I don't need color prints, just black or white depending on the color of the panel (anonized black or brushed aluminium) Printing at all is going to make things more complicated because machine shops don't do it and they will have to bring in a subcontractor, regardless of whether it's colour or no...

The critical issue is printing. If you don't need them printed, there are a lot of Chinese factories that can make machined panels for you. But if you need them printed, well... When I started out a few years ago I contacted a number of Chinese companies through AliBaba and tried to get quotes on pa...

Thanks for the plug, KSS. Something I'd add is that as a largely DIY vendor, I'd rather have people buy full kits because those give me the chance to make a small profit on the parts and get paid for the sourcing work I'd be doing anyway to support my own builds. The percentage margin on PCB+panel i...

My DIY kit sales are up. Assembled sales are down. I don't offer "PCB and panel" and don't feel I can afford to - with the smaller price tags on those, and people opting for them instead of kits, despite a higher profit margin on PCB/panel I'd have to sell thousands to make a living and the market i...

Also, very often a "relaxation" oscillator will be one where the discharge switch and voltage detection are done in a single component - like a neon lamp or a unijunction transistor. These aren't common anymore but were very popular, especially for hobbyists, back in the day when components were mor...

I haven't used svg2shenzhen, but I think the folder full of .mod files is probably what you want. It should be usable as a footprint library in KiCAD's PCB editor. KiCAD uses the term "module" sometimes (including to abbreviate for the filename extension) where it means what most people call a PCB f...

htor wrote: ↑

Fri Dec 04, 2020 3:32 pm

really enjoy reading your articles. thanks for writing them :tu:

Thank you!

Yes, it's a standard series-resistor LED drive circuit. You'll need to find the forward voltage of the LED; subtract that from the driving voltage that you expect in the normal case to find the voltage drop across the resistor; choose how much current you want through the LED in the normal case; and...

Good point, hadn't thought of protecting outputs from potential inputs being patched. What's a good practice to go about doing this? I know of limiting voltages to a range using a circuit with Schottky diodes (from the Doepfer A-100 DIY page) but I'm assuming there is an easier way to do this with ...

An LED is unlikely to add noise, especially not on a logic signal, but you may need to protect the LED itself against the possibility of someone patching an unusually high or low voltage into that output jack. People sometimes overlook that when designing output circuits.

I only noticed the first one because I was for a time trying to figure out how to make a submini toggle that was (off)-on-off. For the matrix switch bus splits of my new format. I wanted the middle to be on so the bat was vertical in what would be considered the default state. Then one way would be...

You are a coward for making insinuations that you cannot back up with facts.

In answer to this posting perhaps? Okay, so... maybe it should be considered bad style to use the output of an op amp internally when that op amp is also driving an output jack. We need to add an op amp just to drive the output jack, separate from the one that generates the internal signal. Now loo...

How is that relevant to this thread?

Just taking the opportunity to drop the Lego harpsichord link: http://www.henrylim.org/Harpsichord.html

I intended my post as a starting point so that we could hopefully work out a set of rules to cover as many situations as possible. I think - much as I said in the "voltage range" discussion - that it's a bad idea to try to simplify this down into a set of rules. At most, we can have guidelines of t...

Sorry, I think I missed that you meant the first four circuits to be for module outputs and the last one to be for module inputs . Not the same situation, and some of the points I raised aren't relevant with that distinction. But others of the points I mentioned remain important, and this illustrate...

So can we get a summary of best input and output practices for those times when we want to do it "right"? For example, does the following diagram from Dr. Sketch-n-Etch from the thread below represent the final word for outputs or is there a better way? Those topologies are more or less the ones we...

My guess would be that the Zeropoint is doing an exponential minus an offset, something like f = exp(a*x+b) - c. The descriptions I was able to find with a quick Google search are pretty confusing. I don't think there is a "standard" meaning for this. As you say, exponential functions don't go negat...

There is pcb-pool (pcb-pool.com). I have not used them for PCBs, but the same company's "panel pool" service is good. You will pay a significant premium to have PCBs made in Europe in any case, and the higher base price for "made in Europe" will probably cost you more than the brokerage fees to impo...

Nobody seems to be mentioning the capacitance of patched cables in arbitrarily patched systems as a *good* reason to have *some* resistance outside the OPA output FB. Another reason to consider not putting the output jack (and therefore the output resistor) inside the feedback loop, is that when th...

There are three different questions here: Is it bad that people say the input impedance is 100k, when it isn't always exactly 100k? Is it bad that people build modules such that the input impedance isn't always exactly 100k? Is 100k a bad value for the input impedance even if modules could always be...

I keep seeing so many posts where folks say such things as: A typical Eurorack input has an impedance of 100k. The more I see this, the more I realise this is such a dangerous assertion. Very rarely is this assertion actually correct, and so any discussions based on it are likely wrong from the out...

Do you mean plug and play, sort of? The issue I can see is that variances in VCO outputs mean you won't always get them properly offset/attenuated. It might be interesting - and would make the module less trivial - to have it automatically adjust for varying input voltage ranges. That could be done...

Having no DC path for the bias current from the AD633 "Y" inputs seems like a mistake. Even though the chip is more complicated than just an op amp, those are effectively op amp inputs and they need DC paths just like any other op amp input. It's hard to think that adding rectification will help - u...