MISC Messages from TPL about SASI Project: f: STEVE MCCRYSTAL t: SYSOP (Rcvd .... SASI bus but making sure not to lock up the CPU (waiting for non-existant data transfers). The end result: you may want to try changing the value of the 10K resistor used on the one shot to something like 100K (60 us) or 1 meg (600 us) [I believe the RC is linear in this case]. The problem where you must send out 7 bytes instead of six sounds like a timeout on the one shot. . A third item, as I mentioned in SDUTIL.DOC, the 220/330 ohm termination is more that absolutely required. Without the termination resistors, the REQ and ACK lines will slowly deassert themselves due to the open collector drivers (7438) which would also present the same type symptoms (7 bytes vs 6 bytes). Also, 74LS38 due not output as much current as the 7438 which is required to drive 220/330 ohm terminators. So I hope you've got 7438 and not 74LS38s on your board. . A third item, check the GND connection between the SASI controller and the JRSASI interface board. You can do this by using a volt meter and checking that the +5 reading is same using a GND on the JRSASI board and a GND on the SASI controller (still measuring the +5 level on the JRSASI board). [that may have not been too clear - my strong point has never been grammar]. . And as a last item, check that the voltage levels are as close as can Thank you Mike for the tips. I've come a little ways since my last request for help. The SASI board seams to be working now. I was trouble shooting the REQ/ACK circuits using a logic/pulse type probe and found while running an automated version of the DEBUG routine (written in BASIC) that the second byte of issue drive characteristics and the third byte of the eight bytes going to the controller resulted in no REQ response on the input side of the 74LS04 but when I placed the probe on the output side of the 74LS04 each byte gave a REQ pulse and the whole DEBUG routine worked fine. That is strange!! This may sound strange but then a I place a .1 micro farad cap between pin 12 and ground and got the same positive results. I was very happy to see this and only hope I haven't addressed a symptom. I proceeded to set up the disk using SDUTIL. Everything worked great. There weren't even any errors in the multipule writes and reads. For the 20 meg Seagate ST225 I issued the values: DOS INITIALIZATION- 1 (0), 2 (615), 3 (4), 4 (400), 5 (300), 6 (8) this is the Shugart 1610-3f, 7 (0), 8 (17), 9 (512), 10 (4), 11 (2), 12 (1), 13 (256), 14 (0), 15 (41820), 16 (41). I set up DEVICE=SASI.COM and BUFFERS=24 and FILES=24 and restarted the harddisk and rebooted the system. I was able to access drive C:. My new problem is the FAT is being altered. I'm getting cluster errors, truncated files and a whole bunch of ugly things that CHKDSK tells me. Is it true that when addressing more than 4080 clusters we need a 16-bit FAT. #25 91 05 Dec 86 07:24:16 From: Gary Myers To: Sysop SEE ALSO #26 Subj: SASI Controller Problem I have recently completed the Jr SASI Project and am now having a problem I need some advice on. All parts of the Jr Sasi board ohm out ok and the format and write sector portions od SDUTIL complete without error, however the thing times out on a read multiple sector command (error 84). I have double checked everything many times. In the debug section all commands complete ok except for the final check listed as "Command-Data Exchange. When I enter I3FFD I get "0B" as a read from the SASI data port. Any hints??? Also, on page 2 of the schematics, the common lines from the 7438s, they are listed as going to (SIO), but not connected. Is this a reference point? Any comments will be appreciated. Thanks, Gary Myers (713)643-6524 #26 95 06 Dec 86 02:37:22 From: Jim Arroyo To: Gary Myers REPLY TO #25 Subj: solution? Hi Gary, This is Jim Arroyo from Los Angeles, Calif. Try placing a .1 mf CAP from pin 12 of U6 (7404) and ground. This fixed a similar problem for me when I was building my hard disk interface. I hope it works out. Good Luck! Jim . Are you aware that we have an assembled and tested version of the jrSASI interface. The assembered version is mounted in a sidecar case and comes complete with the 50-pin cable required to connect the interface to a SASI hard disk controller and the driver and setup software required. Besides the assembed jrSASI interface you still need a SASI hard disk controller, a hard disk and a suitable power supply (at least 1 amp of +5 and 3-4 amps of +12). More? . The advantage of the assembled and tested interface is obviously, that you don't have to build it and risk not having it work. The cost of the assembled version is $169.95. The protoboard is $39.95. You will have to decide whether you think you are skilled enough to build the interface yourself and save or if you should spend a bit more to get the pre-built and tested unit. . I hope this answered your questions. Feel free to ask any others you may have. . Take care... Jon Kosmoski V.P. Engineering, TPL Corp #47 49 16 Mar 87 22:17:19 From: Chuck Holst To: All SEE ALSO #53 Subj: Shugart 1610-3 Controller I just acquired a Shugart 1610-3 hard drive controller to use with the TPL PCjr SASI interface. The specification that came with it claims it is compatible with the Xebec S1410A controller, which other sources say is a SASI controller. However, the specification also keeps referencing the SCSI (ANSI