Re: VAX 11/730 quickie
On 2/24/2016 1:11 PM, Mike Ross wrote: Why are you using MRSP mode (-m switch)? Do you know that is absolutely required? For VAX-11/730 console yes it is absolutely required as far as I know; everything I've read about people using emulation there says that it is. Mike Yes it appears that is correct. Going back to this thread: http://www.vcfed.org/forum/showthread.php?36113-Problems-using-tu58em-with-VAX-11-725-VAX-11-730/page13 there was a whole lot of discussion on bringing up a '730 with TU58EM, and the updates to it (to fix some time delays) to support the TU58 device on the '730 with TU58EM. I have since updated mainline TU58EM to version 1.4m to add Mark Blair's VAX 730 timing and background mode changes (enable via -x and -b switches). Source and cygwin executable here: https://github.com/AK6DN/tu58em Functionally this should be the same as Mark's version 1.4j + his changes. Don
Re: VAX 11/730 quickie
On Thu, Feb 25, 2016 at 9:56 AM, Don Northwrote: > On 2/24/2016 3:43 AM, Mike Ross wrote: >> >> On Tue, Feb 23, 2016 at 4:26 AM, Mike Ross wrote: >>> >>> On Mon, Feb 22, 2016 at 10:14 AM, Mark J. Blair wrote: > > On Feb 21, 2016, at 23:09, Mike Ross wrote: > Actually I do have a Mac within easy range of the 730. Could you do me > a favour and throw a prebuilt OSX binary somewhere I can grab it? I > have flaky internet in the workshop and this Mac isn't set up with > Xcode or any other dev environment... https://dl.dropboxusercontent.com/u/29876211/tu58em.zip >>> >>> Thanks Mark... I played with this. I built a 3-wire cable; I'm sure I >>> have it wired right because traffic happens - and if I disconnect, or >>> switch Rx/Tx, it just sits there but this is as far as it gets: >>> >>> MikeUpstairs:~ Imac$ ./tu58 -s 38400 -p /dev/tty.usbserial -d -v -m -x >>> -r vax73058 >>> info: unit 0 rfile 'vax73058' >>> info: tu58 tape emulator v1.4j (NF6X fork) >>> info: (C) 2005-2014 Don North , (C) 1984 Dan >>> Ts'o >>> info: serial port /dev/tty.usbserial at 38400 baud >>> info: MRSP mode enabled (NOT fully tested - use with caution) >>> info: TU58 emulation start >>> info: R restart, S toggle send init, V toggle verbose, D toggle debug, Q >>> quit >>> info: emulator started >>> >>> (I ctrl-c the VAX) >>> >>> info: flag=0x00 last=0x00 >>> info: seen >>> info: flag=0x00 last=0x00 >>> info: seen >>> info: flag=0x04 last=0x00 >>> info: seen >>> info: flag=0x04 last=0x04 >>> info: seen >>> info: seen, sending >>> >>> (?27 DEVICE ERROR on VAX) >>> >>> (I ctrl-c the VAX again) >>> >>> info: flag=0x00 last=0xFF >>> info: seen >>> info: flag=0x00 last=0x00 >>> info: seen >>> info: flag=0x04 last=0x00 >>> info: seen >>> info: flag=0x04 last=0x04 >>> info: seen >>> info: seen, sending >>> >>> (?27 DEVICE ERROR on VAX) >>> >>> (repeat) >>> >>> Can you spot me being stupid? Does my command line look sane re. >>> various options? >>> >>> (I've trivially renamed the tape image file for faster typing >>> otherwise it's straight out of the box) >> >> Just for giggles I decided to try Bela Torok's Arduino-based emulator: >> >> http://www.torok.info/computing/pdp11/tu58/ >> >> I whipped that up from bits tonight and got it working surprisingly >> quickly. But the VAX still doesn't want to play; from the behaviour >> and messages it's falling over at exactly the same place. The Arduino >> serial console displays: >> >> Continue after 2 INIT flags. online! >> >> And VAX shows the now-familiar ?27 DEVICE ERROR... >> >> Which looks a HELL of a lot like how tu58em is ending. So I don't >> think the problem is with tu58em... >> >> Is there anything I should be looking at on the VAX end here? Any >> configuration or cabling details that could be screwing things up >> before I drag the RS232 analyser out? I've tried changing the 'delay' >> parameter on the Arduino to various values between 0 and 255 to no >> avail - and I've tried a couple of different tape images in case I got >> a bad one... >> >> Mike >> >> http://www.corestore.org >> '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.' >> > > Why are you using MRSP mode (-m switch)? Do you know that is absolutely > required? For VAX-11/730 console yes it is absolutely required as far as I know; everything I've read about people using emulation there says that it is. Mike http://www.corestore.org '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.'
