Twitter   Patreon WIGGLING 'LITE' IN GUEST MODE

 Goto page 1, 2, 3  Next [all]
Author LEDs as ring mod diodes (HUGE CORRECTION!)
Rex Coil 7
 May 10th, 2018 .... I discovered a MAJOR problem with the transformers I used in the ring modulator I used as a basis for authoring this thread. Please understand that some of the observations I made in this thread were erroneous. Please read the ~correction~ post that is dated with the same date as this note, which can be found on PAGE 2. Many things will make more sense once you understand what went wrong. Thank you. ********************************************************** Yet another silly question out of the Rex Coil 7 files. I'm using some LEDs in passive transformer style ring modulators for the diodes. I've already assembled one ring mod with them and I like what I'm hearing. Do they need to be matched, and if so, how does one go about doing that? Thank you!
guest
 not certain if it will make much difference to match them, but if you want to give it a try, you need to send a fixed current through them, and measure the voltage. you can do this with an inverting opamp. put the diode in the feedback in place of the resistor that normally goes there. then measure the output voltage of the opamp. the input should go to 5V or so, and the input resistor should be scaled to be in the range of currents your circuit is producing, probably in the 100uA range (so 50k).
Rex Coil 7
 guest wrote: not certain if it will make much difference to match them, but if you want to give it a try, you need to send a fixed current through them, and measure the voltage. you can do this with an inverting opamp. put the diode in the feedback in place of the resistor that normally goes there. then measure the output voltage of the opamp. the input should go to 5V or so, and the input resistor should be scaled to be in the range of currents your circuit is producing, probably in the 100uA range (so 50k).
I figured you'd have to measure voltage or current running through them, then observe their similarities/differences to find closely matching ones. However I really didn't know ~how~ to go about it.

That having been said, thanks for the explanation!

Like you, I too feel that for this purpose (passive classic ring mod) it really won't make a blinkin' eye's worth of difference if the LEDs aren't precisely matched, really. Now, using germanium diodes, that's a different story altogether.

I still like the LEDs much better, though. For the reasons that I use a ring mod (adding rude upper harmonics, much the same as hot shot guitar players use "pinch harmonics" for expressive purposes), the LEDs are better suited.

I've not heard of anyone ever using LEDs in a passive ring mod before! I am really glad I'm keen on taking the odd path, rather than the well traveled road! Many times that dirt road takes you places far more interesting than the 4 lane highway does.

Thanks again for the tip! I've added it to my little folder of Experienced Ideas.

basicbasic
unrecordings
 I'm interested in this, have you got a circuit diagram to share ? What transformers are you using ? I built one on perf board based on Ken Stone's layout, I'm interested in what LEDs might do
PWM
 unrecordings wrote: I'm interested in this, have you got a circuit diagram to share ? What transformers are you using ? I built one on perf board based on Ken Stone's layout, I'm interested in what LEDs might do

https://www.muffwiggler.com/forum/viewtopic.php?t=175136&start=450
elektrouwe
 I did a lot of LED matching for LED-ladder filters. Easy going with my YATMA: https://www.muffwiggler.com/forum/viewtopic.php?p=2114868 but why would you use LEDs in a ringmod ? High forward voltage is not your friend...
Rex Coil 7
 unrecordings wrote: I'm interested in this, have you got a circuit diagram to share ? What transformers are you using ? I built one on perf board based on Ken Stone's layout, I'm interested in what LEDs might do

This is the CGS Ring Mod (Ken Stone) PCB (I bought all seven of mine from "Modular Addicts" ... \$6.55ea or something). I added sockets that I salvaged from socket strips ... took a bit of doing but it got done! I used the sockets to provide the capability of trying different diodes within the ring modulation circuit.

Used standard 42TM018 transformers (10k primary x 10k secondary transformers, with roughly 6.5khz frequency response) ... Mouser item ... something like \$3.55 each. I bought 14 of them.

PILE OF PICS: (just remember, you asked for this! ... haahaa!)

Then, on to testing different diodes. I used a work-horse panel I have and mounted the Ring Mod on the outside of the panel to facilitate easy changing of diodes on the ring mod PCB ......

And proceeded to testing and comparing different diodes, all very carefully matched (except for the LEDs). I used vastly different input levels to the modulator and the carrier, as well as vastly different output levels. Taking notes on the differences between sound and behavior of the different diodes used.

Oddly enough, I ended up liking the LEDs the best (for MY purposes, if anyone is considering using LEDs as diodes in their passive ring modulator, please be sure to test them out somehow before committing to their use. They are FAR louder on the output side when compared to germaniums or even silicons, LEDs are much more "rude", and distort with a very sharp crunch.) You've been warned!

