Home » RBC Forums » General Discussion » Newbie with Tiny68K (Description of a newbie's experiences with Tiny68K)
Re: Newbie with Tiny68K [message #4243 is a reply to message #4242] |
Thu, 01 February 2018 12:32   |
rhkoolstar
Messages: 276 Registered: October 2015
|
Senior Member |
|
|
Maybe I am missing something, but looking at the schematics and the pcb, the tiny68k terminal is only 4 wires. Rx, Tx, VCC and Gnd.
I'm not surprised when CTS and RTS don't work, they seem not not be wired up at all.
Edit .... disregard, I was looking at the rev 0 board.
Rienk
[Updated on: Thu, 01 February 2018 13:24] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4244 is a reply to message #4243] |
Thu, 01 February 2018 13:26   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
rhkoolstar wrote on Thu, 01 February 2018 15:32Maybe I am missing something, but looking at the schematics and the pcb, the tiny68k terminal is only 4 wires. Rx, Tx, VCC and Gnd.
I'm not surprised when CTS and RTS don't work, they seem not not be wired up at all.
Rienk
On the schematic, check out the 68681. You'll see that Rx & Tx go to the terminal connector pins T7 & T8. You'll also see that OP0 goes to pin T6 (RTS) and IP0 goes to pin T5 (CTS). Finally, +5V & GND go to pins T9 & T10. The terminal connector pins are pointed out in one of Bill's photos.
I can attest that not connecting anything to CTS & RTS means that the Tiny68K board will hang, waiting for RTS. Only if you connect CTS & RTS together will the board start up properly. Of course, if I had a device with the proper handling of CTS & RTS, that would work, too.
smp
[Updated on: Thu, 01 February 2018 13:29] Report message to a moderator
|
|
|
|
|
|
|
|
|
|
Re: Newbie with Tiny68K [message #4262 is a reply to message #4261] |
Sat, 03 February 2018 12:00   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
I'm looking at the CP/M-68K System Guide, and when I execute the command:
it returns:
cpm.rel:26284+2786+16076=45146 (B05A ) stack size = 0
The System Guide says this about the SIZE68 CP/M.REL command:
This gives you the total size of the system in both decimal and hex byte counts. Subtract this number from the highest memory address in your system and add one to get the highest possible address at which CPM.REL can be relocated.
The System Guide goes on to say:
Assuming that the result is aaaaaa, type this command:
A>RELOC -Baaaaaa CPM.REL CPM.SYS
The result is the CPM.SYS file, relocated to load at memory address aaaaaa. If you want CPM.SYS to reside at some other memory address, such as immediately above the exception vector area, you can use RELOC to place the system at that address.
When you perform the relocation, verify that the resulting system does not overlap the TPA as defined in the BIOS. The boundaries of the system are determined by taking the relocation address of CPM.SYS as the base, and adding the size of the system (use SIZE68 on CPM.SYS) to get the upper bound. This address range must not overlap the TPA that the BIOS defines in the Memory Region Table.
So, I see the Transient Program Area (TPA) must be defined in the BIOS. That's where I'll go looking next...
smp
[Updated on: Sat, 03 February 2018 12:05] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4263 is a reply to message #4262] |
Sat, 03 February 2018 12:36   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
I found this in the R5 BIOS:
memrgn: dc.w 1 * 1 memory region
dc.l $20000 * starts above CP/M
dc.l $6e0000 * goes until $7000000. RAM drive above it
* dc.l $dd0000 * goes until $df0000, TPA is ~14meg bytes
Bill, can you please decipher this for me? I can see that the ~14 MB TPA line is commented out. I also remember that you have set up disk drive E: as a large RAM drive. I cannot understand where the TPA is or what size it is by this table.
Thanks very much in advance for your assistance!
smp
[Updated on: Sat, 03 February 2018 12:37] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4264 is a reply to message #4263] |
Sat, 03 February 2018 13:07   |
plasmo
Messages: 916 Registered: March 2017 Location: New Mexico, USA
|
Senior Member |
|
|
I think you have a bad header file with f83.68k. I downloaded forth 83 from cpm.z80.de, unzip it, gkermit them all to CF disk and execute f83.68k and have the same error message as you did. I'm looking for a relocatable executable such as f83.rel. This is because f83.68k may need to reside at different memory to run correctly.
