On Sat, Dec 31, 2016 at 6:45 PM, Ian Finder <ian.finder at gmail.com> wrote: Well, considering the main bits I need are assorted screws and brackets (the correct UNC hardware is very difficult to find in the UK), including the hinge bolts for the logic chasiss, it's going to fall apart anyway. As I said, I may well scrap the outer casing metalwork unless I have a good reason not to. The boards, HDA and PSU are smaller than the whole thing _and will fit into odd corners_. And they are light enough to move around (unlike the RA80!). There are few parts (other than the screws and cables) that I would cannibalise for the R80. Yes, the HDA if it is good and the one in the R80 is dead. The only custom electronic parts are the ROMs on the microprocessor board and they are different (so no point in using those). Other electronic parts I would try to get new. I don't swap boards! -tony

On Sat, Jan 7, 2017 at 6:48 PM, Pontus Pihlgren <pontus at update.uu.se> wrote: Well, actually I think you own an R80 but don't know it :-) As I understand it, the R80 had the sort-of SMD interface. 2 cables (60 and 26 wires). It's not SMD, there are significant differences, but anyway. The RM80 is an R80 with a Massbus interface box in the stand. Or at least I think it is. I've not got any technical documentation on that unit. If you have one, can you tell me what the cabling between the drive and Massbus interface is? And if the 'personality board' (the smaller of the 2 hinged boards under the top cover) matches that in the R80 manuals. The RA80 has the SDI interface. The bare R80 (no external Massbus interface) was used with the IDC (Integrated Disk Controller) on the VAX11/730. That controller can link to up to 4 drives, one can be an R80, the others RL02s. I have the compact 11/730 configuration, a half-height rack containing the CPU, an R80 and a TS05 magtape (Cipher F880). There are basically the following modules in the drive : HDA : I am pretty sure the physical HDA is the same on all drives. The servo information should be the same too. Maybe the data surfaces are formatted differently. PSU : There are at least 2 versions of the PSU which are very different internally (so if you are working on one you need the RA82 printset from bitsavers as well!). But it's not drive-specific R/W PCB. Sits on top of the HDA and AFAIK is the same in all drives Microprocessor PCB. An 8085 and a lot of paralell I/O. It uses the 8155 RAM I/O chips and thr 8755 ROM I/O chips. And a lot of other logic too. The ROMs are different between the R80 and RA80, but the board is the same apart from that. Annoyingly the ROMs in the RA80 are in sockets, those in the R80 are not.... Servo PCB. A lot of analogue and digital circuitry to drive the positioner. AFAIK t's the same in all drives Personality PCB. The host interface. The RA80 and R80 ones are totally different for obvious reasons. Control panel. A few illuminated pushbutton switches. The R80 one is quite simple, the only electtonics on it are the lamp drivers. The RA80 one has the later drive select swtiching (wider ready lamp with a cap that you break pegs off to set the drive number) and also has the drive serial number, revision level, etc set by DIP switches and DIP shunts. It still plugs into the same connector on the microprocessor PCB and thus the RA80 control panel is stuffed with multiplexers. Getting back to my dilemma. I need an R80. I do not need an RA80. I may try to run the internal diagnostics on the RA80. Check the PSU first of course. I almost certainly will take the gas strut brackets, etc from the RA80. Making them is a right pain!. I would consider trying to source the screws, etc, and make the missing cables for the R80, but only if somebody really wants an RA80 intact. I don't want to keep the thing in my machine room for no purpose. Rest assured that any 'useful' bits will be saved no matter what. I will certanly keep all PCBs, HDA, PSU, motor, etc. It would certainly need the RA80 ROMs and contol panel. I think then you could get a working drive, but it may well not be able to read the format of an RM80 HDA. You could probably reformat the HDA to work, but that rather defeats the purpose of what you are doing. -tony

On Sat, Jan 7, 2017 at 8:08 PM, Pontus Pihlgren <pontus at update.uu.se> wrote: I've just re-read the printsets... The microprocesor boards in the R80 and RA80 printsets are different. They have different layouts. But the circuitry looks at first glance to be much the same. Whether this is just a revision level thing I don't know. I will look over it more carefully. At first glace the external connectors seem to have the same pinouts. Interestingly the personallity board in the R80 printset is called the RM80 personality board. Suggesting that as I thought the RM80 is the R80 with a Massbus interface box. The microprocessor board includes the write data encoder and the read data separator. There is a clock/data pair for each to the personality board. The R80 (and thus RM80) personality board just buffers them and sends them to the 26 way host cable. The low level format is thus determined by the controller. I have no reason to assume that an RM80's controller uses the same format as the VAX11/730 IDC. The RA80 personality board has some circuitry between the clock/data lines to the microprocessor board (and thus eventually to the heads) and the host cables. But not that much circuitry, I've not looked at it seriously yet (after all it's the one part I will not be using). Certainly not enough to have a sector data buffer. So my guess is that againg the host does a lot of the low-ish level stuff. -tony