Re: VAX 11/730 quickie
On 2/24/2016 3:43 AM, Mike Ross wrote: On Tue, Feb 23, 2016 at 4:26 AM, Mike Rosswrote: On Mon, Feb 22, 2016 at 10:14 AM, Mark J. Blair wrote: On Feb 21, 2016, at 23:09, Mike Ross wrote: Actually I do have a Mac within easy range of the 730. Could you do me a favour and throw a prebuilt OSX binary somewhere I can grab it? I have flaky internet in the workshop and this Mac isn't set up with Xcode or any other dev environment... https://dl.dropboxusercontent.com/u/29876211/tu58em.zip Thanks Mark... I played with this. I built a 3-wire cable; I'm sure I have it wired right because traffic happens - and if I disconnect, or switch Rx/Tx, it just sits there but this is as far as it gets: MikeUpstairs:~ Imac$ ./tu58 -s 38400 -p /dev/tty.usbserial -d -v -m -x -r vax73058 info: unit 0 rfile 'vax73058' info: tu58 tape emulator v1.4j (NF6X fork) info: (C) 2005-2014 Don North , (C) 1984 Dan Ts'o info: serial port /dev/tty.usbserial at 38400 baud info: MRSP mode enabled (NOT fully tested - use with caution) info: TU58 emulation start info: R restart, S toggle send init, V toggle verbose, D toggle debug, Q quit info: emulator started (I ctrl-c the VAX) info: flag=0x00 last=0x00 info: seen info: flag=0x00 last=0x00 info: seen info: flag=0x04 last=0x00 info: seen info: flag=0x04 last=0x04 info: seen info: seen, sending (?27 DEVICE ERROR on VAX) (I ctrl-c the VAX again) info: flag=0x00 last=0xFF info: seen info: flag=0x00 last=0x00 info: seen info: flag=0x04 last=0x00 info: seen info: flag=0x04 last=0x04 info: seen info: seen, sending (?27 DEVICE ERROR on VAX) (repeat) Can you spot me being stupid? Does my command line look sane re. various options? (I've trivially renamed the tape image file for faster typing otherwise it's straight out of the box) Just for giggles I decided to try Bela Torok's Arduino-based emulator: http://www.torok.info/computing/pdp11/tu58/ I whipped that up from bits tonight and got it working surprisingly quickly. But the VAX still doesn't want to play; from the behaviour and messages it's falling over at exactly the same place. The Arduino serial console displays: Continue after 2 INIT flags. online! And VAX shows the now-familiar ?27 DEVICE ERROR... Which looks a HELL of a lot like how tu58em is ending. So I don't think the problem is with tu58em... Is there anything I should be looking at on the VAX end here? Any configuration or cabling details that could be screwing things up before I drag the RS232 analyser out? I've tried changing the 'delay' parameter on the Arduino to various values between 0 and 255 to no avail - and I've tried a couple of different tape images in case I got a bad one... Mike http://www.corestore.org '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.' Why are you using MRSP mode (-m switch)? Do you know that is absolutely required?
Re: Serial analyzers (was Re: VAX 11/730 quickie)
On 2/24/2016 10:59 AM, Eric Smith wrote: On Tue, Feb 23, 2016 at 8:52 PM, jwsmobilewrote: For serial I use a Saleae 8 bit analyzer. [...] You do have to use a level shifter with the thing because it is designed for 3.3v logic and needs to be adjusted accordingly. The Saleae Logic Pro 16 (and probably the Logic Pro 8, but I haven't checked) are designed for an operating input voltage range of +/-10V, with an abs max rating of +/-25V. Saleae specifically states that it can be directly connected to RS-232. We had the analyzers out "bleeding edge" and the originals were not differential and were not +5 safe. They tolerated the differential 0-5 we had at one point in the circuit, and the engineer I worked with said "hm, you are lucky" with that hookup. I was not using the Pro, that is way late in their product line for me. their newer 4 channel version does seem to also be differential and RS232 voltage level compliant. (FWIW I was on RS485 differential, line levels same as RS232, but function not). Thanks Jim The Logic Pro models are more expensive than the other (amateur?) models, but support analog input on all channels. I now use a Logic Pro 16 for most of my simple logic analyzer requirements, but still use an Agilent 16700 logic analyzer mainframe when I need more than 16 channels.
Re: Serial analyzers (was Re: VAX 11/730 quickie)
On Tue, Feb 23, 2016 at 8:52 PM, jwsmobilewrote: > For serial I use a Saleae 8 bit analyzer. [...] > You do have to use a level shifter with the thing because it is designed for > 3.3v logic and needs to be adjusted accordingly. The Saleae Logic Pro 16 (and probably the Logic Pro 8, but I haven't checked) are designed for an operating input voltage range of +/-10V, with an abs max rating of +/-25V. Saleae specifically states that it can be directly connected to RS-232. The Logic Pro models are more expensive than the other (amateur?) models, but support analog input on all channels. I now use a Logic Pro 16 for most of my simple logic analyzer requirements, but still use an Agilent 16700 logic analyzer mainframe when I need more than 16 channels.