The TPA for Tiny68K is quite large. The size of memrgn (i.e., TPA) is 0x6E0000 or 7208960 bytes (over 7 meg). You can confirm it with the utility that you can download from Gaby's site, TPALEN
http://cpm.z80.de/download/util68k.zip
[Updated on: Sat, 03 February 2018 13:09] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4265 is a reply to message #4264] |
Sat, 03 February 2018 13:11   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
plasmo wrote on Sat, 03 February 2018 16:07I think you have a bad header file with f83.68k. I downloaded forth 83 from cpm.z80.de, unzip it, gkermit them all to hard disk and execute f83.68k and have the same error message as you did. I'm looking for a relocatable executable such as f83.rel. This is because f83.68k may need to reside at different memory to run correctly.
The TPA for Tiny68K is quite large. The size of memrgn (i.e., TPA) is 0x6E0000 or 7208960 bytes (over 7 meg). You can confirm it with the utility that you can download from Gaby's site, TPALEN
http://cpm.z80.de/download/util68k.zip
RATS! I was hoping that it was not the bad header file. If you find another Forth83, please let me know. Thanks very much for the information and the pointer, Bill!
smp
|
|
|
Re: Newbie with Tiny68K [message #4266 is a reply to message #4246] |
Sat, 03 February 2018 13:12   |
 |
lowen
Messages: 226 Registered: August 2016 Location: Western NC USA
|
Senior Member |
|
|
smp wrote on Thu, 01 February 2018 16:40...
Ha! I've never seen CP/M in a ROM.
Well, the CPU280 boots CP/M 3 out of EPROM, all the way to CCP. The ROM isn't accessible as a disk device, but it does have CP/M 3 in ROM.... 
Quote:I have seen the XMODEM application added into CP/M file systems, but it was always added in after the fact. It never came standard with any CP/M file system. Do you know of an XMODEM application coded for the 68K, or a source file that can be assembled for the 68K?
Well, CP/M 68K has a C compiler, and it might not be too hard to get an rzsz derivative to build and run. I don't think YAM was ever ported, but that was as I recall the gold standard for CP/M X/YMODEM-supporting terminal programs back in the day. But rzsz will do everything, X/Y/ZMODEM, and it's simple to use.
I think Kermit was used simply because there is a binary already available from Gaby and others. Kermit is very portable, too.
--
Bughlt: Sckmud
Shut her down Scotty, she's sucking mud again!
|
|
|
|
Re: Newbie with Tiny68K [message #4268 is a reply to message #4266] |
Sat, 03 February 2018 14:25   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
lowen wrote on Sat, 03 February 2018 16:12smp wrote on Thu, 01 February 2018 16:40...
Ha! I've never seen CP/M in a ROM.
Well, the CPU280 boots CP/M 3 out of EPROM, all the way to CCP. The ROM isn't accessible as a disk device, but it does have CP/M 3 in ROM.... 
Quote:I have seen the XMODEM application added into CP/M file systems, but it was always added in after the fact. It never came standard with any CP/M file system. Do you know of an XMODEM application coded for the 68K, or a source file that can be assembled for the 68K?
Well, CP/M 68K has a C compiler, and it might not be too hard to get an rzsz derivative to build and run. I don't think YAM was ever ported, but that was as I recall the gold standard for CP/M X/YMODEM-supporting terminal programs back in the day. But rzsz will do everything, X/Y/ZMODEM, and it's simple to use.
I think Kermit was used simply because there is a binary already available from Gaby and others. Kermit is very portable, too.
Hi Lowen, and thanks very much for your info.
Both you and rhkoolstar have now pointed out systems with a version of CP/M in ROM. Thanks to you both. I guess that even though I have never seen one, they do exist out there.
Yes, if source files can be found for C code or 68K assembly, those files can easily be brought into the CP/M-68K file system for use. As you know, X/Y/ZMODEM/Kermit is critical for importing already compiled/assembled applications. I'm not sure why Kermit is so often found. Perhaps simply because it was earlier than the others? I like the GKermit here on the CP/M-68K because it is the stripped down version made to perform send & receive only, and not all the other bloat often found in the other versions. Kermit tried to be all things to all users, and so I find many different versions across many different machines quite confusing at times. Do I have to set the baud rate? Do I have to indicate that I'm doing binary files? Etc., sometimes for several settings. Oh, well, since I finally have the correct interface device to get GKermit running I'm in good shape now, but I would love to try another method to see if we can get away from needing to use CTS & RTS in the terminal I/O. If you find some source code to try, please share it here!
