This looks and sounds lovely - shame I have no DIY soldering skills to speak of:

https://www.dreadbox-fx.com/antiphon/

  • Analog Monophonic Synth Voice
  • 8 Sine Wave Oscillators
  • Oscillator 7 and 8 can be used as LFOs with a range of 1min up to 10kHz
  • Individual Oscillator Outputs and pitch control
  • 2-Pole non-resonant Low Pass Filter
  • 2-Pole non-resonant High Pass Filter
  • Exponential VCA
  • Analog Spring Reverb with Gain control
  • 150mm 3-Spring Reverb tank by Belton
  • External Input can accept up to Guitar Levels

Desktop Effect:
– Spring Reverb
– Tremolo
– Ring Modulator / Bit Crush
– Filter Effects
– Overdriven Reverb

  • 22 Patch Points
6 Likes

Pretty cool for ~$180. It could be a fun source of FM madness.

2 Likes

Has anyone built a mic pad/attenuator?

I only just discovered that, surprisingly, my RME UCX has no built-in pad, so when I’m doing my snare/feedback stuff, I’m coming in too hot and clipping the input. I’m distorting the shit out of it after anyways, but having an extra gain stage clipping which isn’t consistent on my other sound cards (all of which have selectable 20dB pads) isn’t ideal. So I want to make a physical 20dB pad that I can put in-line when I’m using my RME to make all the setups the same. (actually 15dB would be even better since the RME has a gain jump from 0 to 10dB, so I can’t actually boost it by just 5dB like I can on my MOTUs after engaging the pad)

Something like the Shure A15AS would be great, but it’s pretty expensive and less available on the UK/EU (used that is).

After a bit of google-fu I came across this webpage, which goes into a ton (maybe too much…) detail about attenuators:
http://www.uneeda-audio.com/pads/

It definitely explains everything, but even as someone who isn’t technically illiterate the “you can do it this way or if you want you can do it this way, but also you can do it this way” way of explaining everything confuses the shit out of me.

I’m trying to recreate a fairly standard use case (20dB cut, having no impact on the sound of the microphone, letting phantom power through), which lines up with ‘Example 1’ on the page, but even then, I’m not really sure what configuration/wiring to use etc…

The stats on my gear are quite similar too, though not exactly.
Mic: output impedance of 150Ω
Preamp: input impedance of 2000Ω

So has anyone made a DIY pad?
Did you get all mathy with it?
How did you enclose it?
etc…

1 Like

Ok, I’ve sat down and tried to work out some of the math for this and have gotten somewhere, but not completely.

So with my starting stats of 150Ω/2000Ω I can follow example 1 from that detailed webpage for the post part.

  • Convert 20dB to a ratio. The formula is: k = 10^(db/20). That’s 10 to the power (db/20). The calculator says 10. 20 dB is a ratio of 10:1. If you have a table of K factors, you’ll find this value in the column with K at the top.
  • To achieve this value of loss, the resistors that make up the voltage divider need to have values that fulfill this relationship: k = 1 + (Rseries/Rshunt). But where do you start? There are a jillion values you could start with, but the one that makes sense is the shunt resistor, which effectively sets the value of the output impedance.
    So, Rshunt = Zout = 150 ohms (the microphone imedance).
  • Using algebra, manipulate the formula to solve for Rseries when we know Rshunt and k.
    Rseries = Rhunt * (k - 1). Since k = 10, k - 1 = 9. In the K factor table, there is a column for the value K - 1.
  • Continuing, Rseries = Rshunt * 9 = 1350 ohms.
  • Since this is a U-pad, for a balanced system, we split the series arm into two equal parts. 1350 / 2 = 675 ohms.
  • Resistors only come in particular values, so you need to convert these calculated values into standard ones. For 5% tolerance resistors, there are 24 values per decade (a decade of values is a range of 10:1, the first two digits are the same, and they get multiplied by factors of 10 to get the higher values). For 1% resistors, there are 96 values per decade. We’ll use 5% resistors here. 150-ohms is a standard value, and 680-ohms is the nearest standard value.
  • Summarizing: get 2 680-ohm 5% resistors and 1 150-ohm 5% resistor. Connect as shown in the diagram for a U-pad (balanced-L). If you want to use 1% resistors for improved common mode rejection (CMR) and maybe sonics, the nearest 1% values are 681 ohms and 150 ohms.