On Sat, Jan 7, 2017 at 8:22 PM, Tony Duell <ard.p850ug1 at gmail.com> wrote: However, today I came across the microprocessor board from my R80 (and which will be going back in the R80). It seems to have the same layout as the board in the RA80 printset (rather than the one in the R80 printset). I've not checked the schematics, etc, but I suspect my original thoughts are right and that it's the same circuitry with different ROMs (and the layouts in the 2 printsets are just 2 different revisions of the board). -tony

On Sun, Jan 8, 2017 at 6:20 PM, Robert Armstrong <bob at jfcl.com> wrote: It's somewhat SMD-like, but it's not SMD. Some of the signals have the same names (like the control signal bus from controller to drive being called 'Tag') but you can't plug an SMD drive into the 11/730 IDC. One oddiity is that the SMD drive's select lines are inputs to the drive (as you'd expect), on the R80 the lines are outputs from the drive giving the coding of the ready lamp cap (which sets the unit number like an RL02). I've not looked into making an interface between the 11/730 IDC and an SMD drive, I might have to if the R80 proves unrepairable. I've always felt that the RA80 is the 'oddest' member of the family. As a result I suspect an RM80 HDA will swap with the others and vice versa, but I've not tried it. The RM80 manuals aren't on Bitsavers. The RA81 and RA82 HDAs seem quite different. They have more heads for one thing, but I get the feeling that the servo system is quite different. One manual suggests not only do they have the dedicated servo surface, but also embedded servo information on the data surfaces. AFAIK the R80, etc only has the dedcated servo surface. The servo PCB seems to be the same in the R80 and RM80. That for the RA81 is a lot more complex with a microprocessor on it. I had wondered if you could kludge an RA81 or RA82 HDA into an R80, I think it would be very hard. Not only am I the wrong side of the Pond, but the HDA is one part I am keeping from the RA80, no matter what I do with the rest of it. I don't know if it's good, of course, but if it is it will be kept as a spare for the R80. The HDA is the only major unit I can't repair (yet!). -tony

On Sun, Jan 8, 2017 at 6:46 PM, Robert Armstrong <bob at jfcl.com> wrote: I don't know about the 11/730's IDC. I was just looking at the R80 schematics, the unit number lines are output on the 60 wire cable. Maybe the RM80's massbus box uses them or something? A quick look at the printset and technical manual suggests that the 11/730 IDC compares the drive select number from the appropriate output register with the unit number from the R80 (set by the ready lamp cap). If they match it uses the R80 logic, if not then it uses the RL02 logic. I have no idea if the nomal software will work if the R80 s anything but unit 0, but it looks like the hardware was designed to allow for it. The only (other) SDI drive I have is an RA60, and I have no packs for it. I doubt I will ever use it. Apart from the ROMs (and those 8355s/8755s are not suported by most programmers/readers now available), there are no custom parts on the PCBs that I've spotted. There are custom chips on the PCB on the front of the HDA and unlike one Micropolis drive I work on, you can't replace them without dismantling the HDA (in a clean room). There are also plug and socket connectors on that PCB -- inside the HDA. I hope those don't suffer from bad contacts! Quite likely. Which means we are less kind to them than we should be in that we don't keep them running (or at least I won't) -tony

On Sun, Jan 8, 2017 at 8:57 PM, allison <ajp166 at verizon.net> wrote: Is it therefore likely that if an HDA has lasted this long it is likely to have been made with the 'good' adhesive? I have an RA60 (but no media for it). What was the problem with that one? At least you can replace heads in the field (although getting new heads is going to be 'fun'). Problem for me is that I have the compact 11/730 system. The CPU box is pretty much full -- CPU, IDC (I don't have the floating point board, but it's a dedicated slot anyway so I can't put something else there), 4M RAM, TSU05 controller. I don't have space for another disk controller (certainly not a UDA50). I really want to keep the machine in the short rack too. Which means using an R80 or something that emulates one. I wonder how hard it would be to interface some other storage device to the IDC. -tony