smp
[Updated on: Sat, 03 February 2018 14:27] Report message to a moderator
|
|
|
|
|
Re: Newbie with Tiny68K [message #4271 is a reply to message #4270] |
Sun, 04 February 2018 06:58   |
plasmo
Messages: 916 Registered: March 2017 Location: New Mexico, USA
|
Senior Member |
|
|
I've already ported Lee Davison's EhBasic to Tiny68K. It was written for EASy68K so used the same EASy68K trap services so the porting was trivia (org to 0x4000 and turn off interrupts). EhBasic is relocatable, so you can move it anywhere in memory without recompiling. I ran your make-$1-from-changes BASIC program and it took 90 seconds to complete (I have to change line 70 because EhBasic didn't understand 'PRINT #3'). You may recall that it took TinyBASIC 2 minutes to complete.
The program ported to Tiny68K is attached below, search for "**X" for my changes to the original code.
-
Attachment: EhBasic.zip
(Size: 245.35KB, Downloaded 261 times)
[Updated on: Sun, 04 February 2018 06:58] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4274 is a reply to message #4271] |
Sun, 04 February 2018 08:52   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
Hey, thanks a million, Bill! Greatly appreciated. Cool to have another BASIC in my arsenal.
smp
UPDATE: WHOO HOO! This is my 35th message, so now I'm a Member, rather than a Junior Member. Yay!
[Updated on: Sun, 04 February 2018 08:53] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4275 is a reply to message #4274] |
Sun, 04 February 2018 10:56   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
Today, I've been playing around with EhBASIC. I also decided to try out my 12 MHz and 16 MHz oscillators.
Here are my eyeball times for my DOLLAR.BAS program with the three different oscillators installed:
8 MHz 12 MHz 16 MHz
----- ------ ------
1:45 1:11 0:53
The easiest check on this performance is to see that at 8 MHz the run took 105 seconds and at 16 MHz the run took 53 seconds, pretty much right on at half the time.
This is excellent. I am still using the MC68000P8 that came installed with my board. I have not yet dared to pull the processor and try the MC68000P12F 16 MHz chip I purchased on eBay. Maybe another time when I have my courage up.
Is there any risk to keep using the MC68000P8 at 16 MHz? It seems to me no, but I'd like to hear other opinions on this.
Thanks!
smp
UPDATE: I also tried out the ASCIIART.BAS program at 16 MHz - only 43 seconds!
[Updated on: Sun, 04 February 2018 11:09] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4276 is a reply to message #4275] |
Sun, 04 February 2018 11:44   |
plasmo
Messages: 916 Registered: March 2017 Location: New Mexico, USA
|
Senior Member |
|
|
I'm surprised that -P8 part can run at 16 MHz. Now, I've never tried with Tiny68K so I have no empirical data to show how unusual that is. At 16Mhz the DRAM access is probably out of spec, on paper, but not as far out as the 68000. The primary danger to overclocking is heat; your part is rated at 70 degree C and they are probably around 50 degree C now, so no, there are no danger. Another trick of getting more performance is to raise the voltage; 5.25V is the spec, but they can go as high as 5.5V. Parts also get hotter at higher voltage, so risk of damage increases with voltage. Put your finger on the part, 50 degree C is when you are barely able to keep your finger on the part. You can always put a fan on it if you are really worry, but I wouldn't bother.
|
|
|
Re: Newbie with Tiny68K [message #4277 is a reply to message #4276] |
Sun, 04 February 2018 12:09   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
Ah, yes, heat. I've been checking this afternoon, and the 68000 and the Altera device, and the memory are all just barely warm to the touch. I have a 5V @ 2A wall wart supply, and my 5V is reading 4.94V. I'm not going to go for a higher voltage at this point. I was just trying to see what the board can do with a few component changes. If the DRAM is now out of spec, that will probably bother my conscience, but until I see any real flakey stuff going on, I'll probably stay with 16 MHz for the time being.
Fun stuff! Please let me know if you know of any other 68K Forth available anywhere out there. Or any other interesting programs I ought to be trying.
I found the original reference manual for the 6502 EhBASIC, and I extracted the language reference pages in the attached document, if anyone is interested.
smp
|
|
|
|
|
|
Re: Newbie with Tiny68K [message #4294 is a reply to message #4290] |
Mon, 05 February 2018 13:26   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
plasmo wrote on Mon, 05 February 2018 10:18Also take a look at CPM68K in http://www.retroarchive.org/cpm/cdrom/CPM/CPM68K/
zamp has written up instruction on how to unsqueeze the files:
https://www.retrobrewcomputers.org/forum/index.php?t=msg& ;amp ;amp ;amp ;amp ;amp ;th=222&goto=3757&#msg_3757
I tried the F83V2-68.LBR. It unsqueeze OK, but still won't execute on Tiny68K
Hi Bill,
Thanks for the pointers. I downloaded that stuff. I see that the files in F83v2-68.LBR are almost all files that have been compacted by other methods, but apparently not the F83.68K file. I extracted it using LU, and I verify that I got the same "Insufficient memory or bad file header" error message.