Re: VAX 11/730 quickie
On Tue, Feb 23, 2016 at 4:26 AM, Mike Rosswrote: > On Mon, Feb 22, 2016 at 10:14 AM, Mark J. Blair wrote: >> >>> On Feb 21, 2016, at 23:09, Mike Ross wrote: >>> Actually I do have a Mac within easy range of the 730. Could you do me >>> a favour and throw a prebuilt OSX binary somewhere I can grab it? I >>> have flaky internet in the workshop and this Mac isn't set up with >>> Xcode or any other dev environment... >> >> https://dl.dropboxusercontent.com/u/29876211/tu58em.zip > > Thanks Mark... I played with this. I built a 3-wire cable; I'm sure I > have it wired right because traffic happens - and if I disconnect, or > switch Rx/Tx, it just sits there but this is as far as it gets: > > MikeUpstairs:~ Imac$ ./tu58 -s 38400 -p /dev/tty.usbserial -d -v -m -x > -r vax73058 > info: unit 0 rfile 'vax73058' > info: tu58 tape emulator v1.4j (NF6X fork) > info: (C) 2005-2014 Don North , (C) 1984 Dan > Ts'o > info: serial port /dev/tty.usbserial at 38400 baud > info: MRSP mode enabled (NOT fully tested - use with caution) > info: TU58 emulation start > info: R restart, S toggle send init, V toggle verbose, D toggle debug, Q quit > info: emulator started > > (I ctrl-c the VAX) > > info: flag=0x00 last=0x00 > info: seen > info: flag=0x00 last=0x00 > info: seen > info: flag=0x04 last=0x00 > info: seen > info: flag=0x04 last=0x04 > info: seen > info: seen, sending > > (?27 DEVICE ERROR on VAX) > > (I ctrl-c the VAX again) > > info: flag=0x00 last=0xFF > info: seen > info: flag=0x00 last=0x00 > info: seen > info: flag=0x04 last=0x00 > info: seen > info: flag=0x04 last=0x04 > info: seen > info: seen, sending > > (?27 DEVICE ERROR on VAX) > > (repeat) > > Can you spot me being stupid? Does my command line look sane re. > various options? > > (I've trivially renamed the tape image file for faster typing > otherwise it's straight out of the box) Just for giggles I decided to try Bela Torok's Arduino-based emulator: http://www.torok.info/computing/pdp11/tu58/ I whipped that up from bits tonight and got it working surprisingly quickly. But the VAX still doesn't want to play; from the behaviour and messages it's falling over at exactly the same place. The Arduino serial console displays: Continue after 2 INIT flags. online! And VAX shows the now-familiar ?27 DEVICE ERROR... Which looks a HELL of a lot like how tu58em is ending. So I don't think the problem is with tu58em... Is there anything I should be looking at on the VAX end here? Any configuration or cabling details that could be screwing things up before I drag the RS232 analyser out? I've tried changing the 'delay' parameter on the Arduino to various values between 0 and 255 to no avail - and I've tried a couple of different tape images in case I got a bad one... Mike http://www.corestore.org '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.'
Re: Serial analyzers (was Re: VAX 11/730 quickie)
On 2/23/2016 7:11 PM, Curious Marc wrote: I was wondering if I should add a 4952 to my HP collection. It's tempting, these are cute machines. But except if I am using synchronous RS232, I was not sure what I would getting that a laptop with a good terminal emulator and a serial port would give me. Can you convince me otherwise? What do you guys use it for? Marc On Feb 22, 2016, at 3:58 PM, Jay Westwrote: I've always eyed the 4952... being as my penchant is HP and much of my test equipment is same-vintage HP gear (I'll give a vote for the 1631D logic analyzer, combo LA and digital scope - the scope is sorta poor, but handy - the LA is great for what I work on). However, I spent a significant portion of my career in front of a serial analyzer so I'm pretty familiar with them [...] J For serial I use a Saleae 8 bit analyzer. It is an 8 bit logic analyzer, but that just means you get 8 capture channels. It will do any combination of serial, I2C, and SPI out of the box. CAN bus is extra. Recording capacity is till the disk fills up. You can export traces to files of text. You do have to use a level shifter with the thing because it is designed for 3.3v logic and needs to be adjusted accordingly. A note on knockoffs. Please buy the original if you can, as the company has not cut off the use of their software with knockoffs of the analyzer. (Off topic of this thread, open new thread if it is something to reply to, please) Thanks jim
RE: Serial analyzers (was Re: VAX 11/730 quickie)
Marc wrote I was not sure what I would getting that a laptop with a good terminal emulator and a serial port would give me. Can you convince me otherwise? What do you guys use it for? - Perhaps there are terminal emulators out there that can do the following, but I'm not aware of any that do these things that are the most common use cases 1) Display transmit on one line and receive on the other line, positionally correct. Very easy to visually see what one side sent, and what the other side replied. 2) Display the characters not just in ascii text, but in ascii mnemonics, and better yet... binary, octal, hex. FAR easier to see what is REALLY being transmitted and received. Ex - so if the unit is set to 81N, are the parity bits right? You can tell visually at a glance without really thinking. 3) And given the display can be in binary, it's easy to see field values that do not align on even count positions - ex: 8 bits where the first 3 bits mean something in the protocol, and the next field is 5 bits, etc. 4) Automatically trigger on events - ex. Watch the line until you see this sequence from the DCE. Then insert this string as a reply, wait for this response, then let the DTE continue. Oh, and start logging if this pattern occurs 3 times. 5) Automatically calculate checksums by any one of a designated set of algorithms, and verify the checksums being sent and received. Start logging, up to 2K size, once a checksum mismatch occurs. 6) Run BERT tests to either side. 7) Often provide a breakout box to jumper and/or reroute wires. 8) Save data to a disk for later review, or print to paper. 9) Easily deal with async, sync, bsync, x.25, etc. When writing your own communications protocols, a datascope makes testing/troubleshooting a very quick process. The alternative ... not so much. Many decades ago, I had a program for DOS (that I have since lost) called "Breakout II". It came with a cable and required two serial ports. It was a "software" breakout box. It was sorta handy... but it was no datascope. J