I decided that I have a use for germanium based ring modulators, and LED based ring modulators. That having been said, I've decided to use both in my project synth. I have 2 voices, comprised of 2 VCOs each. I have ring mods in each voice (meaning I have two ring mod circuits in my synth, one per each 2-VCO voice). I like having both germ-type and LED-type ring mods available for both voices, so I now have FOUR ring mods in my synth, with them all mounted in a center strip that divides the upper and lower rows of modules in my synth cab.

Here's a very early picture of my synth, taken a few years ago. You can see what I mean when I say "Center Strip" there, it divides the synth vertically, and offers roughly 1.5" of usable space between the cabs.

Here's the results of that installation design. Note that the socketed ring mod PCB is not used in my synth, it's been replaced with ring mod boards with permanently installed red rectangular LEDS ... so in the Ring Mod stack ... of which there are 2 .. one stack per voice ... consists of a germanium stoked ring mod, and an LED stoked ring mod, with an "A/B" toggle to swap between them. there's also "flip-flop" toggles in each ring mod pair that allow swapping which VCO is the carrier and which VCO is the modulator. There's also another pair of toggles per-ring mod pair that allow swapping between different input sources for the selected ring mod. I won't go any deeper than that here .... I've documented the piss out of how I'm doing all of that in my ongoing project thread ... I'll post a link to that thread at the bottom of this posting:

Here's the "Center Strip" controls for the ring mod stack on the left voice. The 4 small screws surrounding the little ~blue~ toggle switch are the mounting screws for the ring mod stack.

And here's the "Center Strip" controls for the ring mod stack on the right voice.

And for anyone that doesn't approve of ring modulators .....

LINK TO PROJECT THREAD (this takes you to page 7 where all of my most recent building activity takes place) .... https://www.muffwiggler.com/forum/viewtopic.php?t=78836&start=150
SoundPool
 oh good- more ring mod obsessiveness, makes me happy. not to get too far from the diodes but I'm curious if anyone has put as much into trying different transformers as diodes in the classic passive set up. I've got quite a few different transformer sets here, but I haven't really set anything up yet for doing methodical testing. I imagine the transformers might make a bigger difference in an active set up where you are probably closer to hitting the saturation points of the transformers, but obviously coloration will occur in a passive one too. I haven't coughed up the dough to check it out but would be curious if there is a discernible difference between higher end/proper audio transformers like Edcors or Carnhills.
commodorejohn
 Okay, now I'm curious what this looks like in action...
Rex Coil 7
 SoundPool wrote: oh good- more ring mod obsessiveness, makes me happy. not to get too far from the diodes but I'm curious if anyone has put as much into trying different transformers as diodes in the classic passive set up. I've got quite a few different transformer sets here, but I haven't really set anything up yet for doing methodical testing. I imagine the transformers might make a bigger difference in an active set up where you are probably closer to hitting the saturation points of the transformers, but obviously coloration will occur in a passive one too. I haven't coughed up the dough to check it out but would be curious if there is a discernible difference between higher end/proper audio transformers like Edcors or Carnhills.

Edcor makes some really nice ones, roughly \$30 bucks each. 10k x 10k - 20hz to 20khz freq response. XSM series.

Here's a Ken Stone ring mod set up with Edcor transformers. Be prepared to do some DIY to get the transformers mounted.

Is the increase in freq response worth the extra cost and hassle? That's up to you. A pair of Edcors will run you about \$60 plus \$15 in shipping. So is \$75 bucks worth of transformers worth it?

Dunno.

Search You Tube for a video done by CZ Rider, he posted a short video using his old Moog modular with an Edcor based Ken Stone ring mod. It sounds phenomenal!!!

But, I can't say it sounds any better than a ring mod built with 6khz freq response. There's a shitload of conjecture and speculation about it, but no hard core evidence that supports the claims. Keep in mind that there are many factors that may be contributing to what you may hear in CZR's video.
Rex Coil 7
 Per my post above .....
cornutt
 Not knowing very much about the "true" ring mod circuit... Do you have to used different transformers, or different taps, to make up for the higher diode drop of an LED?
Rex Coil 7
 cornutt wrote: Not knowing very much about the "true" ring mod circuit... Do you have to used different transformers, or different taps, to make up for the higher diode drop of an LED?
I didn't do a thing, I just stuck them in there and went for it. There's far FAR less signal level loss with the LEDs, and as far as using different transformers .. nah .. it's a totally passive circuit .. there's no gain stages in it or anything like that to worry about.