I also downloaded UTILS.LBR. Here is a listing of whats in it:
-C > -o utils
lu: old library file UTILS.LBR has 32 entries, 0 free
-O > -l
Name Type Start Length CRC
==================================
DIRECTORY 0 8 0000
CONFIG .HQ 8 5 A504
FROMHEX .CQ 13 12 AD04
HEX .MQN 25 8 EA44
IBMREAD .6QK 33 158 2584
IBMREAD .CQ 191 36 F859
IBMREAD .MQN 227 8 9C92
INDENT .CQ 235 164 DA99
INDENT .DQC 399 29 5402
INDNTCOD.HQ 428 8 D5C5
INDNTGLO.HQ 436 28 C97C
IO .CQ 464 62 CE6F
KILLDU .6QK 526 142 256D
KILLDU .C 668 5 D331
KILLDU .MAN 673 4 80E2
LEXI .CQ 677 58 933A
LKINDENT.SUB 735 1 2276
PARSE .CQ 736 51 9EEC
PRCOMMEN.CQ 787 48 5C4C
PTOC .6QK 835 133 A46B
PTOC .CQ 968 21 E984
PTOC .MQN 989 7 6FF4
SORTFL .6QK 996 149 6DC9
SORTFL .C 1145 6 3EC9
SORTFL .MAN 1151 3 289A
TOHEX .C 1154 9 2C76
UTFILES .SQB 1163 6 D37E
WEED .MAN 1169 3 F0D7
WEED .SUB 1172 1 CDA4
XDIR .CQ 1173 76 53C4
XDIR .MQN 1249 11 9D78
XDIRLINK.SUB 1260 1 379D
==================================
Used sectors 1261
Deleted sectors 0
Total sectors 1261
Active entries = 32, deleted = 0, free = 0, total = 32
-L > -c
lu: old library UTILS.LBR closed
The files on retroarchive.org are going to be difficult to retrieve if there are a number of unpacking methods that have to be used. It appears that we will need additional tools besides the LU tool. See how the file extensions have been changed from 68K to 6QK, and from SUB to SQB, and from H to HQ, and from C to CQ, etc. Some other file compacter will be needed to unpack those files after they've been extracted from the LBR file.
Does anyone have any advice to offer on this? Are those additional tools actually in the XLU68K.LBR file? Or, will the -eu command get the job done from the LU tool? I definitely need another tutorial like the one that zamp posted for how to use the LU tool!
Thanks for listening!
smp
UPDATE: BTW, I'm doing all of this with my 16 MHz oscillator still installed. Everything seems to be operating OK.
[Updated on: Mon, 05 February 2018 13:50] Report message to a moderator
|
|
|
Re: Newbie with Tiny68K [message #4295 is a reply to message #4286] |
Mon, 05 February 2018 14:12   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
norwestrzh wrote on Sun, 04 February 2018 19:02> Please let me know if you know of any other 68K Forth available anywhere out there.
Have you tried: http://home.earthlink.net/~schultdw/cpm68/simulator.html
There is a F83 in his simulator disk that is pretty nice -- big vocabulary.
Roger
P.S. It is fairly common to be able to overclock quality Motorola 68k CPU's. I've seen 8 MHz parts go 10 or 12, but I've never seen an 8 MHz part run at twice the nominal clock!
Hi Roger, and thanks very much for this pointer.
I am not a very competent 68K guy at this point, so please bear with me. It appears to me that the F83 you refer to is probably in the file DISKC.CPM.FS. Is there any way to extract it from that file, short of make-ing the simulator and then manually pulling the file out somehow? I'm not very familiar with running simulators at all, so I have no experience here.
Thanks in advance for your advice!
smp
[Updated on: Mon, 05 February 2018 14:12] Report message to a moderator
|
|
|
|
Re: Newbie with Tiny68K [message #4298 is a reply to message #4297] |
Tue, 06 February 2018 06:44   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
plasmo wrote on Tue, 06 February 2018 07:09The -eu command in LU.68K should take care of the extract & unsqueezing. So all you need to do is open the file (-o utils), extract & unsqueezing (-eu *.*), and exit (-x). You should see all the files extracted.