Re: Serial analyzers (was Re: VAX 11/730 quickie)
I was wondering if I should add a 4952 to my HP collection. It's tempting, these are cute machines. But except if I am using synchronous RS232, I was not sure what I would getting that a laptop with a good terminal emulator and a serial port would give me. Can you convince me otherwise? What do you guys use it for? Marc > On Feb 22, 2016, at 3:58 PM, Jay Westwrote: > > > I've always eyed the 4952... being as my penchant is HP and much of my test > equipment is same-vintage HP gear (I'll give a vote for the 1631D logic > analyzer, combo LA and digital scope - the scope is sorta poor, but handy - > the LA is great for what I work on). > > However, I spent a significant portion of my career in front of a serial > analyzer so I'm pretty familiar with them [...] > > J > > > > >
RE: Serial analyzers (was Re: VAX 11/730 quickie)
I had written... > Just my 2 millidollars worth... To which Tony replied -- Don't you mean 20 millidollars? -- I'd have intended to say 2 centidollars... but if I say I meant 20 millidollars then I can claim it was merely a typo and I dropped the 0 *cough* J
RE: Serial analyzers (was Re: VAX 11/730 quickie)
> > I've always eyed the 4952... being as my penchant is HP and much of my test > equipment is same-vintage HP gear (I'll give a vote for the 1631D logic > analyzer, combo LA and digital scope - the scope is sorta poor, but handy - > the LA is great for what I work on). I would love to find the upgrade kit to turn my 1630 into a 1631, but they are not exactly common. > However, I spent a significant portion of my career in front of a serial > analyzer so I'm pretty familiar with them. As much as I really WANT the 4952 > to be "it", it isn't - for me. Mainly because I have found that having a > unfolding or removable (or permanently in the usual spot) keyboard takes up > far too much bench space - or is unusable when standing the unit upright > (very often it was convenient to set the LA on the floor and on the models I > use - the keyboard and screen face straight up). Right... That is an interesting point. With the sorts of machines we work on you want to take the test gear to the machine and not vice versa and there is often nowhere convenient to put a large instrument I was given a unit badged by Black Box (I can get the details which is almost small enough to be handheld. It has an 8 line LCD display (in monitor mode, alternate lines give the data in each direction, one is in inverse video so it's quite clear) and a membrane keyboard with nice clicky metal domes. I forget the internals, but it's something like a Z180, nothing really custom apart from the ROM. One oddity is the power switch, which is both a slide switch (to completly disconnect the battery) and keys on the keyboard. Given that the RAM contents (data and config) are backed up by a lithium coin cell even when the slide switch is turned off, it is not clear why they did it this way. I suspect the HP units do rather more, but this does everything I need apart from not going down to 50 (or 45.45) baud. [...] > Just my 2 millidollars worth... Don't you mean 20 millidollars? -tony
Re: VAX 11/730 quickie
On Mon, Feb 22, 2016 at 10:14 AM, Mark J. Blairwrote: > >> On Feb 21, 2016, at 23:09, Mike Ross wrote: >> Actually I do have a Mac within easy range of the 730. Could you do me >> a favour and throw a prebuilt OSX binary somewhere I can grab it? I >> have flaky internet in the workshop and this Mac isn't set up with >> Xcode or any other dev environment... > > https://dl.dropboxusercontent.com/u/29876211/tu58em.zip Thanks Mark... I played with this. I built a 3-wire cable; I'm sure I have it wired right because traffic happens - and if I disconnect, or switch Rx/Tx, it just sits there but this is as far as it gets: MikeUpstairs:~ Imac$ ./tu58 -s 38400 -p /dev/tty.usbserial -d -v -m -x -r vax73058 info: unit 0 rfile 'vax73058' info: tu58 tape emulator v1.4j (NF6X fork) info: (C) 2005-2014 Don North , (C) 1984 Dan Ts'o info: serial port /dev/tty.usbserial at 38400 baud info: MRSP mode enabled (NOT fully tested - use with caution) info: TU58 emulation start info: R restart, S toggle send init, V toggle verbose, D toggle debug, Q quit info: emulator started (I ctrl-c the VAX) info: flag=0x00 last=0x00 info: seen info: flag=0x00 last=0x00 info: seen info: flag=0x04 last=0x00 info: seen info: flag=0x04 last=0x04 info: seen info: seen, sending (?27 DEVICE ERROR on VAX) (I ctrl-c the VAX again) info: flag=0x00 last=0xFF info: seen info: flag=0x00 last=0x00 info: seen info: flag=0x04 last=0x00 info: seen info: flag=0x04 last=0x04 info: seen info: seen, sending (?27 DEVICE ERROR on VAX) (repeat) Can you spot me being stupid? Does my command line look sane re. various options? (I've trivially renamed the tape image file for faster typing otherwise it's straight out of the box) Mike http://www.corestore.org '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.'