The transformers used in passive ring mods don't have different taps, they only have the ends of the windings and the center tap. These transformers are simply what are called "matching transformers", and are typically used for impedance matching. You simply need to make sure that you use 1:1 ratio transformers .... same input impedance as the output impedance. In the Ken Stone ring mod, it calls for 10kohms input impedance and 10kohms output impedance (same on the in and out ... therefor 1:1 ratio). Myself, I personally do not believe that using "20 to 20" (20hz to 20khz) frequency response spec is necessary. But other people may tell you a different story.

That said, the transformers that the Ken Stone ring mod calls out work just dandy! Ans they're easy to work with since they fit right on the board all nice and neat like.

Using LEDs in the circuit simply allows more signal to pass through without the circuit sucking the level down so hard. And LEDs seem to have a more harsh sound (which is a GOOD thing, because the germaniums can really tend to suck the top end out of the sound, and LOTS of low end goes away as well).

There's absolutely nothing you can do to hurt anything if you experiment with different diodes in a passive ring mod. Absolutely nothing at all. In fact, you'll find that the amount of signal volume is really cut down a lot after running it through a passive ring mod, so much so that I use a Dot Com Q118 Instrument Interface with the gain switch set on either 100x or even 1000x to bring the sound level back up to the same as VCOs that aren't going through the ring mod.

When you use something like a MoogerFooger ring mod, there are gain stages both before and after the ring mod circuit to compensate for the volume loss. This goes for just about any active ring mod device ... there's gain stages used on the ins and outs. The components of the passive ring mod reduce the signal level, it's just what happens in a lot of passive circuits.

So ... ~no~ ... you don't have to use different transformers or anything to compensate for using LEDs, in fact, LEDs suck far less of the signal level down than using silicon or germanium diodes. Germs are the most quiet out of all of them, silicons are next, and the loudest are the LEDs, which also suck the least amount of bottom end out of the signal as well.

SoundPool
 I'm curious to try out the Edcors, but here in Europe they are pretty expensive for what is normally a more budget option transformer. I think I can get Carnhills for the same price or even cheaper. So far I have just tried out some smaller trafos- the ones you have in the photo above, some other cheap aliexpress finds, and then a few NOS transformers I found here that are unmarked or unidentifiable but have center taps, and some crystal radio output transformers. I should set up a rig for some proper testing. I figure worst case if the differences are negligible I can build some passive transformer coloring boxes out of everything for experimenting with.
Rex Coil 7
 elektrouwe wrote: I did a lot of LED matching for LED-ladder filters. Easy going with my YATMA: https://www.muffwiggler.com/forum/viewtopic.php?p=2114868 but why would you use LEDs in a ringmod ? High forward voltage is not your friend...
It's a passive circuit, the only thing that the higher forward voltage will do is reduce the output less than diodes with lower forward voltage.

It still "ring modulates", at least enough to my own liking.

 elektrouwe wrote: ... but why would you use LEDs in a ringmod ? High forward voltage is not your friend...

I think a better question is ... Why not?

Repeating myself, it's just a passive circuit ... it's not as if higher forward voltage of the LEDs is going to increase the output voltage of the ring modulator above whatever you piped into it in the first place, y'know? No matter how much forward voltage the LEDs have, the output voltage is STILL going to have less voltage than the input voltage level is. No harm done.

If I'm missing something, please tell me. My best days are days when I've learned something new.

emmaker
 Yes I'm wondering about the forward voltage also. Isn't the point of the diodes to act as 'switches' in the circuit? So you want the lower forward voltages of germanium and schottky diodes for less distortion. With LEDs and silicon diodes won't you have more distortion? So is it that there is more distortion with the LEDs which is more 'appealing' to some people?
Rex Coil 7
 emmaker wrote: Yes I'm wondering about the forward voltage also. Isn't the point of the diodes to act as 'switches' in the circuit? So you want the lower forward voltages of germanium and schottky diodes for less distortion. With LEDs and silicon diodes won't you have more distortion? So is it that there is more distortion with the LEDs which is more 'appealing' to some people?
How would lower forward voltage reduce distortion unless the sound you've pumped in to the ring mod was already distorted? What, exactly, is causing the distortion you're speaking of? I'm not being a dick, I'm asking a pragmatic question because I honestly do not understand.

Besides, it's been my experience that diodes with lower forward voltage actually distort more than ones with higher forward voltage. They clip far earlier as the voltage of a given signal is increased (that is to say, the louder the input signal is the greater the tendency to distort with lower forward voltage .... I've built hundreds of overdrive units that use that very concept to determine how loud the output can be before the diodes clip .... germaniums clip with far less input gain than LEDs do ... I have an entire product line that I used to make that took full advantage of that fact). It takes a lot more input gain for higher forward voltage diodes to clip than it does for (let's say) germaniums or silicons.