I was able to follow that and more-or-less make sense of it. This gives me the values I would need if I’m wiring up a U-configuration pad, and will show up to the mic as an impedance of 1500Ω. Where it gets confusing is that my RME has an actual impedance of 2000Ω. I don’t know how fussy this kind of stuff is, but I guess I’d want the pad to show up as transparently as possible, so I want to try to solve for that.

On that webpage, the next step sort of covers this:

Find the values for the 30 dB pad using the table. The pad input impedance is 4550 ohms (verify this for yourself). What do we need to parallel this with to get 1500 ohms? This additional parallel resistor is R4 in the diagram.

Use a variation of the parallel resistor formula: Rtotal = 1/((1/R1)+(1/R2)). Solving for R1 we get 1/((1/Rtotal)-(1/R2)). The calculator says 2238 ohms. 2200-ohms is the closest 5% value.

I find that really confusing, but using my trusty Max was able to implement the formula and get the same values:
34%20pm

EXCEPT

In my context, my “total value” (Rtotal) is 1512Ω and the value I want at the end is 2000Ω.

I don’t know enough about maths to make proper sense of what’s going on, but it appears that no value of resistor will give a higher value than the input here:
47%20pm

I’m guessing because this is some kind of voltage divider, the impedance produced can’t be higher than what you start with.

SO

Is the difference between a preamp impedance of 1500Ω to 2000Ω a “big deal”, or is there a way to compute the values such that I can get a 20dB loss and still have an input impedance of 2000Ω?

Just started the DannySound EN129 kit. It’s all through hole with great instructions.

This is a through zero oscillator which is is sort of rare. I’ve only seen a few Oscillators that are through zero and they are usually very expensive.

I got most of the main PCB done tonight.

https://www.thonk.co.uk/shop/dannysound-thru-zero-kit/

2 Likes

Just passing on something I found: if you’re building an SMT project, and it calls for 0805 package components, try 0603 instead. They’re honestly not much harder to handle, and I found it far easier to make reliable joins with the smaller components, especially the capacitors.

Edit: it’s also possible that I’ve been doing NLC builds with 0805 when they should use 0603. NLC does recommend 0805 in their build guides.

At any rate, don’t be afraid to go smaller!

I would take @infovore’s advice on soldering over mine, so read ahead!

3 Likes

Conversely, I would strongly recommend not doing this. Why? The pad spacing is different. You might be able to connect an 0603 between 0805 pads, with suitably massive blobs of solder, but there’s no guarantee parts will fit. Component footprints designed for a size are almost always specific to that size, and recommending “using the wrong size parts” isn’t, I don’t think, a good idea.

“Reliable” joins come down to, mainly, solder, a clean iron at the right temperature, and flux. If you’re pouring solder into a joint and it’s just not taking, it’s likely you’re low on flux in the joint, and whilst solder often has flux in it, a dash more rarely hurts. A fluxpen is a good investment: dabbing it on pads you’re about to join, or over the legs of ICs you’re about to solder, and watch that metal flow.

5 Likes

It is very possible I’m generalizing based on my experiences with NLC builds, where which recommend 0805 parts but with caps there ends up being almost no exposed pad to solder to. Resistors are ok, but I often have trouble getting the solder under the capacitor contact and also seeing whether or not I have a join. As you say a flux pen might also help me here.

Since you’ve spoken up and have some direct knowledge: I just built a 16n. Were those pads sized to 0805 or 1206? They were much easier going, even with 0805 parts.

Additionally, my “go with smaller” post was prompted by building this Mutable derivative: https://modularaddict.com/mutated-rings-pcb-1164

That BOM recommends 0603, but the pads are generously sized (and correspondingly tightly packed) to the point that I wondered if they had actually been sized up to 0805.

I’m also gathering reading around that pad size is not entirely standard.

Edit edit: I did some measuring and comparison. NLC does in fact lay out for 0805 components, and just seems to favor less generous landing pad sizes. I think you’re right, a flux pen is in order.

Yeah, this is the key part.

There’s a recommended pad sizing, but remember, most of those are designed for things that will be oven assembled. A pad that’s fine for paste + hot air might be harder to work around with tweezers.