Re: VAX 11/730 quickie
On Tue, Feb 23, 2016 at 3:35 PM, Mark J. Blairwrote: > >> On Feb 22, 2016, at 16:12 , Mike Ross wrote: >> >> Mark am I missing something or is there no make option for Linux in your >> tweaked tu58em? > > I think you simply type "make" to build it on a unix-like system (e.g., > Linux). ...which blows up immediately with large numbers of errors and warnings when it gets to serial.c - looks like it's hardcoded for Windows header files. No worries I'll puzzle it out; the main branch tu58em builds cleanly with no issues on Linux. Mike http://www.corestore.org '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.'
Re: VAX 11/730 quickie
> On Feb 22, 2016, at 16:12 , Mike Rosswrote: > > Mark am I missing something or is there no make option for Linux in your > tweaked tu58em? I think you simply type "make" to build it on a unix-like system (e.g., Linux). -- Mark J. Blair, NF6X http://www.nf6x.net/
Re: VAX 11/730 quickie
Mark am I missing something or is there no make option for Linux in your tweaked tu58em? Mike On Feb 23, 2016 4:28 AM, "Mark J. Blair"wrote: > > > On Feb 22, 2016, at 02:50, tony duell wrote: > > > > The pinouts are the same, the printset of course gives details of the > latter. It's > > RS232 levels, TxD, RxD, Ground, and it is 38400 baud. > > I didn't get around to examining the wiring in my VAX last night, but I > determined the wiring from the TU58 manual, anyway. > > When I was debugging the connection between my VAX and tu58em on my Mac, I > ended up buying an old serial protocol analyzer. Notably, I specifically > avoided one with the same type of tape drive; I got one with a nice, > reliable 3.5" floppy drive! ;) > > -- > Mark J. Blair, NF6X > http://www.nf6x.net/ > >
RE: Serial analyzers (was Re: VAX 11/730 quickie)
I've always eyed the 4952... being as my penchant is HP and much of my test equipment is same-vintage HP gear (I'll give a vote for the 1631D logic analyzer, combo LA and digital scope - the scope is sorta poor, but handy - the LA is great for what I work on). However, I spent a significant portion of my career in front of a serial analyzer so I'm pretty familiar with them. As much as I really WANT the 4952 to be "it", it isn't - for me. Mainly because I have found that having a unfolding or removable (or permanently in the usual spot) keyboard takes up far too much bench space - or is unusable when standing the unit upright (very often it was convenient to set the LA on the floor and on the models I use - the keyboard and screen face straight up). The ones I use these days are a Spectron D101 or D101X (I have them both). See http://ebay.com/itm/262261970901 The one I used for work projects - don't own - and may buy soon... is a Spectron D2000 which for some reason is labeled in this auction as a "Telenex AR Datascope 2000". See http://ebay.com/itm/301859511695 Since either of these have a front panel that is "on the front" and doesn't stick out at all... they seem to work better real-estate-wise sitting on the bench facing towards me, or sitting on the ground on the end - screen/keyboard facing up - so they are more accessible when working on a system. Just my 2 millidollars worth... J
Re: Serial analyzers (was Re: VAX 11/730 quickie)
On Mon, Feb 22, 2016 at 10:54 AM, Mark J. Blairwrote: > The bigger picture: > > https://twitter.com/nf6x/status/701796809413304320 Nice! We had a couple 4951s set up at all times, mostly for sync debugging of our own products, but occasionally, we stuck them between two machines running Kermit or Blast or some other async transfer program to figure out why they weren't happy. We had one RS-232 and one combo RS-232/differential (2 connectors) pod. I see a few cheap machines on eBay, but they are missing the pod cover. Good for parts, but without finding/fabricating an external interface, not immediately useful. The complete machines seem to still fetch a pretty penny. Great series of machines. -ethan
Re: Serial analyzers (was Re: VAX 11/730 quickie)
The bigger picture: https://twitter.com/nf6x/status/701796809413304320 -- Mark J. Blair, NF6Xhttp://www.nf6x.net/
Re: Serial analyzers (was Re: VAX 11/730 quickie)
> On Feb 22, 2016, at 07:35, Ethan Dickswrote: > > On Mon, Feb 22, 2016 at 10:28 AM, Mark J. Blair wrote: >> When I was debugging the connection between my VAX and tu58em on my Mac, I >> ended up buying an old serial protocol analyzer. Notably, I specifically >> avoided one with the same type of tape drive; I got one with a nice, >> reliable 3.5" floppy drive! ;) > > HP 4952? Good guess! https://twitter.com/nf6x/status/701796162223779840 -- Mark J. Blair, NF6X http://www.nf6x.net/
Re: VAX 11/730 quickie
