#1

I’m going to lay bare my general naiveté when it comes to DIY electronics…

In preparation for receiving a norns this week, I wish to set up a system for sending MIDI from norns to my Elektron Analog Keys, which can output that data as control voltage to my Buchla (which does not have a MIDI/CV decoder module). The AKeys can be set to provide Buchla-friendly 1.2v/oct and 10v gates, but the Buchla should also NEVER receive negative voltages on any of its inputs and, apparently, the AKeys is capable of bipolar CV output. This is where my question comes in:

I have built a breakout box which takes the TRS signals from the AKeys and splits those into isolated pairs of banana jacks, along with the required common ground jack (see photo below). Following advice, I have used schottky diodes (1N5817) to wire the plugs which should remove all negative voltages.

Before I hook things up and make a costly mistake, I would like to test for pass-thorough of any negative voltages. I have a cheap-o multimeter on hand (photo), which does not seem to give a negative voltage reading from either the breakout box, or testing a TRS cable attached to the AKeys CV outs and sending low octave notes. Could someone recommend the best method for safely testing this breakout box once and for all?

Thanks!

box
box.jpg1080×1080 689 KB
multimeter
multimeter.jpg1080×1080 850 KB

#2

Nice little box.

To make sure your multimeter measures negative voltages you can measure the voltage of a battery and then swap leads (black to + and red to -). Of course you can do that with the TRS cable from your AKeys. To be safe you have to be sure to send a negative voltage before the diode (anode) and measure zero after the diode (cathode).

I don’t know If you thought of this but there is a 450mV voltage drop over the diode (which is about 4.5 semitones)… but I guess you can just tune the Buchla accordingly.

#3

i assume you mean buchla 200e since old guys are basically tolerant of -V.

but, if youre gonna hit ADCs directly then yeah, schottky diode is a good idea.

metering is about all you can do, but do be sure you test your diode box with -V input.

AFAICT the analog keys should indeed be able to go negative, possibly by changing this offset value and/or sending very very low notes:

if nothing else you can indeed just reverse a 9v battery as suggested above.

be aware that you may see a small amount of voltage drop and/or nonlinearity from the diodes.

1 Like
#4

@sonoCircuit Thanks! I have the aesthetics down, just not the electronic know-how :wink:

@zebra Yes indeed, 200e modules, but I will mainly be connecting first to a BEMI 208 which apparently should have some protection from negative voltages, and then passing those voltages through to 200e modules.

Based on the photos of the breakout box writing, and the mode setting of the multimeter, does it at least appear that I have done those two things correctly (orientation of diode, etc)? I will then proceed with testing the reversed battery and more adjustments on the AKeys to send a negative range CV.

#5

The diode orientation is correct as well as your multimeter settings.
I can’t really see where the lower ground wire is going to though.

#6

@sonoCircuit You caught it too…one of the two ground wires had broken away. I just got out the solder and repaired it!

I’ve also tried the reverse battery and the meter does indeed display a negative voltage, so the next step is trying a cable connected to the AKeys.

Thanks again

1 Like
#8

I was unable to get a negative pitch voltage sent from the AKeys, but I did try the reverse battery technique attached to a cable connected then to the breakout box, and was able to pass a negative voltage through to the banana jacks. Of course, just realized that there is no diode on the ground. Am I correct that, if a negative voltage is sent from the AKeys (or any other instrument) that the ground banana will also require a diode for protection? I’m not exactly clear on how a negative CV voltage functions in practice.

#9

Hmm… that’s odd. A measured voltage is just referenced to ground so make sure when measuring the black lead is connected to ground and you use the red one to measure.

The way you hooked it up seems good to me. There is no need for a diode protection for the ground banana jack. The schematic below is basically what you did, which is fine AFAIK.

Schematic
Schematic.png1160×952 4.82 KB

Try to build a bipolar power supply with batteries (see pic below) if you don’t have a bipolar power supply. Use the centre tap as ground and now you have a positive and negative voltage to test your circuit.

Hope you manage to work it out! Cheers

540x405

1 Like
#10

Thanks very much for the assistance! I’ll try out the 2x 9V bipolar setup next.

