From rwallace@world.std.com Sat Sep 2 14:00:55 PDT 1995 Article: 4900 of vmsnet.pdp-11 Path: nntp-server.caltech.edu!news.cerf.net!ihnp4.ucsd.edu!mvb.saic.com!info-pdp11 From: Roger N Wallace Newsgroups: vmsnet.pdp-11 Subject: Re: Q: BDV11 Rev. "A" vs "E" Message-ID: Date: Sat, 2 Sep 1995 16:08:42 +0059 (EDT) Organization: Info-Pdp11<==>Vmsnet.Pdp-11 Gateway X-Gateway-Source-Info: Mailing List Lines: 97 On 31 Aug 1995, Leonhard Schneider wrote: > Hello, > > I have an OEM PDP-11/03 system with a BDV11 Rev. A bootstrap/terminator > board (M8012-YA). > > Several MicroNotes mentioned a > > "BDV11 CS Rev.E" with "ECO M8012-ML005 installed". > > Could anybody please explain me briefly > > - differences between Rev. "A" and "E", > - what "CS" stands for, > - what "ECO M8012-ML005" means, > Briefly, the fifth ECO against the BDV11 connects on-board terminations to the four highest address/data lines (BDAL18 through BDAL21) of the Q22 bus. This permits the board to terminate the bus in a 22-bit system correctly. A similar modification is possible on certain variants of the REV11 (boot/ refresh/terminator) module, although I don't know whether it was officially accepted like the BDV11 ECO. Background: The BDV11 was originally intended for use in 16 and 18 bit PDP11 systems. In the versions of the board usually supplied, with the exception noted below, the 18 low-order address/data lines, as well as the control and interrupt lines, were connected to on-board terminations. However, the terminator networks, integrated in several DIPs, contained some _unused_ terminations. By connecting wires from the unused terminations to the appropriate fingers on the card-edge connector, one can properly terminate all 22 address/data lines. I don't think that DEC ever supplied a 22-bit BDV11 as a "new" product. (The real DEC historians who frequent this group may correct me on this.) However, Customer Service _would_ perform the 18 to 22-bit modification of an existing BDV11 in a system that was being upgraded to a 22-bit processor. I believe that this is the reason for the "CS Rev. E" portion of the identification. The actual content of ECO #5 from the Field Maintenance Print Set is: 5-1 Add wire from finger BC1 to E1 pin 15 5-2 Add wire from finger BD1 to E1 pin 14 5-3 Add wire from finger BE1 to E1 pin 13 5-4 Add wire from finger BF1 to E1 pin 12 It is easy to inspect the board to determine whether this modification has been performed. If you are handy with a small soldering iron and wire-wrap wire, you can easily perform the modification yourself. I have done it on a half-dozen or so BDV11's. There is only one possible downside to the installation of this ECO. The Q-bus lines corresponding to BDAL18-BDAL21 are used as test points in 11/03 processors, and the CPU does not expect them to be "bussed" with a terminator at the far end. If such a processor is installed in a 22-bit backplane _and_ the far end of the bus is terminated with a BDV11 that has been modified by installation of ECO #5, the system may not operate properly. Q-22 backplanes recently manufactured often contain jumpers for "18-bit" or "22-bit" operation. When these jumpers are _removed_, BDAL18-BDAL22 are disconnected from the processor slot, so connections of any other boards to these lines do not affect the processor. If the backplane is a "16-bit" or "18-bit" type that does _not_ provide connections to BDAL18-BDAL21 at each card slot, installation of ECO #5 on the BDV11 will have no effect on system operation. An OEM 11/03 may use lines BDAL18-BDAL21 for special functions, or may contain 22-bit cards that require these lines to be terminated even though they are not connected to the CPU. This could be the reason for the modification of the BDV11 in your system. In the foregoing description, I noted that control lines and the 18 low-order address/data lines were terminated on the standard BDV11. One qualification to this statement arises because of an error in the BDV11 circuit board etch. The trace that was supposed to connect BIRQ6 (terminal AB1 of the card-edge connector) to a termination was placed on the wrong side of the board, and instead connected terminal AB2 (user option, -12 VDC) to the termination. This error was corrected by ECO #2, as follows: 2-1 Cut