> On Feb 22, 2016, at 02:50, tony duellwrote: > > The pinouts are the same, the printset of course gives details of the latter. > It's > RS232 levels, TxD, RxD, Ground, and it is 38400 baud. I didn't get around to examining the wiring in my VAX last night, but I determined the wiring from the TU58 manual, anyway. When I was debugging the connection between my VAX and tu58em on my Mac, I ended up buying an old serial protocol analyzer. Notably, I specifically avoided one with the same type of tape drive; I got one with a nice, reliable 3.5" floppy drive! ;) -- Mark J. Blair, NF6X http://www.nf6x.net/
Serial analyzers (was Re: VAX 11/730 quickie)
On Mon, Feb 22, 2016 at 10:28 AM, Mark J. Blairwrote: > When I was debugging the connection between my VAX and tu58em on my Mac, I > ended up buying an old serial protocol analyzer. Notably, I specifically > avoided one with the same type of tape drive; I got one with a nice, reliable > 3.5" floppy drive! ;) HP 4952? I have two with tapes from the old days (they contain a formatter, and we did wipe some scratch tapes for storage of our analyzer programs, including a PU Type 2 BIND simulator that filled the program memory, but it worked well enough for us to debug SNA BIND sequences in the absence of a real PU Type 4). -ethan
Re: VAX 11/730 quickie
> On Feb 21, 2016, at 23:09, Mike Rosswrote: > Actually I do have a Mac within easy range of the 730. Could you do me > a favour and throw a prebuilt OSX binary somewhere I can grab it? I > have flaky internet in the workshop and this Mac isn't set up with > Xcode or any other dev environment... https://dl.dropboxusercontent.com/u/29876211/tu58em.zip -- Mark J. Blair, NF6X http://www.nf6x.net/
RE: VAX 11/730 quickie
> > On Feb 21, 2016, at 14:41 , Mike Rosswrote: > > Thanks for that. Are you able to provide confirmed working details & > > pinouts for the cable? IIRC it was just three wire; Rx/Tx/Gnd? Would > > help if I could have confirmed working setup there. > > I'll need to dig inside my machine to verify the wiring; I'll try to remember > to > do that sometime today. I'm pretty sure that it's a simple 3-wire connection > at 38400 baud. It is. There are 2 positions you can connect the TU58 to the 11/730 CPU, both 10 pin headers. One is on the backplane, the other is on the WCS board. The latter is used in whichever version it is that has the CPU in a more conventional box (boards standing vertically) with the TU58 mounted in the rack cabinet. The pinouts are the same, the printset of course gives details of the latter. It's RS232 levels, TxD, RxD, Ground, and it is 38400 baud. -tony
Re: VAX 11/730 quickie
(replies inline) On Sun, Feb 21, 2016 at 9:27 AM, tony duellwrote: > >> Well I haven't figured out exactly what the problem was but I'm >> embarrassed to report it was indeed serial comms finger trouble. I > > I have found that those little in-line RS232 testers with 7 or so bicolour > LEDs monitoring the important signals are very useful when working on > a machine with a serial terminal. If you get flickering on the TxD or RxD > LEDs then it is sending something. Yes and I usually use such; I have an assortment of things from inline testers to breakout boxes to Tektronic analysers. But this hookup was so simple and obvious nothing could go wrong (!) >> could have sworn that VT220 was fine and the cable wired correctly... >> but to cover all bases I tried it with a USB serial port on the Mac >> that sits in the lab. >> >> It worked. >> >> First time. >> >> CONV011 > > There should be a pause between the 'CONV0' and the '11' IIRC Indeed that's how it comes up. Very short pause since I have most of the memory out as part of potential gotcha elimination - and the last two tests are memory tests IIRC; they take a little longer. > You should then get 2 lines of read errors for DD1 and DD0 in that > order. If you get a 'Device Error' then the TU58 controller is not > responding at all, either it is not plugged in or it has serious problems. Yep that was shown in the link I gave. In text it's: CONV011 ?27 READ ERROR DD1 ?27 READ ERROR DD0 ROM>^C ?27 READ ERROR DD1 ?27 READ ERROR DD0 ROM> > I have 2 genuine 11/730 console tapes. Not that it does me any good, they > both have dropouts and are shedding oxide. The output of the read amplifier > in the TU58 is 'interesting' shall we say. > > I am currently rebulding a standalone TU58. My aim is to somehow find a good > tape (the hard part) and then to dump an 11/730 console tape image to it. That > means writing some kind of program to talk to the TU58 (another hard part). I did wonder how one might go about writing a tape image to a physical tape... I'm happy to get the thing loaded via the emulator for now - but I'd like to get real working tapes eventually. If it helps your 'hard part' I have at least a couple of brand new TU58 tapes - still sealed in plastic wrap... Mike http://www.corestore.org '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.'
