Teletype hardware: open

[Gerald_Stevens]

[toneburst]

@okyeron

my trick with these oleds was to use a low profile header on the main board, then use a standard (square pin) male header put through from the top. Then clip the top bits off the top after soldering.

I tried this with one of my two builds. The only potential issue here is that if you don’t trim the pins low enough on the top, there’s a possibility of them shorting against the bottom of the panel.

Another option is to install the pin header the right way up, them (carefully) lever the plastic spacer strip off somehow.

I think I did that with my 2nd Teletype build, but it’s a slightly scary process, as I worried I’d damage a trace if the flathead screwdriver I was using slipped.

A third option is to screw the display board down with the spacers at the other side, and rest it on the header socket, then remove the pins from a pin-header strip, and simply push them through the header holes in the display board into the socket below, then solder them one-by-one.

[jmsiener]

I ended up using some chonky diode legs (Tayda 1n4001 to be specific) for my display like someone mentioned earlier in the thread. I used the socket listed in the bom but the diode legs gave a similar diameter to the machined header. I got it pretty low profile, flush with the socket, but I feel like it’s janky with just diode legs although it’s in there super solidly.

I actually used my pi4 for the dfu-programmer part - super painless that way.

Glad to hear you got it working!

[Gerald_Stevens]

Should the USB jack just rest below the panel? The tabs won’t fit into the slot without bending… which I will screw up. But it seems that sitting under the panel, the jack is a bit taller than the standoff from a Thonkiconn jack which sets the panel height for the rest of the board. Thoughts?

@tehn

quickly looking into the other issues:

On the subject of light pipes - how long should they be? Here are some 8.9 mm ones that Mouser stocks : 749-PLP1-350-F

This Alpha pot is OK - but has a round shaft / no D
311-1940F-15R1-B10K

[tehn]

@Gerald_Stevens

On the subject of light pipes - how long should they be? Here are some 8.9 mm ones that Mouser stocks : 749-PLP1-350-F

PLP1-350-F is the correct part number. updated BOM

---

if your PCB doesn’t have slots for the USB, yes, bend the legs. or cut them flush and SMD them with a lot of solder.

[Gerald_Stevens]

I was talking about how the USB fits into the front panel. It has right angle tabs that keep it from going into the rectangular panel slot. Is it supposed to sit underneath the panel with no engagement?

[thopa]

I think so, yes, when soldered the usb plug aint going anywhere.

[jmsiener]

Yeah, it does make the bottom part of the panel sit like .5mm above the jacks but it works out well enough. I feel like the standoff listed in the BOM is standard size so it’s a bit taller than 10mm.

[philomates]

Hi there, thanks for all the details here. I wanted to add my experience doing this build with a friend to potentially help others.

We ordered 3 per-populated PCBs from pusherman.

as most mentioned we needed these parts:
C5 (the one on the board is non polarized), PF1, PF2, Y1, RN1, U1, IC4, U40, U2, U3, U4, IC2, USB connector, POP (switch), JP1, JP2, the 10k pot, the thonkiconns jacks, the screen, the screen headers, and lightpipes

additionally, as some mentioned, we needed to buy R37, R38, R46, R47, R52, R53, R66, R67 (20k resistors)

I don’t think others mention this, but C9, C48, C55 were also missing on our boards

we also ordered R43, R44, and C2, but didn’t need to use them

later I noticed one of the boards had a busted 100n cap and I replaced it with a through-hole I had laying around to avoid shipping costs

My advice is to order the PCB and check out what is missing (closely, we missed some parts the first time) because every board seems slightly different or has faults.

Initially, when we turned on the built modules and plugged the USB into the computer, they weren’t recognized. After a lot of debugging, we tracked down various shorted pins on the different builds. Fixing them made the module show up on our computer using the lsusb command.