The 16n pads are just using the standard EAGLE 0805 parts, I believe.

In KiCad, you’ll find alternative footprints that have the suffix _handsolder, which are a tad more generous with spacing, and I tend to use those out of habit.

The key part of the ‘spacing’ is the gap between the two pads, not the size of the pad itself, and that’s why I wouldn’t recommend using parts that are too small.

2 Likes

The interesting thing is, NLC is very much a hand-assembled/DIY-friendly operation, and Andrew does a fair amount of workshops and education as well. I’m surprised that the pads would be laid out with a higher degree of difficulty, especially since they’re not so tightly packed that they couldn’t be a bit longer.

Flux pen! Thanks for the pointers.

There is another reason I would like to recommend soldering 0603 on 0805.
I’ve got a very complete assortment of 0603 parts. I’ve soldered many 0805 boards with 0603 parts since that’s the parts I have around and it saves me from having two assortments for all parts (3 actually with through hole). I’ve never ran into trouble using 0603 on 0805. There were a few instances where I had to be creative and build a little bridge between pads that were to far apart.

Has anyone got any audio circuits around chips like cd4060 for making audio oscillators. Inspired by the Olegtron (I have the version one) https://www.olegtron.com/olegtron-4060. Temporarily mislaid my copy of Nicolas Collins - perhaps there is something in there? Google search hasn’t brought up much.
Might have to open up the olegron and see if it’s simple enough / cloneable. I am hoping for a simple basic circuit with some patchability which can be chained or mixed in different ways.

2 Likes

If you’re interested in building synth circuits out of CMOS chips then the best resource on the internet is the Electro-Music Lunetta forum:
http://electro-music.com/forum/index.php?f=160

2 Likes

There is a series of tutorials on Hackaday called Logic Noise. That’s exactly what you’re after I think - highly recommended.

3 Likes

Yes! series starts here: https://hackaday.com/2015/02/04/logic-noise-sweet-sweet-oscillator-sounds/

3 Likes

Can anyone recommend a simple/cheap/compact DIY power amp circuit/kit?

Basically I’m building a setup that will use a couple of transducers which are being fed audio from my feedback snare mic stuff. While testing I’ve been using an old spare bluetooth power amp thing which works ok, but is annoying (it “speaks” when you turn it on, it has a press+hold power mechanism, etc…).

Since it’s driving a couple of transducers with fairly limited frequency response, having good “tone” isn’t really a concern. So mainly looking at simple-ish and compact, and ideally easy to power. Two (independent) channels would be good too I think, or perhaps just a simple mono one, which I can build two of (I’m not sure if I’m going to use one or more transducers).

edit:
ordered a generic ebay power amp thing to see how that goes.

I was going to mention , i use these (Sure amps) all the time for many projects - sound pretty good

3 Likes

Getting ready to revamp my Monoprice 600800 with a Mic-Parts TL3 circuit upgrade and new head basket (replaced the capsule with an RK-47 last year). I’m also interested in a high-end mic build - possibly a U47 or M49. Anyone have experience in that area?

Primarily recording acoustic and electric guitars, male vocals. Looking for a dry, straight forward, smooth sound.

Regarding investigations with the 4060 chip. I found a circuit diagram for the olegtron on their forum


and I got a breadboard version of it working much more simplified than this with a couple of resistors and a couple of caps.

this keeps the voltage starve and the freq pot.
This is a MP3 of around 18 short clips of the types of beats and riffs you can get out of it.

5 Likes

Another circuit I am investigating involves mixing the output of a group of crystal oscillators. This is based on a short description and a bad photograph in the Nicolas Collins book (which I have still temporarily mislaid.) . I am using a selection of different clock rates.

I think the idea was that the imperfect nature of using diodes to mix signals together causes some kind of cascading down of the normally ultra high frequencies of these clocks into the audio range. I haven’t got anything like the amazing sounds he describes and using a few different diodes I haven’t had any results yet. But got quite an intriguing falling drone using 6 LEDs. It seems to take around 15 mins to descend from a mid note to a low rumble and though the effect is subtle there are some odd unstable modulations happening every so often. Has anyone else played around with this type of circuit?

a short recoding of the drone note at two different point of its descent. I think I used the caps in the picture to filter off some high whine. It’s all very temperamental.

10 Likes