etch Side 2 between E1 pin 11 and finger AB2 2-2 Add wire Side 1 between E1 pin 11 and finger AB1 Note that standard DEC PDP11 systems do not use either BIRQ6 or the optional -12 V line. Use of a BDV11 that has not been modified in accordance with ECO #2 in an OEM system where the optional -12V line (AB2) is powered would probably "smoke" (burn out) the terminator pack. In a "former life", I had occasion to look through about 20 BDV11's, mostly from MINC systems that were being retired. Only one board did _not_ have ECO #2 installed. Roger Wallace From rwallace@world.std.com Tue Apr 18 22:02:44 PDT 1995 Article: 4181 of vmsnet.pdp-11 Path: nntp-server.caltech.edu!pagesat.net!decwrl!svc.portal.com!sdd.hp.com!ihnp4.ucsd.edu!mvb.saic.com!info-pdp11 From: Roger N Wallace Newsgroups: vmsnet.pdp-11 Subject: Re: XXDP Docs? Message-ID: Date: Tue, 18 Apr 1995 23:42:24 +0059 (EDT) Organization: Info-Pdp11<==>Vmsnet.Pdp-11 Gateway X-Gateway-Source-Info: Mailing List Lines: 59 On Tue, 18 Apr 1995 terry@spcvxa.spc.edu wrote: > > That's odd - they just took an order for one from me, $114, 7 days lead time. > The trick seems to be to *not* order it through the E-store, but from DECdir- > ect. > > Terry Kennedy Operations Manager, Academic Computing > terry@spcvxa.spc.edu St. Peter's College, Jersey City, NJ USA > +1 201 915 9381 (voice) +1 201 435-3662 (FAX) > If you are fortunate enough to get the KDF11-B3 upgrade kit, could you let the group know the extent of the XXDP documentation you receive? Last one I purchased (about two years ago) came with a fairly extensive manual explaining how to install the new ROMs, but the XXDP documentation was limited to a single sheet with release number and change information. If you dig through the floppies themselves, there are some "help" and information files here and there, but no complete documentation. Some of the diagnostics themselves have embedded "help", but this is typically of a sort that gives new meaning to the term "minimalist". Incidentally, the ROMS in the KDF11-B3 kit _are_ useful if you have an 11/23 system. The boot code is "BDV11 compatible", meaning that it supports the same windowing (two windows starting at 173000) and page-control scheme (PCR at 177520) as used in the BDV11. The BDV11 switch packs are a superset of the boot configuration switches available on a KDF11B CPU board. The ROMs themselves can not be used in a "stock" BDV11, as they are 8kx8 and the BDV11 supports only 1Kx8 and 2Kx8 devices. However, if you copy the boot code from the two 8Kx8 ROMs into 8 2Kx8 ROMs, install the 8 ROMs and "jumper" the BDV11 appropriately, it works fine -- supporting MSCP "sniffer" boots, as well as TK50 and ETHERNET boots, and also providing some menu-driven diagnostics -- none of which were present in the original BDV code. Alternatively, but requiring more surgery, you can pick a low-byte/high-byte socket pair on the BDV and run the two additional on-board address lines necessary to support the 8Kx8 ROMs "as is". (Having the maintenance print set helps for this exercise!) After either "conversion", the BDV11 switches have to be interpreted according to the KDF11B manual -- their function is re-defined by the new boot code. If you don't mind spending the quad slot, the BDV11 can be used in 22-bit systems. The four additional terminators needed for this are actually present in the existing DIP resistor networks -- but not connected to BDAL18 through BDAL21. Solder in some wire-wrap leads to complete the connections and you're in business. An ECO, ca. 1978, describes this procedure, but I'm not sure is was ever released to the field. DEC folklore trivia: I have a BDV11 on which, in the center of an open "field" area far from any adjacent "wiring" on the etch side of the board, a small "spot" of interconnect metal remains. On close inspection, this spot is, unmistakably, the outline of a roadrunner -- the real thing, not the cartoon character. Board serial number is AG933FMY50, which I understand does identify the board shop which made it, at least to someone who knows the "code". There's got to be a story, or at least a "footnote to history" here. Does anyone in the assembled group of PDP11 historians know how the roadrunner got past layout Q/A inspection? Roger Wallace