#11

Here are the results of my MacGyver-style test using the bipolar 9V battery setup. I have the ground (sleeve) of one end of the cable attached to the “ground” of the pair of 9V’s, and the tip of the cable alternates between the + and - of the 9V’s. Unless I have made an error in setup, it does appear that negative voltages passed to the bananas. aargh.

neg
neg.jpg1500×1125 1.49 MB
pos
pos.jpg1500×1125 1.47 MB

#12

I‘m a bit baffled… do you have any spare diodes to test separately?

EDIT: Further question: are you using a stereo jack or a switched mono jack? Just to be sure.

#13

Ah…you are correct. I have TRS stereo jacks correctly installed inside the breakout box, but was mistakenly using a TS mono jack to test the other end of the cable. I do not have another spare TRS jack so will try connecting the wires directly to the cable plug and re-test.

EDIT: Just re-tested connecting the bipolar battery directly to the TRS cable and same results - negative voltage is being passed to bananas. To confirm, this is the diode I am using:

https://www.digikey.ca/product-detail/en/diodes-incorporated/1N5817-T/1N5817DICT-ND/190530

#14

part looks fine. shouldn’t need anything else for half-wave rectifier (which is what you’re building.)

try this basic test with just 9v battery, diode, meter.

9v positive terminal -> black probe
9v negative terminal -> anode
cathode-> red probe

expected result: 0v

switch the battery around:

9v negative terminal -> black probe
9v positive terminal -> anode
cathode-> red probe

expected result: +9v

go from there.

#15

@zebra @sonoCircuit Thanks for continuing to assist me with this mad issue. Following the advice above, here are the results:

neg
neg.jpg1080×1080 996 KB
pos
pos.jpg1080×1080 1 MB

I also happened to have the same model schottky from a different supplier and, while the negative value was less (~3.80) it was still passing negative voltage. Then tried inverting the two different schottky, but same results again.

Insane, no?

#16

pretty insane. my best attempt at explanation is that these diodes have been blown up by exceeding their reverse voltage spec (20v @1A for 1N5817.) or i dunno, baked in an oven or something. (i guess even shorting with a lower voltage and no resistive load could do it? not sure. [ed: yes it could; this part will blow up with 25A forward current. your voltage source probably won’t like it either.])

an even more basic test of a diode is to use the 200ohm resistance setting of your multimeter; you should see low resistance from anode->cathode and high resistance the other way. but if they read similar then they are toast.

really not much else to say, pretty simple things.

3 Likes
#17

Testing all of the extra diodes I have on hand (plus the ones installed in the box already) I get the same results using the resistance test. With red probe on side of TRS jack and black probe on side of banana jack the reading is ~92.5. With the probes swapped the reading simply shows “1 .”

Should these be the results I am hoping for, and in the correct order?

Res%201
Res 1.jpg1080×1080 1.13 MB
Res%202
Res 2.jpg1080×1080 1.1 MB

#18

if the first photo is 92 ohms in the reverse direction, thats waaay too low. it should be basically infinite

i dunno if the second photo is saying 1 ohm or “off scale”

i’m not sure what’s going on with your stuff. the wiring looks fine but i’m feeling like your diodes are just bad.

to make sure i’m not completely demented and havent, like, forgotten how a battery or DMM works, i pulled out a random small-signal diode (1n4148, not schottky but v similar to your part) and a cheapo multimeter. resistance meter shows 8MOhm in fwd direction (higher thanmd anticipated, and a schottky might be lower) and off-scale in reverse direction (as expected.)

with 9v battery i read 9v in fwd direction and -0.2V in reverse. as expected. i believe this is with a 220kohm resistive load in the DMM.

so i dunno what else to tell ya, sorry!

#19

I can only back up @zebra, sorry it’s not working out for you. Did you buy the diodes especially for this purpose and may I ask where? You could drop them a line and maybe get some working ones.

#20

@sonoCircuit The diodes are in two batches - one set from Digikey, and the other from a no-name-brand eBay seller, and I have tested them all with the same results.

Given what is at risk (ie potential of frying Buchla modules), and my diy ineptness, I think it’s time to send this box to someone with significantly more skills than myself. I very much appreciate the time that both you and @zebra have put into helping me on this project.

1 Like
#21

I was going to order some replacement diodes. Would either @sonoCircuit or @zebra have a suggestion of a more suitable part # and supplier? I’m well out of my depth here. Thank you!