emulating a FUTABA 20×2 VFD M202SD01HA lexicon pcm 80 81 90 91

Hello friends, I need to emulate a FUTABA display, and I need 10 digital lines in input. It’s a parallel data, and I need 8 Datalines in, 1 Write and 1 Busy signal (signals coming from the old device)

I want to use an STM32f103c8t6 + i2c OLED 1,3 display 128×64 (ssd1306 or sh1106, it’s indifferent)

the display is an OBSOLETE from 90’s FUTABA VFD 20×2 M202SD01HA and is not compatible with HD44780 protocol
and is from Reverb Lexicon PCM80 (or pcm81)

here the documentation about timing and data transfer

the data transfer is happening using parallel data transfer

so I need some suggestion about libraries that I can adapt to do the parallel data transfer.

data coming from the pcm80 should be interpreted and transferred to the OLED

I measured, and the space is RIGHT!

thank you!
synthi

this is the current display
like this one

Hi, @synthi and welcome.

I would be wildly surprised if any “Arduino display library mapped to the VFD commands.” That being said, I have worked with the OLED displays and their libraries are straight-forward.

Assuming that Arduino is the right platform (I think there are better solutions, more later), then there are two obvious ways to pull-off this translation: (very simplistic vuew, not fully baked)
– Bring in the parallel stream, decode instructions and data….
IF data, buffer the data character in SRAM, set a flag, return control to the state machine which will invoke the OLED library to update the OLED display.
IF instruction, decode and translate to 1 of more OLED library calls, set a flag, return control to the state machine which will call the OLED library to pricess the instruction(s).

– Bring in the parallel stream, decode.
IF character send to OLED library, loop.
IF instruction, send to OLED library, loop.

Both approaches have strength and weaknesses.

IMO, Arduino is not my 1st choice. I think FPGA is the most desirable if your design has volume. If your design is just 1-off for your personal use then maybe Arduino… but you’re in for a wild ride IMO.

Futaba sells a LCD emulator.

RU left-brain or right-brain … that is, are you up to the challenge? IMO, not a beginner project or even an intermediate project.

So, distilling if all, STM32duino & OLED may well fit your need. You will need to build the entire parallel logic interface both hsrdware and software. The OLED stuff is pretty well canned but I have no idea if the OLED library is optimized to be usable.

If you want to see OLED + STM32duino, I have some examples here: https://www.hackster.io/rayburne/projects

Ray

Hello Ray

Thank you for your suggestions and your experiments with OLED
Will be one shot project… and yes I’m up for the challenge
and I switched from Arduino to STM just because of the missing ports on the Arduino platform

the space is important too: I have only 99mm x 34mm for the substitution
(protoboard + stm32f103 + oled + cabling)

just an idea
The DISPD0 to DISPD7 needs to be connected to PA0 to PA7
DISPBSY to PB6
DISPWRT to PB7
DISPRXD to GND

I need some example from people having done some parallel polling

thanks
Antonio

synthi wrote:…
I need some example from people having done some parallel polling

thanks
Antonio

In my Github is the project
<https://github.com/sheepdoll/AVRVFDCLOCK.git>

This emulates a standard 14 pin LCD on a 14 segment alpha numeric VFD. I used the popular h44780 protocol so that it is drop in for any LCD. The protocol is basically a shift register.

A lot of the FUTABA vfd’s used this protocol. Others were basically a glass teletype display that used ASCII codes and escape sequences to position the cursor. I prefer the 14 pin version.

If I ever find that elusive thing called uninterrupted time I was working on connecting this to one of my STM32s.

I did the very same thing last year to fix an old guitar effects box with that exact display. I’ve attached my code. I used an arduino nano and a ST7735 LCD display, so this code probably wont work for you, but hopefully it should point you in the right direction.

thank you to everyone

I’m getting some result
I’ve attached an interrupt to the PB5 port on RISING (WRITE is HI)

the interrupt is
if the SELECT is LOW
setting BUSY HIGH
reading from the PA0 to PA7
setting BUSY LOW

but I’m getting very strange chars from the data read (invstigating)

uint8_t GetData()
{
uint8_t char_in =
(digitalRead(Data0) << 0) +
(digitalRead(Data1) << 1) +
(digitalRead(Data2) << 2) +
(digitalRead(Data3) << 3) +
(digitalRead(Data4) << 4) +
(digitalRead(Data5) << 5) +
(digitalRead(Data6) << 6) +
(digitalRead(Data7) << 7);

return char_in;
}

corrected some timing,
now the reading loop is this one

void interrupt()
{
if (!(GPIOB->regs->IDR & 0x0004)) // ignore if select signal is not present
{
//turn on the busy signal
GPIOB->regs->ODR |= 0b0001000000000000;

uint8_t data = GPIOA->regs->IDR & 0x00FF;

//process the new byte
processData(data);

delayMicroseconds(35);

