My Otari MX-80 Journey.

Found a nice video of switch maintenance/cleaning for the CB140.

https://www.youtube.com/watch?v=52AFE4pLnhE

So I pulled and disassembled all the red  Armed/Safe switches (BL1D-2M4) and found many of them to be pretty messed up. There's a pin that's spring loaded, and has a narrow tip with a plastic cap on the end, which is pushes the innards around to actuate the  switch. The plastic caps on many have fallen off, some are broken, the tips of some pins are bent or broken, and even though the switches still worked, they are unreliable, and best replaced. Of course, the BL1D and BL1E are discontinued, but Digikey carries the BT1D and BT1E, which are the same form factor, but a toggle instead of the paddle rocker on the BL versions.

So I bought 26 of each,  BT1D -2M4-Z and BT1E-2M4-Z at ~$5 each, (ouch), and tore one apart for comparison. Pics below.

The footprint and height are exactly the same, which is the main thing, as is the function. On disassembly however, I found the guts differ just enough to make them less than directly interchangeable.

The pins on the original are stepped and bored out on one end and hold the spring inside, and are then inserted into the rocker actuator.  On the new ones, the design is simplified, with a straight pin, no bore, and a slightly wider diameter spring of equal length. Unfortunately this precludes just swapping pins from the toggle to the rocker, as the lack of bore on the pin makes it protrude too far to use without modification.

I'm not feeling picky and considering the cost of these, I'd rather not modify them and risk causing myself trouble down the road. I'm fine with the toggles, but I'd still like them to be red, so I'll get rubber toggle caps for them.  My other concern is the big hole exposed by the lack of the big plastic paddle rocker. I think I'll just cut a strip of plastic to run across the underside of the remote control shroud with holes drilled for the toggle switches. Sounds like a bit of work, so maybe after I've gotten things functional I'll tidy that up.

Carry on the good work, better a working remote than nonworking!

radardoug said:

Carry on the good work, better a working remote than nonworking!

Click to expand...

Thanks Doug. The support is most appreciated.

So, even though things are looking good with my CB-140 remote (once I can get back to soldering), I'm still trying to figure out if this thing can be used without one via Serial control. All signs point to yes, though info on this, and the actual the ASCII commands mentioned in the manual are elusive.

The manual's wiring diagram for the remote suggests an RS422 connection plus some other stuff, as there are Transmit and Receive A and B connections. So what I want to do now is splice into those serial connections, without disrupting the rest (+24V, Tach Pulse, and Fwd/Rev) and see if I can monitor communication between the Machine and the Remote in hopes of capturing commands.

I inspected my cable and it seems pretty straightforward. Looks like the wire is Mogami W2689 or at least whatever the version from 30 years ago was. Belden 9541 also seems like a reasonable alternative. So either I order some of this, or I butcher a 9 pin serial cable and use some 24awg 3 wire for whatever that won't cover. I've got a bunch of DB25 and DB9 M and F connectors, and some serial to USB adapters that should work. Will need to sort out the software side of things though.

This part of the project could be it's own thread, and may become one at some point, but for now I'll just keep it here.

Looking in other directions I've found bits of info here and there on running the MX-80 with a DAW and synchronizer. I stumbled onto a fresh ebay listing and got 6x Timeline Lynx units for around $80 shipped, which should show up in a few days. I'd need to cobble together cables for those as well, but was able to find info herehttp://www.technicalaudio.com/pdf/TimeLine/ so that's a possibility as well, but only for transport, which is still potentially useful, though ultimately being able to perform all of the functions of the Remote, or at the very least, Arming channels and the Input/Sel-Rep/Repro switch functions, which is the whole point, since without that, an MX-80 missing the remote is dead in the water.

Still plenty of other maintenance to do, and need rollers replaced, which has stalled with the COVID-19 situation.

thecr4ne said:

So, even though things are looking good with my CB-140 remote (once I can get back to soldering), I'm still trying to figure out if this thing can be used without one via Serial control. All signs point to yes, though info on this, and the actual the ASCII commands mentioned in the manual are elusive.
...
Will need to sort out the software side of things though.

Click to expand...

Kermit was always the goto serial protocol software:

  http://www.kermitproject.org/

IIRC I used it to get files (using printer commands) off of an extremely old Data General computer once. Maybe it can help you.

I made a little RS422 to 232 converter when I was working on my Lexicon LARC project. Then ran that into an old DOS computer using Procomm. Then later I wrote some custom software in Turbo Basic on the same computer, so I could filter some of the continuous commands. Put a switch on there so I could read from the frame or the LARC.

KERMIT protocol was all about getting arbitrary data through limited character sets and often on noisy lines.
http://www.columbia.edu/kermit/kermit.html

Yes, your DG.

I don't see how it helps for short lines, except the KERMIT program did do a dumb data-terminal mode (like MS Terminal). Wouldn't you still have to send all possible character combinations to see what does what? (Unless there is a working remote you can wire-tap.)