Perhaps my understanding of things is incorrect. I'm good with admitting that. But from years (over a decade) of using diodes as clipping components, my experience has proven to me that LEDs don't clip until they are hit with A LOT more input gain than germaniums and silicons.

Now, if we're talking about ~tone~ ... that's different. Germaniums produce a lot more of a mid-range-heavy sound and even act like high pass filters to a degree since some of the low end is reduced, where LEDs produce a much more open and uncolored sound, pretty much what goes through them is what comes out. But when you drive Germs to clipping, they get all fat and midrange heavy, where LEDs become more rude and ~sharp~, with a LOT more output gain since the LED allowed far more voltage to pass though them compared to germaniums or silicons.

Look, I have no idea what theory behind using this or that is. At the end of the day, all I really know is that using LEDs made the sound go ~woop woop woop~.

And I'm good with what I hear. Technical sins be damned.

All of these ~doubts~ are precisely why I made a ring modulator with socketed diode holes, to be able to test out different ones to see what works for me, and what doesn't. There's really nothing to "worry" about, it's not like if a different diode is used the synth will catch fire or some sort of damage will result.

Sometimes, ya just gotta put the book down, and GO FOR IT! A passive ring mod is a safe thing to tinker with in that regard. The absolute worst thing that can happen is you won't like what you hear.
emmaker
 Nothing wrong with trying to understand something in a polite way. My understanding of the circuit might be completely wrong here. But here is how I understand it. Thinking about this now I might have some more ideas. Those I'll get to at the end. The diode bridge acts as a switch routing the positive and negative signals to the second transformer. So the forward bias voltage acts as a switching point. With germanium and schottky diodes that switching point is about 0.3V, silicon 0.6V and LEDs are going to be 1.5V-3V. So if the LEDs switch at let's say 2V is there going to be a big gap in the switching area versus the germanium diodes? Here's the thing that I just thought of. Is the modulator run in the linear region of the diodes and the forward voltage is really the clipping voltage of the circuit? Going to have to do some research on this. Going to have to learn audio transformers in LTspice. That would answer my question. Also found this video on YouTube. While it for RF it shows the harmonic relationship between the two signals. Also it's a good demo on how to use a digital scope to see the harmonics. Ring modulator and digital scope.
Rex Coil 7
 emmaker wrote: Nothing wrong with trying to understand something in a polite way. My understanding of the circuit might be completely wrong here. But here is how I understand it. Thinking about this now I might have some more ideas. Those I'll get to at the end. The diode bridge acts as a switch routing the positive and negative signals to the second transformer. So the forward bias voltage acts as a switching point. With germanium and schottky diodes that switching point is about 0.3V, silicon 0.6V and LEDs are going to be 1.5V-3V. So if the LEDs switch at let's say 2V is there going to be a big gap in the switching area versus the germanium diodes? Here's the thing that I just thought of. Is the modulator run in the linear region of the diodes and the forward voltage is really the clipping voltage of the circuit? Going to have to do some research on this. Going to have to learn audio transformers in LTspice. That would answer my question. Also found this video on YouTube. While it for RF it shows the harmonic relationship between the two signals. Also it's a good demo on how to use a digital scope to see the harmonics. Ring modulator and digital scope.
Some of your statements are far over my head. I have no formal electronics education, everything I know has come from decades of falling on my face and getting back up to try again. Sometimes my ignorance shows, as in this case, I don't have any problem with that, where the troubles begin is when my ignorance gets in my way!

As it has now.

I cannot answer your questions, I can't even point you to solutions. All I can tell you is that I am quite pleased with the results, and I am so busy with getting my synth built that I have not taken the time to learn ~why~ the modifications to the ring modulator have created pleasing results for me.

Three months ago, my synth was in working order. Then I decided I wanted to get a normalizing project within it that I had put off for several years done.

That said, now my synth is literally stripped down to the PCBs and an empty cab. As I find solutions to various modifications and improvements, I simply move on to the next task on the (rather long) list of things I need to do so as to get this synth back up and operating.

It's actually pretty damned comical! I started off just wanting to normalize the 1v/oct distribution to all 4 of my VCOs, and then another idea sprang forth ... I mean, may as well, "now that it's taken apart, I might as well do ~this~ .... ~that~ .... ~this one other thing~ .... ~oh yea, and that thing too~ .... ~wow!, it would be so easy to add this~ .....

ugh .....