Re: VAX 11/730 quickie
On Sun, Feb 21, 2016 at 5:33 PM, Mark J. Blairwrote: > >> On Feb 21, 2016, at 02:57 , Mike Ross wrote: >> >> Looking back at a long >> thread on vintage-computers last year it appears that tu58em had >> timing issues and was unusable on 11/730 but is now patched and >> working correctly... watch this space! > > As a recap, I made a fork of tu58em which adds -x/--vax flags to work around > a timing issue that I encountered when using it with my VAX-11/730: > > https://github.com/NF6X/tu58em/tree/nf6x > > I was running tu58em on a Mac. I don't know if the same timing issue is > present when running tu58em on other platforms. If you want to try my > modified fork, make sure you pull code from the nf6x branch instead of the > master branch; the master branch contains the unmodified tu58em code that I > patched. Mark, Actually I do have a Mac within easy range of the 730. Could you do me a favour and throw a prebuilt OSX binary somewhere I can grab it? I have flaky internet in the workshop and this Mac isn't set up with Xcode or any other dev environment... Thanks Mike http://www.corestore.org '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.'
Re: VAX 11/730 quickie
On 2/20/2016 7:03 PM, Mike Ross wrote: I might just try all switches *closed* on the basis that maybe it was wired wrong but... no doesn't make sense; the system would have been operational when decommissioned; the switch settings as I received it must be valid... Make sure the switches are actually working. Often, they're not. - J.
Re: VAX 11/730 quickie
On Sun, Feb 21, 2016 at 8:23 AM, tony duellwrote: > > [...] > >> red lights. The baud rate switches are set for 4800 - the default. > > Eh? There is no way to set it for 4800 baud (the installation manual, > printset, and my tests all agree on what baud rates are available). > > How do you have the switches set? I just double checked. 3 was open; 4 was closed. That's default 2400. Switch 8 was also closed; I have no idea of its function. All other switches were open. I set ALL switches to open - including 3 & 8. 3 & 4 open gives 9600 baud. I have no idea what setting 8 open does - but it didn't change behaviour. Still nothing on console. G. I might just try all switches *closed* on the basis that maybe it was wired wrong but... no doesn't make sense; the system would have been operational when decommissioned; the switch settings as I received it must be valid... Mike http://www.corestore.org '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.'
RE: VAX 11/730 quickie
[...] > red lights. The baud rate switches are set for 4800 - the default. Eh? There is no way to set it for 4800 baud (the installation manual, printset, and my tests all agree on what baud rates are available). How do you have the switches set? Note that if switch 2 (of the DIP switch at location E47 on the WCS board which is the one that controls the console baud rate) is closed then the 8085 consol processor is held in a wait state. If you have done this then of course you will get nothing on the terminal. -tony
RE: VAX 11/730 quickie
> > Of course I don't know mine works, I need to get a working TU58 tape... Perhaps I should have been a bit clearer, it does pass the console POST, then gives read errors on both drives (as there is no tape in them), then gives the ROM> prompt. So mine is getting further than yours, but until I have a working microcode tape, I don't know how much of the rest of the machine is good. > That's good to know on the LEDs. Yes I checked the PSU first; three > red lights. The baud rate switches are set for 4800 - the default. OK... Sounds like your PSU is OK. But worth checking the voltages just in case! > think I have the console cable correct - and just in case I buggered > the Rx/Tx I tried it with them swapped. Nothing. Never so much as a When I was routing the cables in my machine (as I was rack-mounting it, I had to put the R80, RL02, console, remote diagnostics, TS05 and DMF32 cables all in at once), I actually buzzed out a few non-ground connections between connectors on the distibution panel and IC pins on the right board (from the printset) just to be sure I hadn't got them mixed up or turned over or anything. It might be worth just checking that the RS232 chips on the WCS board do connect to the pins on the terminal you think they go to. I don't want to 'teach you to suck eggs' (especially since I think you've been doing this longer than I have) but bitter experience has taught me that checking 'the obvious' as you go along means you only have real faults to trace. > Anything you can tell me about the POST on your machine - the timing > - what appears on the console and when etc. - would be helpful. I will turn mine on later to check. IIRC it prints the first 4 or so characters of the header line, then pauses (RAM check I think), then the rest of the line, then the device error messages with a short pause between them, then the ROM> prompt (from which you can only enter ctrl-C to try to re-read the tape). One of the manuals on bitsavers (I think it's the diagnostic manual) explains this, and mine seems to do what the manual says. Have you checked signals round the 8085 yet? Is it being held reset? Is it clocking, accessing ROM, etc? > I plan to use an Arduino-based TU58 emulator when the time comes = > although I would like real working TU58s too! I am more interested in having an 'original' machine than a machine to run VMS on :-). What I mean by that is that I do want to have the TU58 working, rather than some replacement for it. I have replaced the drive pucks, in fact I've just been turning another pair for my standalone TU58 (which I will use to write a console tape if I can ever find a good cartridge). The console tape I have has dropouts, is shedding oxide, etc. The read waveform looks horrible. I have socketed the 8155 I/O chip in both TU58s (11/730 CPU and standalone). If I pull it, I can ground pins on the socket to select a drive, start the motor, control speed (normal or seeking), direction etc. I can then use a 'scope on the output of a read amplfier or a logic analyser around the data separator to see what is going on. Just got to keep an eye on how far the tape has gone so I can reverse it before it runs off the end. If I was going to use a solid-state replacement for the tape I would probably use the original controller board, put the 8155 on a daughterboard plugged into the 8155 socket on the controller, use the port pins of the 8155 to link to some kind of flash memory device and re-write the 8085 firmware. At least then it would be partly the original. Using a processor with more components than the rest of the machine (perhaps I exagerate...) seems a bit silly. -tony
Re: VAX 11/730 quickie
On Sun, Feb 21, 2016 at 1:56 AM, tony duellwrote: >> >> See row of red LEDs on M8391? On power-up the leftmost LED lights and >> goes out. Then the rightmost two LEDs light and stay on... > > I can't remember exactly what mine does, but I do remember 2 LEDs at one > end being on when it settles down. So I think yours is probably OK. > > Of course I don't know mine works, I need to get a working TU58 tape... That's good to know on the LEDs. Yes I checked the PSU first; three red lights. The baud rate switches are set for 4800 - the default. think I have the console cable correct - and just in case I buggered the Rx/Tx I tried it with them swapped. Nothing. Never so much as a garbage character on the terminal - and I know the terminal works ok! Nothing funky in the settings there as far as I can see. Anything you can tell me about the POST on your machine - the timing - what appears on the console and when etc. - would be helpful. I plan to use an Arduino-based TU58 emulator when the time comes = although I would like real working TU58s too! Mike http://www.corestore.org '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.'