GPIOB->regs->ODR &= ~(0b0001000000000000);
}
}

Ok, for that nearly 500USD device (current used market price) it’s a good idea to investigate a little bit further:
Some guys with the same problem got their lexicon to life again:
https://www.gearslutz.com/board/geekslu … oblem.html

I reseated all the ICs
the error is giving m is E4
that is “Display error”
In some way, the device knows I’m not using the original FUTABA

If I override the block, pressing UP and TEMPO, the reverb starts and works correctly

what I’m getting is some spurious character in the parallel port

for example I’m getting 0x30 when I’m expecting 0x20
but in other screen I’m getting the correctly 0x20 and 0x30 !

other thing:
one of the command for the FUTABA is “DP” (display position)
after DP there is always a byte between 0x00 and 0x27 (40 positions in the FUTABA)
but sometimes I’m getting wrong positions, like “252” or “168”
and is happening always in the same screens: I suspect there are some timing issues
OR the DATA port is not correctly cleaned after reading

thanks

madias wrote:Ok, for that nearly 500USD device (current used market price) it’s a good idea to investigate a little bit further:
Some guys with the same problem got their lexicon to life again:
https://www.gearslutz.com/board/geekslu … oblem.html

Ok, the device is alive, the rest are additional features
All in all this sounds like a job for:
http://www.aliexpress.com/item/1set-New … 33314.html
Beyond the analyzing things you can measure the real timing with it, maybe you are really a little bit off (but it’s strange that it happens in the particular screens.)
Display error: I would keep as a “childlock feature”.
If most of the remaining errors are reproduceable I would build as many exceptions in code as possible, because this code is only for this single device and not for mass productions.

DONE!!!!!
solved the problem of spourious chars
I cleaned the port for every read
now the full callback for reading is

void interruptFunc()
{
//turn on the busy signal
GPIOB->regs->BSRR = 0b0000000000010000;

uint8_t data = GPIOA->regs->IDR & 0b0000000011111111;

//clean bits
GPIOA->regs->BRR = 0b0000000011111111;

//process the new byte
processData(data);
//delayMicroseconds(200);

GPIOB->regs->BRR = 0b0000000000010000;
}

Any chance you can post the complete code in case anyone needs to fix their PCM80

RogerClark wrote:Any chance you can post the complete code in case anyone needs to fix their PCM80

Congratulations synthi!
It would be a shame dropping such a special sounding device because of a lousy display!
I hope that we will never see in the future Ebay “sellers” with a “Lexicon PCM80/81 Display upgrade kit for only 250USD” using your code (without credits).
Just one more thinking: Would a simple 2002LCD wouldn’t fit better into the lex? –> http://www.aliexpress.com/wholesale?cat … xt=lcd2002

madias wrote:Congratulations synthi!
It would be a shame dropping such a special sounding device because of a lousy display!
I hope that we will never see in the future Ebay “sellers” with a “Lexicon PCM80/81 Display upgrade kit for only 250USD” using your code (without credits).
Just one more thinking: Would a simple 2002LCD wouldn’t fit better into the lex? –> http://www.aliexpress.com/wholesale?cat … xt=lcd2002

I started buying the STM32F103 ($2.50 !!!) because I had to buy an adapted display for €180 + shipping

so I decided, let’s learn something…

attached a little video where I’m using a i2c 0.96 bicolor oled (128 x 64) ($4.23 !!)
I did 2 different text size experiment (but the display is still TOO SMALL! like a quarter of dollar)

you can see the speed of changes

madias wrote:Congratulations synthi!
It would be a shame dropping such a special sounding device because of a lousy display!
I hope that we will never see in the future Ebay “sellers” with a “Lexicon PCM80/81 Display upgrade kit for only 250USD” using your code (without credits).
Just one more thinking: Would a simple 2002LCD wouldn’t fit better into the lex? –> http://www.aliexpress.com/wholesale?cat … xt=lcd2002

FYI

during the weekend I did a protoboard and I missed completely measures…
I forgot that the connector protruded for about 20 mm !

I need a female connector with 90 deg of rotation (parallel to the board)

currently I used velcro to maintain the screen in place
and the idea is to fix the (insulated) protoboard too with velcro, to the top panel

IMG_5502.JPG (36.34 KiB) Viewed 1044 times

for the lovers

I’m attaching the PSEUDO schematics and the BIN file for and OLED 128×64 with I2C SH1106 controller (works on SSD1306 too, but I tested more on 1106)

the 2 switches in the scheme let you play with 4 different font configurations

OFF OFF 2 lines small << —– best
ON OFF 2 lines big 1 small
OFF ONN 1 small 2 big
ON ON 4 lines big

FutabaEmulation.pdf

(60.68 KiB) Downloaded 82 times

in the PDF , the IC / connector labelled U? does not seem to be connected ??

the object labeled PCM80_to_FUTABA port is just a LEGEND for the connector 2×10 (that is connected to the STM32)

OK

Thanks