I actually don't remember spit about kermit. I thought it was for all-things-serial-communication. Didn't know it was it's own protocol.

Today I finally swapped all the switches in the CB-140 Remote unit. Once the old ones were removed it went pretty quick.  It's all back together, but I haven't connected it and fired it up yet as I've got some rollers removed from the machine at the moment. I'd still like to put a strip of something under the big square holes to make it look a little cleaner and prevent crap getting in, but that will be for another day.  Pics below

Serial control quest update:
I found a 15 pin control cable I had around and used it to make a ~1ft DB-25 M to F cable and added leads off the Receive-A/B/Ground on each end. The idea here being to hook each end to a USB to RS-232 adapter, via an RS-422/485 to RS-232 adpater, and then use software to monitor the commands. Software side of things TBD. I still haven't fired up the machine to test this or the new switches on the CB-140 because I removed the Tach roller in hopes of getting it re-done, but that stalled due to the choice of either waiting out the Viral hiatus of one vendor, or spending twice as much with another, maybe I'll put it back together as is to test one of these days, but I just haven't yet. Also, I noticed the CB-140's DB-25 connector doesn't have pins 19 and 21, which is encouraging in that it suggests everything is being communicated via serial...or maybe it means something else, but it doesn't really impact anything...hopefully. Anyway here's some pics.












Anyway, here's pics of my dumb little adapter. Here's hoping it actually gets me somewhere.

Today I finally fired this thing up again with the remote connected. On first attempt it was locked up with a few lights flashing and I thought I killed it. I double checked the cable connections and after powering the MX-80 off and on about 3 times, it locked in and worked.

I tested the functionality of the new switches. I found that the 3 position output switches all worked except 1 and 16, both of which are stuck on "Sel-Rep" when the controller is set to individual mode. Those channels do appear to follow the master buttons for "All Input", "All Sel-Rep" and "All Repro". Moving on to the "ready/Safe" switches, I flipped the "All Safe" switch off, and the remote locked up again.

I turned everything off, disconnected and opened the CB-140. I didn't see anything obviously wrong, but I went ahead and re-flowed the solder on the switch pins again  thinking maybe there some cold joints causing trouble. I also tested continuity on the problem switches to confirm that the switches themselves were fine. All good.

Closed it up, fired it up. Output switches unchanged, 1 and 16 still misbehaving.  Ready/Safe switches all appear to be working as expected.  Big sigh of relief.

I did open it up once more to try to diagnose the issue with Output switches 1 and 16, but nothing's obvious, and I didn't feel like pulling the switches altogether, so I moved on.

Next, after double and triple checking both the original Otari cable, and my little jumper with the serial "receives" split off at each end, I added that in line and hooked it up to the computer via the USB-RS232-RS422 adapter chain pictured in my last post, and ran an app called "Serial Port Monitor" which has a 14 day free trial, and tried out the monitoring.

The MX-80 side of the cable's receive lines were where I saw all the action. Once I opened the connection I saw a steady stream of what I'm pretty sure hacking a mainframe looks like.... Anyway, the ASCII output looks like this:
=deleted by JR===
That's just a small snippet, which I can't make heads or tales of. 

Being both mentally and emotionally exhausted, I'm calling this a successful day. CB-140 lives, even if there's still an issue to work out, and I think I have proof of concept on the serial command monitoring front. I think the learning curve beyond this is gonna be steep so I'll get back to it when I do.

Looking at that, it could be that you have the wrong baud rate. Try different baud rates and see if the data stream looks better.
Alternatively, they may be using characters that dont resolve to a readable ASCII stream. Then you will need to pick it apart and see  if there is logic to it. There are two data streams, to and from the machine. You need to read both. If you are getting a continuous stream it may mean the re is data like time being sent, or position. Also some systems will send an interogate character to make sure the other end is connected.

Thanks Doug,
I started with 9600 baud rate, as that is what's listed in the MX-80 manual elsewhere. I'll fiddle with it some more one of these days and report back.

I do most of my RS232 snooping using an old DOS computer I have, good old software available, and I can custom write software for specific snooping.  My program for snooping on the Lexicons can print the readable characters as ASCII, and then show the control characters as numbers.

Hi,
Could you tell me what kinf of switches do you use ?
Many thanks,
Cheers,
P.

The service manual specifies the correct lubricants to use.
You must use a reference level calibration tape to set up the machine as the first step in calibration, and setting reference level (essential to set record and bias level).
A full single speed version will show you frequency response.
The good news is the most of the calibration tapes made use tape formulations that last. I have use 25 year old tapes that have been run many, many time with no problems. As long as there is no sticky / shed you should be OK, so a used reference tape is not out of the question.
It is also important to resolve the machine with a synchronizer, if you want consistent speed. Many synchronizers had remote options that give full auto locator functions, I have a Zeta 3 remote you can have if you can obtain a Zeta 3!