RE: VAX 11/730 quickie
> > See row of red LEDs on M8391? On power-up the leftmost LED lights and > goes out. Then the rightmost two LEDs light and stay on... I can't remember exactly what mine does, but I do remember 2 LEDs at one end being on when it settles down. So I think yours is probably OK. Of course I don't know mine works, I need to get a working TU58 tape... -tony
RE: VAX 11/730 quickie
> Just started working on mine - been a back-burnered project for a long > time. Unfortunately all the cables were cut when it was dismantled; I > was lucky to grab just the CPU. I had to re-fit all the cables on mine (a painful job), but at least I had them all, uncut. > Got it powered up ok - no drama there. I faked cables to the port on Are all 3 LEDs on the PSU distribution board (under the transparent cover at the top rear of the PSU) glowing? If not, you have a PSU problem. > the WCS which drives the console and hooked up my trust VT220 at 2400 > baud. Nada. No self-test prompt; no ROM> prompt. The oriignal console cable was a straight through ribbon cable with a header socket (to fit the WCS board) at one end and a DB25 plug at the other. It plugs into a filter module on the distibution panel, again straight through with a DB25 plug on the outside. This is wired as a DTE (terminal), you need a null modem cable. Only pins 1,2,3,7 are used, the official cable (I had to make a copy of it) wires 1 and 7 straght through, swaps 2 and 3 and swaps 6 and 20, but you don't need the last pair. Have you set the DIP switch on the WCS board correctly? You can certainly have 300, 1200,2400 or 9600 baud. IIRC some of the other switches do very odd things, so make sure they are set as the manual says. I assume you have checked you are plugged into the console port and not the remote diagnostic port. > Maybe I've messed up the faked console cable; I'll check - already > tried obvious things like making sure Rx & Tx were crossed (it's a > three wire cable according to the schematics - Rx Tx Gnd; no flow > control). But I'd like to know more about the assorted LEDs on the CPU > boards; maybe there's a clue there if it's not getting far enough into > the self-test to display console output. There's a fair few LEDs on > the M8391. But the doc I've looked through on Bitsavers doesn't seem > to document their meanings and interpretations. Can anyone help on > that point? They're shown on page 208 of the printset on bitsavers. They display the memory controller microcode address (nothing to do with the CPU microcode, the memory controller has its own ROM firmware). I don't think the memory controller is used by the console ROMs, so I doubt if this is the problem. Start by checking the power supply rails (if you haven't already done so), then see if the 8085 console processor is running, and if it tries to access the console serial chip. -tony
Re: VAX 11/730 quickie
On Sat, Feb 20, 2016 at 12:22 PM, Mike Rosswrote: > Just started working on mine - been a back-burnered project for a long > time. Unfortunately all the cables were cut when it was dismantled; I > was lucky to grab just the CPU. > > Got it powered up ok - no drama there. I faked cables to the port on > the WCS which drives the console and hooked up my trust VT220 at 2400 > baud. Nada. No self-test prompt; no ROM> prompt. > > Maybe I've messed up the faked console cable; I'll check - already > tried obvious things like making sure Rx & Tx were crossed (it's a > three wire cable according to the schematics - Rx Tx Gnd; no flow > control). But I'd like to know more about the assorted LEDs on the CPU > boards; maybe there's a clue there if it's not getting far enough into > the self-test to display console output. There's a fair few LEDs on > the M8391. But the doc I've looked through on Bitsavers doesn't seem > to document their meanings and interpretations. Can anyone help on > that point? > > Hopefully this link works: > https://www.facebook.com/photo.php?fbid=10208521465663997=867897c786 OK apparently link didn't work... here's the pic: http://www.corestore.org/730boardsLEDs.jpg See row of red LEDs on M8391? On power-up the leftmost LED lights and goes out. Then the rightmost two LEDs light and stay on... Mike http://www.corestore.org '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.'