Reading a 1Mhz -15V to Gnd clock

Status
Not open for further replies.
Please tell us what instrument this is for.

You've said organ/keyboard, Yamaha, 14 note polyphony, analog with digital control and it's pre-MIDI so old. So curious it's killing me.

Yamaha’s proprietry keycode system, used in everything from CS-01 to Electones. They use a weird key matrix, on a 61-note keyboard it’s 5x12, which Highly Liquid’s UMR2 doesn’t support. I figured if you can bypass the keyboard assigner, you can extend notes beyond what key’s are available. The vibrato inplementation is also based on modulating the master clock (limited range), which the sync clock is based on, so looking to get the teensy to act as both a slave and master for the master clock, if that’s possible could do some pretty neat stuff with pitch modulation.
 
I’d have to alter the resistor choice, 1k is probably the max impedence working with a clock at 1mhz. R2 and R3 add’s a dc bias to the ac coupled signal, so they can likely remain at the current values.

I’ll get this running in my simulator, i’m really enjoying figuring out the theory first in simulation before breadboarding or sodlering ;)
 
i’m really enjoying figuring out the theory first in simulation before breadboarding or sodlering ;)

Ha, that's pretty much the exact opposite of how I work. Really not fond of simulators. They're great for analyzing, but hard to design anything new that isn't already designed by someone else.

I can usually come up with stuff like msg #31 on a whim. But I didn't anticipate the need for less than 10K to get good performance for 1 MHz.... until I looked at the waveforms on my scope.

A well equipped workbench also really helps! ;)
 
Ha, that's pretty much the exact opposite of how I work. Really not fond of simulators. They're great for analyzing, but hard to design anything new that isn't already designed by someone else.

I can usually come up with stuff like msg #31 on a whim. But I didn't anticipate the need for less than 10K to get good performance for 1 MHz.... until I looked at the waveforms on my scope.

A well equipped workbench also really helps! ;)

Indeed it does. Unfortunatley I had to relocate to Australia this year and all my schtuff is in storage in Seattle, including my awesome soldering station. Everything like that is double the price here and i’m
not flush atm.

Ive found it helps the simulator, esspecially being able to slow the simulation time below 1us per second. All the stuff you were saying about those resistor choices i could see the result on the fly, the ramp/fall times, how close the LOW signal got to ground, picking those sweetspots would be a pain on a breadboard, and i hate brekibg up my resistors ;)

Appreciate the help, i’ve just started rebuilding my expansion board. Just got done with mirroring the onboard hex buffer stage. These yamaha’s use custom chips that would be painful to replace, i’m glad i took a step back to implement this in a foolproof way. I’m not even game using switching regulators, so i think i’ll protect the teensy and yamaha cmos by using seperate linear regulators.
 
Ha, that's pretty much the exact opposite of how I work. Really not fond of simulators. They're great for analyzing, but hard to design anything new that isn't already designed by someone else.

I can usually come up with stuff like msg #31 on a whim. But I didn't anticipate the need for less than 10K to get good performance for 1 MHz.... until I looked at the waveforms on my scope.

A well equipped workbench also really helps! ;)

I’ve gotten a little further but im having some issues. I’m using the T3.6 3.3v pin as the 3.3v source at the collector (source). The base (gate) is connected to the internal keyb. ground, the only way i can get the circuit to function is tying the gnd pin on the teensy to the keyb. 0v/gnd.

I thought everything was going well, i mirrored a t3.6 input pin to go out an output pin, but (to my horror), when i measured the input pin it was going negative by about -800mv. I have a 3.3v zener on each of the input pins reverse biased to gnd, without those i think i would have damaged the pin. For now i don’t want to touch anything until i understand what is going on. I’m think it may have to do with tying both gnd’s together, but without doing that nothing works.
 
Ah, too much faith i had in the simulation. I spent a couple hours playing around with resistor combinations and i found the sweet spot. I did need to add a 22p cap + 3.3v zener at the teensy pin to get it to behave, but now i’m getting almost 0v to 3v logic. I ran a test running the 3v logic signal at an input pin mirrored to an output pin, using digitalRead/Writefast, and the timing is spot on with the signal coming from the keyboard. Thats impressive.

This is good news. It means i can filter everything from inside the keyboard through the teensy, and then feed it back to the sound chips This should allow for some fun stuff, like octave shifting, adding 3rds/5ths/7ths or single note chords, perhaps even an arpeggio, all without using midi playing from the builtin keyboard.

Check out some scope screenshots;

https://ibb.co/cQysmV9
https://ibb.co/yYMjNhh

*this is before i tweaked the values, so low is around 500mv.

I have a question, the 3.3v output on the teensy say max 250ma. If you drew 250ma, would that leave nothing for the pin outputs on the teensy?

Also one other question, can you configure the slew rate on the outputs?
 
Where did you see the 250ma current spec? On page 6 I found 25ma.(https://www.nxp.com/docs/en/data-sheet/K66P144M180SF5V2.pdf)
Not familiar with the Teensy regarding slew rate control.

The simulator is model dependent, so if the part model included all the parasitic elements then it would behave more like the real world circuit. Thus all wires would also need to be simulated with their parasitic components. This make the simulation very complex and probably very slow due to all the calculations of each element's contribution to the output results.

What does your final interface schematic look like?
 
Where did you see the 250ma current spec? On page 6 I found 25ma.(https://www.nxp.com/docs/en/data-sheet/K66P144M180SF5V2.pdf)
Not familiar with the Teensy regarding slew rate control.

The simulator is model dependent, so if the part model included all the parasitic elements then it would behave more like the real world circuit. Thus all wires would also need to be simulated with their parasitic components. This make the simulation very complex and probably very slow due to all the calculations of each element's contribution to the output results.

What does your final interface schematic look like?

Data Pins are limited to 25ma, but there are two 3.3v pin that can power external devices. I think i’ll need a seperate power source eventually, but its useful in the short term not needing to use another psu for the 3.3v. I’ll likely use a 5v switching supply that drives a 3.3v linear regulator long term.

https://forum.pjrc.com/threads/38027-Teensy-3-6-Power-supply

I haven’t figured out 3.3v to -15v logic yet, i gotta basically gotta do a two-way translator yamaha->teensy->yamaha.
 
Status
Not open for further replies.
Back
Top