This is my first monome-based hardware and I wouldn’t have gotten into it if it wasn’t open and diy, so my deep gratitude and appreciation to everyone who made this possible; very inspiring

[philomates]

Does anyone have tips on testing and calibrating the CV IN?
I found (Teletype) IN / PARAM Calibration (Done) which seems to suggest that the input range is 0 - 10 V, but I quickly looked up the AD7923 analog-to-digital chip’s specs and it seems to say the chip can only read up to 2.5V.

Anyways, when using IN.CAL.MIN with 0V and IN.CAL.MAX with 10V but then when I read values back via IN it maxes out at ~2.5 V

[bpcmusic]

I haven’t looked at the TT’s schematics, but the standard way to do this is to scale the input voltage to a range that the ADC can read. This, 0V-10V becomes 0V - 2.5V.

The TXi does something like this and actually gets the values inverted at the ADC. In software it flips and scales its understanding of them to represent -10V to +10V. (For example, 0V = +10V, 1.65V = 0V, and 3.3V = -10V.)

[Gerald_Stevens]

OK - finally had the courage to test what’s working and what isn’t. I’m kinda f*ck3d because I hard-soldered the screen on - I’ll buy the low profile 0.1" header @okyeron recommended and try to outfit it with that so I can debug. Good news is the firmware / OS is working - that’s a LOT of tiny components I must have gotten right. Bad news is - CV3 and CV4 outputs aren’t working, and TR2 isn’t putting out 5V. And PARAM only varies from 15000 to 16000, which is borked. Any recommendations on which chips are responsible for these outputs / and the PARAM pot?

This is all doable de-bugging wise, key thing is everything works well enough to pinpoint what doesn’t work. It’ll be a bit of a slog, but I’m pretty stoked having dug into teletype’s manual.

[KitKatAndy]

There are some AD-converters and the main brain microchip as far as I get into it. I suggest take a few brief looks into the eagle file and follow all the signal lines which don´t work. I would suggest to start at U2 pins. Perhaps simple reflow of corresponding pins will help.
Concerning the param pot: Do you use a 10k b? If yes, I don´t have a clue right now whats going on. Checking all corresponding resistors / cold joints.?

[jmsiener]

Well if CV 3&4 aren’t working I believe they’re coming from the same chip (AD5687) so that’s a good place to start looking but unfortunately they’d be beneath the screen. The CV in and Param knob are handled by the AD7923. Looks like TR 2 is buffered-amp’d by the tl074-make sure the 20k resistor is present or there will be no feedback. Hope you get it man!

[Gerald_Stevens]

I reflowed all those chips with added flux. Still have problems. I noticed that partially inserting a plug causes the CV LED to light up, probably as it briefly shorts tip to sleeve. I need to look at the schematic to see how this can be. I’ll probably rip the offending jacks out so I can check the tiny resistors and LEDs. Already ripped the screen off.

[reijo]

at which voltage should the CV led start to shine?

I used different leds and they shine as soon as I do CV 1 V 3

So I was wondering if this is normal behavior or if I should change the led resistor values

---

Oh and another thing @tehn:

If you use the ocotpark link to order from mouser you end up getting the wrong amount of 1k und 10k resistors. It´s kind of weird. the octopart list says 16 10k and 8 1k resistors (which is the right amount)
but if you click on “buy now” you end up having 8 10k and 16 1k resistors in your basket…

Edit:
had another look at the octopart list. 1k and 10k are reversed.

[frankchannel]

@reijo

I used different leds and they shine as soon as I do CV 1 V 3

So I was wondering if this is normal

I think that is normal. The forward voltage of the LEDs is between 2-3 volts, below that they don’t conduct.

[jmsiener]

The TT I built doesn’t light up until CV 1 VV 240 making sense with ~2.3V forward voltage. Seems normal, output measures correctly.

[reijo]

Okidoki. Thanks for your replies.
Then I‘m having a fully working TT

[Gerald_Stevens]

Still debugging why outputs CV3 and CV4 are dead. U3 isn’t putting out the 2.5V it should like its brother U4 does. Both are getting their 3.3V. Why does U3 get VREF from U7 but U4 doesn’t? Those two pin 1’s on the 2 chips are totally different voltages as a result.