HAAHAA!!!!!
emmaker
 I don't care how much you study electronics from the books. Sometimes you just need to build something and put a scope/meter on it to see how it works and the nuances of the circuit. Being self taught from building stuff will get you there too. That's how I started at 10. I could read schematics and solder but didn't really understand all the theory and math. Just might take a bit longer. One recommendation I'd make is take good notes and keep them around. 5 years from now you maybe wondering how you did something and it's a lot simpler to look at your notes verses going through the learning process again. I work as a HW/SW engineer in test/measurement and computer systems with more emphasis on SW. Digital and lf analog stuff I can handle. Getting into hf analog takes a very different way of thinking. I did have an rf project I worked on a few years ago. Funny thing was that all the theory about radio I learned 50 years ago was still valid. Difference was it went from tubes, coils and transformers to chips. Theory was still the same.
Rex Coil 7
unrecordings
 That makes for some good reading (glad I asked) For my perfboard version I had a problem sourcing transformers that were less than a tenner each. I found some Triad ones that if memory serves were £5 I'm wondering now what LEDs in parallel with other diodes sound like...
Rex Coil 7
 unrecordings wrote: That makes for some good reading (glad I asked) For my perfboard version I had a problem sourcing transformers that were less than a tenner each. I found some Triad ones that if memory serves were £5 I'm wondering now what LEDs in parallel with other diodes sound like...
I've experimented with diodes as clippers "ad nauseam" when I was building guitar pedals for a living. I've found that with diodes in parallel the signal seeks out the path of least resistance (so to speak, I know that is not technically correct, but it is the end result). So if you put (let's say) a germanium in parallel with an LED, since the LED is seen by the signal as more "open", the end result is that the sound that comes out of the speakin' hole doesn't sound any different than if you'd just used LEDs alone.

Now, I'll repeat myself here, I know that there are technical explanations and "cuz reasons" stuff that folks with Electronics Engineering degrees and trigonometry education will scoff at in response to my above statement. So I feel the need to explain that what I am speaking of is ....

.... at the end of the day, whatever technical reasons that are behind the actions .... I'm talking about how certain actions change or affect what comes out of the speakers.

I've done so bloody much empirical testing ("let's see what happens when I do ~this~") with diodes in the signal chain, that I've amassed a mental catalogue of those change's end effect.

For instance, here's an expansion module that I used to make for use with some of my overdrives .... it has two rows of diodes, each row applied to the separate halves of the signal's polarity (positive and negative) .....

That particular one is set up so that it "inserts" between the audio signal and ground (chassis). But I've done the same thing with circuits where that box of diodes is inserted into the signal chain in the feedback loop of the gain stage opamp. In that case, a TRS jack/plug is used instead of a TS jack/plug. I simply wire a TS or TRS jack into the overdrive's signal chain, and then use a simple patch cable (again, either a TRS ot TS depending on the overdrive ciruit's nature) to connect the box to the overdrive box, and then those diodes are then inserted into the signal chain. Unplug the box, and the signal falls back to a set of default diodes wired across the switching jack's terminals.

As you add more diodes into the chain, their net effect becomes less and less noticeable. But with my little box, the player could apply what ever diodes to each pole of the AC audio signal to suit their style of playing. When a guitar string is plucked, depending on how the player uses a pick (or his fingers if he's a finger player), the string will orbit above the pickup in various shapes of orbits, and those orbits are often times not exactly centered over the pickup's poles. That said, depending on how the string was plucked, and how the orbit evolves as it vibrates, there may be more signal voltage on the positive half of the wave, or perhaps the negative side of the wave. There's a lot of "pickin' style" that comes into the equation.

That said, the little box allows the player to fine tune how the signal is clipped, on both the positive and negative halves of the signal.

Ok, so I've made my point .... "I've played around with diodes a lot". And while my experience has to do with clipping, from the little bit of testing I've done so far with the passive ring modulator circuit, it all seems to be relevant and applicable knowledge.

As I've said, I am no kinda expert on anything, and I don't claim to be one. But I have done a lot of goofballing around with diodes, that much I can say

So there's the short answer (short for me .. haahaa!) to your inquiry about placing diodes in parallel to have some net effect on the ring mod.

I could be DEAD WRONG as well, let's keep that in mind.

unrecordings
 I'm sure I read somewhere that for diodes in parallel, put a very low value resistor in series with each diode to balance them or something - or maybe a pot in series with the LED... The annoying thing for me is, right now I don't have the parts, so I can't sit and experiment
 Goto page 1, 2, 3  Next [all] Page 1 of 3