negative rail. Not that anyone would apply such a voltage… . And anything rediculous going to the input will have to go through a large resistor, so even 220vac may be survivable. Try that connected directly to the non-inverting input of an opamp and see which end the smoke comes out. Have you seen the post at Design mistakes in synth schematics - North Coast Synthesis Ltd.? It covers a lot of ground of how to bullet-proof module ins and outs, and make sure that , e.g., a gate output on module “A” will be recognized on module “B”, for all “A” and “B”. Recommended reading!
There are indeed benefits to both.
If you’re dealing with input voltages close to (but within) the negative rail, or very close to the positive rail, inverting is certainly better. But such voltages are fairly uncommon with synth modules.
As for gain > 1, you can put a voltage divider before a non inverting amplifier. It’ll reduce the input impedance, if that matters, but no more so than an inverting amplifier.
It’s true a non inverting configuration is more sensitive to overvoltage. But in an inverting configuration, 220 V will almost certainly send the voltage on the inverting input pin way beyond the voltage rail and hence the absolute maximum rating for the device. With a large input resistor perhaps it would indeed survive nevertheless. Then again, with a non inverting configuration you can always add a current limiting resistor into the non inverting input if you’re worried about it; normally there will be no current but in an overvoltage situation it might help. Then again again, if you have 220 V on your patch cable, you’ve got bigger problems.
If you have multiple signals, you can sum them (times gain) with an inverting configuration, even if the sum (without gain) exceeds the rail voltage, and still keep the input pin voltage at zero. With non inverting you have to put them into a resistor node (reducing the input impedance) and then you get the average, with the input pin voltage nonzero. In most multi-input instances inverting is probably preferable, especially if one or more inputs might or might not be connected.
All in all the choice depends on the situation, but for a single input within -5 V to +10 V (assuming ±12 V rails) where an un-inverted output is desired, I will usually prefer a single non inverting stage to two inverting stages.
I’m familiar with that article, it’s an excellent one!
2 Likes
Another PCB order
PCBs for three modules, panel for a fourth one. I’m modifying how I do panels so I wanted to just do one and see how it worked out before going ahead with others.
9 Likes
deleted post- bad edit
1 Like
Oh, I will once I start using them, which I haven’t so far. So, patience my friend …
2 Likes
Stuff arrived
Mostly common components, the most unusual of the lot is a pair of Chinese clones of VTL5C Vactrols, so I can compare their behavior to my DIY builds.
And a new PCB I designed! It’s very similar to @analogoutput 's protoboard, with a few key differences: first there’s two per board, to take advantage of the deals on 100×100mm PCBs, and second, it’s built to make it easy to to use the power rails with solder bridges, to minimize the wiring required. The trade-off is that you get one hole fewer per vertical group, and now center rows to run arbitrary signals through.
Here it is manually cut in half, and compared to similar boards.
I’m also awaiting busboard PCBs following the same principle! They should be in a box somewhere in China right now.
(disregard the wrong trace, image is from a wip i posted elsewhere)
I’ll release the files once I make a project with it and confirm the design isn’t flawed.
3 Likes
100k’s never last long do they.
2 Likes
Blue box happened
I wanted to get as much use as possible out of a single 100×100 board, so I designed them like the protoboards I have just made a thread about: you can cut them wherever you want, but if you use them intact, they’re all connected.
I don’t really have an immediate use for them, but I will probably try to put one together, just to make sure it’s OK.
Rev 1 has at least two little mistakes, can you spot them? On the third row, the silkscreen has the +12V and -12V rails flipped, and on the fourth, the +12V label is in the wrong column. I changed my mind about the orientation of the headers while designing the board!
3 Likes