X3T9.2) Specification. Specifically, it says the 1610-3 "is designed as a peripheral controller to be used with the low end of the SCSI industry standard interface as defined by the SCSI specification no. (ANSI X3T9.2). Nowhere does it state explicitly that the 1610-3 is a SASI board. So, is it a SASI controller or a SCSI controller? If I find out soon enough, I can return it and get a used Xebec S1410 instead. Otherwise, I'm stuck with it. Any experienced help would be appreciated. --Chuck #50 18 30 Mar 87 19:19:48 (RECV'D) From: Chuck Holst To: Jon Kosmoski REPLY TO #34 Subj: ST225N As an addendum to your reply about the ST225N, I would like to say that the only references to the controller I have seen in Computer Shopper ads say that it is SCSI, not SASI, and therefore will not work with the TPL interface. To: Paul Baxter REPLY TO #40 SEE ALSO #52 Subj: Mail-order ST225 In answer to one of your questions, I bought a ST225 from the cheapest outfit advertising in Computer Shopper, N.D.R.C in Dallas. It took nearly two weeks to arrive (about 4 - 5 days longer than it should have), had no documentation, and was packaged in an undersized main board box with minimal foam rubber cushioning. Considering how packages get tossed onto the sorting belts at UPS, I think this is inadequate packing for such a delicate piece of gear. I had expected lots of plastic peanuts. On the other hand, the total bill came to under $300, not counting phone calls. It also arrived in a sealed pink poly bag with a test printout indicating it is a factory new drive. And the people I talked to at N.D.R.C. were polite, if not efficient. I only hope it still works when I get the project all assembled. I also located a source of used Xebec S1410 controllers at $95 each and supposedly new Shugart 1610-3 controllers at $79 each. The latter is supposed to emulate the Xebec S1410, which would make it a SASI controller, but the hardware spec (with unreadable schematics, $8 extra) refers only to the SCSI spec, albeit the "low end" of the SCSI spec. I think it is a SASI controller, but I'm not absolutely certain. (Can anyone enlighten me?) Both controllers are available from Computer Surplus Store, 226 Phelan Ave., SanJose, CA 95112. Delivery was prompt on the 1610-3, and it came with a 90-day warranty. #53 18 01 Apr 87 00:40:31 (RECV'D) From: Jon Kosmoski To: Chuck Holst REPLY TO #47 Subj: Shugart 1610-3 Hi Chuck Relax, your 1610-3 is perfect for jrSASI. In fact, that's the exact controller that TPL sells for use with the jrSASI interface board. As far as whether the 1610 is a SASI or a SCSI, I'm not 100% sure myself. As far as I know, SCSI is a superset of SASI. I would guess that the 1610 is a SCSI-compatible controller and as such also handles the SASI subset of SCSI just fine. Hope this helps... Jon Kosmoski V.P. Engineering, TPL Corp I get the impression that the people at TPL are so overwhelmed with a project that pays better than trickle-in orders from PCjr owners that they haven't had time to take care of this end of the business. I expect that when things ease up, they will get back to servicing our questions and orders. I have to admit, though, that I am getting impatient to get the proto board I ordered, so I can get started. I already have everything else I need except the power suppy and a few cables. I have been doing research to answer my own questions, and am about 90% certain that the Shugart 1610-3 will work. The best introductory article on SASI/SCSI I've run across is a two part article by Steve Ciarcia in the May and June 1986 issues of BYTE. Ostensibly about adding a SCSI interface to the SB180 computer, it is really a tutorial about what SCSI is and how it works. SASI turns out to be a SCSI that is intended to work with only one host and one peripheral controller; therefore, certain features of SCSI that allow multiple hosts to talk to multiple controllers on the same bus are not implemented. Now that I know this, I can tell from reading the hardware spec that the 1610-3 is intended for SASI applications. It appears that the difference between SASI and SCSI on this particular board is merely a matter of software implemention, or perhaps I should say "firmware", since it looks like all I would have to do to change it to a SCSI board is to change one PROM. Ciarcia's interface is different from TPL's. Most of the work is done by one chip, a NCR 5380 SCSI bus-interface chip. Although Ciarcia is using it as a SASI interface, he says it will work as a full SCSI interface if he only changes the software. Perhaps someone out there with more design skills than I have will take this circuit and adapt it for the PCjr -- hint, hint. --Chuck the drive as an external (brand varies, in my case it's a Tandon) with a cable ending in a jr sidecar. It's formatted and fully configured; I just plugged it in and away I went. You boot off the floppy, but a boot PROM chip is now available for the sidecar controller! Comes with a warranty and a pledge that if you ever sell your jr, you can trade in the sidecar for a standard PC hard disk controller. All of this is not cheap by PC standards -- these are going for about $600. On the other hand, it works as advertised, right out of the box, doesn't draw on the jr's power system and gives you a few options. Drawback, if you can call it that: The disk is formatted under DOS 2.1, meaning it has 8k clusters -- wasteful if