IBM Mainframe terminal stuff

Mike Ross tmfdmike at
Fri Dec 18 21:01:29 CST 2015

I've just repaired my 3277 which had lain idle for Goddess knows how
long - decades possibly.

I stripped, cleaned, and inspected it. I found a loose capacitor - one
of the tiny black rectangular ones - lying in the bottom of the thing.

I found out where it came from on the analog board and put a new one in.

Powered it up slowly on the variac. Nada. Flyback wasn't getting power
but all PSU voltages checked out ok.

Then I found the PSU wiring on the analog board - those edge
connectors - was bad; the connector was aged brittle plastic and had
cracked. I bodged a fix for that, tried again, and it came right up.

When I powered it down, I couldn't get it back up again initially...
found I could manage it by dialing the variac back down to the point
where the screen started to roll, then bringing it back up again. The
whole thing was very flaky. Left it running for several hours and now
it's fine; powers right up and comes online every time - touch wood!
Guess the capacitors needed time to settle down. But I agree, not one
of IBM's more impressive products - and that squares with what I've
heard from others too.

Here's mine running:

I've got another one coming next week - this one in much nicer
cosmetic condition complete with all the panels - the one I have is
'naked'. This will leave me with two (hopefully) working 3277s but
only one keyboard between them; the new one is coming without a
keyboard - anyone got a spare? And no the 3278/3279 keyboards will NOT


On Sat, Dec 19, 2015 at 3:45 PM, Paul Berger <phb.hfx at> wrote:
> Its possible, but you need a modem eliminator mostly to provide the clocks
> for the sync data line, other than that is is similar to a null modem.
> I remember the control units being pretty much trouble free however the 3277
> terminals,3284 and 3286 printers where a different story.  What I remember
> of the 3277 is they seemed to never have solid problems always seemed to be
> intermittent.  We would swap groups of cards between 3277s in an attempt to
> isolate and leave notes in the terminals saying what cards where swapped and
> why, and you would find notes from years before buried in the dust from
> people who had done the same thing before.  The 3284 and 86 printers could
> be a bit of a nightmare, like the rest of the system they where all hardware
> and the printing was timed by toothed where and emitters that detected the
> teeth playing with them usually led to trouble.  I remember times when the
> carrier would take off at full speed and slam into the right frame reverse
> and slam into the left frame while you scrambled to get to the power switch.
> The 3287 printer with the adapter to attach to 3271/2 was a God send.
> Paul.
> On 2015-12-18 10:32 PM, Mike Ross wrote:
>> I have a 3271 Model 2 - the remote bisync one. It's in embarrassing
>> condition; I'll take pics once it's cleaned up! Only IBM product I
>> have with significant wood in it; the top panel of the unit is made of
>> formica-covered chipboard!
>> No docs.
>> I wonder if it might be possible to finagle it into working with my
>> 3277 by a direct hookup to a serial port on a P/390, or a machine
>> running Hercules?
>> Mike
>> On Sat, Dec 19, 2015 at 2:39 PM, William Donzelli <wdonzelli at>
>> wrote:
>>> Do you have a 3272? I found one during the cleanup of the Black Hole.
>>> I think I can get ALDs and docs for it.
>>> --
>>> Will
>>> On Fri, Dec 18, 2015 at 8:30 PM, Mike Ross <tmfdmike at> wrote:
>>>> grrrr 3271 and 3272 controllers I should have said! They're much
>>>> dumber than 3174 - no program load device whatever; all done in
>>>> hardware. Just an on/off switch and a couple of status lights.
>>>> Mike
>>>> On Sat, Dec 19, 2015 at 2:27 PM, Mike Ross <tmfdmike at> wrote:
>>>>> Oh sounds like fun!
>>>>> I'm working with 3277 which uses a different protocol; connects to
>>>>> older mainframes (360/370) via 3171/3172 controllers - channel
>>>>> attached or remote. Also connects to System/3 direct to the CPU
>>>>> hardware as a console. I have a 3172 controller; physically rough and
>>>>> needs restoration but *should* work if I can fake the remote
>>>>> connection and modem - bisync etc. But just now I'm working with
>>>>> Henk's 3277->RS232 converter - trying to get it hooked up to an
>>>>> emulated System/3 -
>>>>> I've played this game a lot - including telnetting into Linux from
>>>>> 3270 screens - see
>>>>> Please let me know when you manage to ship those 3174 diskettes to
>>>>> Henk; I know he's keen to get them ASAP.
>>>>> Mike
>>>>> On Sat, Dec 19, 2015 at 12:00 PM, Guy Sotomayor <ggs at>
>>>>> wrote:
>>>>>> I’ve been tinkering around with my IBM Mainframe terminals and I’m
>>>>>> pretty close to having everything the way that I want it.
>>>>>> I currently have an IBM 3174-61R Establishment controller connecting
>>>>>> to my “mainframes” via TCP/IP.  This allows me to not only connect to my
>>>>>> Multiprise 3000 S/390 but also some PCs running Hercules to allow me to run
>>>>>> older versions of the mainframe OS’s.  I currently have 3 IBM 3179 terminals
>>>>>> hooked up.
>>>>>> I’ve upgraded the 3174 to be able to boot from a hard disk (it was a
>>>>>> supported option).  I actually purchased the upgrade from a reputable vendor
>>>>>> but after many attempts we could not actually get it to do anything useful.
>>>>>> I learned about DavidG’s MFM emulator and that’s now installed in the 3174
>>>>>> which happily boots from it.  I still have to figure out a permanent mount
>>>>>> for it since the standard 3174 HDD mounts don’t work for the MFM emulator
>>>>>> board.
>>>>>> One of the last things that I wanted to figure out was how to support
>>>>>> multiple sessions.  I had configured the 3174 to support multiple sessions
>>>>>> on each terminal but following the documentation, I could never figure out
>>>>>> how to actually switch between the sessions.  Well today I managed to do
>>>>>> that.  Success!  I finally figured out the key sequence for switching
>>>>>> sessions.  It doesn’t match the IBM documentation.  It’s not clear to me if
>>>>>> that’s because I haven’t configured something properly on the 3174 or if
>>>>>> there’s some other reason behind it.
>>>>>> Since I also have the AEA feature that allows me to connect ASCII
>>>>>> terminals to the 3174, my next project will be to hook up a couple of the
>>>>>> terminals that’re deemed “compatible” (VT100, ADM5 are on the list) and see
>>>>>> how that works.  But that also involves navigating the 3174 configuration
>>>>>> process which is opaque to say the least (there are configuration screens
>>>>>> but you need a manual because the fields only have numbers (e.g. 721) to
>>>>>> identify them and no other indication as to how they need to be filled in).
>>>>>> Oh, the other bizarre thing about the 3174 that’s been configured with
>>>>>> TCP/IP is that I can also connect to “unix” servers with the 3174’s (and the
>>>>>> ASCII terminals too).  It’ll be interesting to see how that works on a CUT
>>>>>> mode terminal.  ;-)
>>>>>> TTFN - Guy
>>>>> --
>>>>> 'No greater love hath a man than he lay down his life for his brother.
>>>>> Not for millions, not for glory, not for fame.
>>>>> For one person, in the dark, where no one will ever know or see.'
>>>> --
>>>> 'No greater love hath a man than he lay down his life for his brother.
>>>> Not for millions, not for glory, not for fame.
>>>> For one person, in the dark, where no one will ever know or see.'

'No greater love hath a man than he lay down his life for his brother.
Not for millions, not for glory, not for fame.
For one person, in the dark, where no one will ever know or see.'

More information about the cctalk mailing list