you, like me, have lots of small text and batch files. (So reformat with 3.x, or PKARC everything on the hard disk -- but take your time, you've got 22mb). The jr really flies, now. Am thinking of getting Software Carousel so I can use part of hard disk for program swapping. This setup is one of several that jr owners appear to have settled on, others being the Racore top-mount unit and the third a semi-do-it-yourself job that goes for about $400 plus some sweat equity. #225 38 17 Feb 87 15:38:21 From: Jon Kosmoski To: John Wager Subj: PCjr to PC Bus Interface Hi John Sorry to take so long to respond. I'm nearly as busy as Mike, who was indeed considering such a project. He did actually build one for his own personal use and I know he was running at least one video card designed or the AT on his PCjr. Perhaps he would be willing to give or sell the design details to you. I will ask him about it and for him to leave you a reply on this BBS. Jon #228 39 17 Feb 87 16:15:23 From: Jon Kosmoski To: John Wager Subj: ST225N Hi John We have not tried it ourselves, but as you said, if it is a "standard" SASI controller, it should work fine. I thought I heard at one time that the 225N uses an enhanced interface, SCSI, instead of SASI, but I could be wrong. If it's SASI, then it should work. Jon #230 40 17 Feb 87 16:53:21 (RECV'D) From: Jon Kosmoski To: Matt Carpenter REPLY TO #200 Subj: non-jrSASI hard disk system troubles Hi Matt I afraid our experience here at TPL is pretty much limited to standard Xebec and other brand SASI hard disk controllers and our own jrSASI interface sidecar and software. My only comments are that our software will definitely NOT work with any one else's PCjr hard disk system. If you decide to give up on DISKIT, I would say that the case, drives and power supply could probably be salvaged. If the controller is equiva-- lent to a Xebec 1410 SASI hard disk controller it could be used also. At the very least, you would need to buy a jrSASI interface sidecar (including software) from TPL. I'm sorry I couldn't be more help. Jon Kosmoski V.P. of Engineering - TPL Corp. #231 35 04 Aug 86 12:53:15 From: Glenn Beer To: Mike Dougherty Subj: SASI using the Shugart 1610-4 SASI Ctrl'r I have been attempting to run a ST-225 disk using the TPL 800102-000 rev a host board and the sdutil software. So far I have eliminated the host, and meybe the ctrl'r. But it seems to me that the controller is incompatable with sdutil. The cables are all purchased assembled and have been checked for continouity from pc trace to trace. Using debug I have been able to issue the proper commands to the ctrl'r however after formatting (cmd04) I recieve an error message about track 0 not found after reclabration. Is there something you or any one else has come across about this combination that could account for this error, or is my choice of controller incorrect. Glenn Beer 511 N. Poplar Ave. Manteca, Ca. 95336 (209)823-7032 #270 30 27 Feb 87 05:21:06 From: Bert Stark To: Sysop Subj: Sasi interface Is there a way to speed up the sasi interface? I know setting the interleave helps a lot but is it possible to possibly change the 74LS--- chips to 74S-- or even 74HCT--- type chips. The interface works great and does things that often simulate DMA. I run Qmodem 2.2 off the hardisk on the Jr. and do not have DMA. Am I possibly barking up the wrong tree and possibly should I look at the software end of the enhancement ? I use the Segate 225 20 meg and as of yet have no complaints. Thanks in advance . I guesss I am another caught up in the ever widening circle of making things happen faster. Bert Stark: San Antonio, TX . #289 20 03 Mar 87 13:21:59 (RECV'D) From: Jon Kosmoski To: Chuck Holst SEE ALSO #292 Subj: SCSI vs. SASI Hi Chuck . Thanks for your interest in the jrSASI project. Yes, SCSI is a superset of SASI, but no, our interface will not work with a SCSI controller. SCSI and SASI are different enough that changes to the driver and setup software would be required for SCSI. . Yes, we do have sidecar cases to go with the prototype board. I'll have to check on the pricing and get back to you. . Are you aware that we have an assembled and tested version of the jrSASI interface. The assembered version is mounted in a sidecar case and comes complete with the 50-pin cable required to connect the interface to a SASI hard disk controller and the driver and setup software required. Besides the assembed jrSASI interface you still need a SASI hard disk controller, a hard disk and a suitable power supply (at least 1 amp of +5 and 3-4 amps of +12). case and comes complete with the 50-pin cable required to connect the interface to a SASI hard disk controller and the driver and setup software required. Besides the assembed jrSASI interface you still need a SASI hard disk controller, a hard disk and a suitable power supply (at least 1 amp of +5 and 3-4 amps of +12). . The advantage of the assembled and tested interface is obviously, that you don't have to build it and risk not having it work. The cost of the assembled version is $169.95. The protoboard is $39.95. You will have to decide whether you think you are skilled enough to build the interface yourself and save or if you should spend a bit more to get the pre-built and tested unit. . I hope this answered your questions. Feel free to ask any others you may have. . Take care... Jon Kosmoski V.P. Engineering, TPL Corp