This is the DRAM diagnostic I used to test out the 16 meg memory. It should take about 3 minutes to run each iteration at 16MHz. It will repeat the test with different seed values indefinitely. To test the serial port, you can use gkermit to send a file to PC and read it back and compare the two files using fcomp. To test the CF, you can initialize drive E (RAMdrive) and pip e:=d:*.*[v] (copy with verify). This suite of tests should provide pretty good fault coverage and give you more confidence in running at 16MHz.
Thanks a million for all your help and advice, Bill!
I am running the memory test now - one iteration has completed successfully already. I have been using GKermit, mostly pulling things into the Tiny68K board, but I've gone both ways without any hiccups. After a while on the memory test, I'll check the CF, too. This is great. Thanks very much.
Thanks also for the advice on the LU tool operation. While I am not new to vintage computers, I haven't used some of these tools before and it's nice to hear that I'm not wandering in the weeds too badly.
smp
|
|
|
|
Re: Newbie with Tiny68K [message #4300 is a reply to message #4299] |
Tue, 06 February 2018 11:19   |
smp
Messages: 49 Registered: January 2018 Location: Bedford, NH, USA
|
Member |
|
|
norwestrzh wrote on Tue, 06 February 2018 13:10>> It appears to me that the F83 you refer to is probably in the file DISKC.CPM.FS. Is there any way
>> to extract it from that file, short of make-ing the simulator and then manually pulling the file
>> out somehow? I'm not very familiar with running simulators at all, so I have no experience here.
Sure!! Are you familiar with cpmtools?? Do you use it??? Easy to extract it that way.
It is a two part application: F83.68k loads F83.BIN. If you don't use cpmtools or if it is too
much of an effort to build and use it, I can make you copies -- either binary, or S-record.
Need a way of getting the two files to you.
Roger
Hi again, Roger,
I obtained a win32 copy of the cpmtools. when I run CPMLS on the file DISKC.CPM.FS, I get a 0: followed by about 30 blank lines, and then I'm back at the command prompt. I assume that's the directory listing for user 0, but all blank lines? Huh?
As you can see, I'm not familiar with what I should be doing here. If you would be so kind as to provide me copies of f83.68k and f83.bin when you have some spare time, I would greatly appreciate that. I will use GKermit to pull these files into CP/M on my system, so binary files would be fine. Can you post them here, so anyone else who's interested can also benefit?
Thanks very much!
smp
|
|
|
|
|
|
|
|
|
|
Re: Newbie with Tiny68K [message #4309 is a reply to message #4304] |
Tue, 06 February 2018 16:29   |
plasmo
Messages: 916 Registered: March 2017 Location: New Mexico, USA
|
Senior Member |
|
|
smp wrote on Tue, 06 February 2018 15:13plasmo wrote on Tue, 06 February 2018 07:09 ... To test the CF, you can initialize drive E (RAMdrive) and pip e:=d:*.*[v] (copy with verify). This suite of tests should provide pretty good fault coverage and give you more confidence in running at 16MHz.
OK! I performed the CF test as described a few times with no errors reported.
It sure looks to me like I have a pretty solid 16 MHz Tiny68K system! Do you think perhaps the MC68000P8 is actually a faster part labelled as a P8? I know that chip makers sometimes do things like that for efficiency. Since the 16 MHz parts are being made, rather than change production lines, just label the faster parts with the slower label?
I don't know, but I'm certainly happy that my board is running at 16 MHz and seems to be nice and stable. That was a pretty cheap upgrade, too. The 16 MHz oscillator cost $2.47 (plus mailing) at Digi-Key.
smp
I certainly don't expect your board to run at 16MHz. In my standard margin test I only test to 12MHz and never even thought about trying 16MHz because I don't expect it to pass. Now you've got me curious and I tried 16MHz on a couple of my boards. One passed and other failed. Hmmm, that's pretty interesting. Now I wonder just how close to the edge the 16MHz board is. One way to find out is to raise the clock by 10% and run through the test suite or lower the voltage 5% and see if it still pass the tests. If so, I'll continue to operate it at 16MHz. If not, I'd roll back the clock by 10% and run at 14.3MHz or chose to run at 16MHz anyway. We are not talking about mission critical component so running it up close to the edge is just for fun.
|
|
|
Current Time: Sat Feb 08 20:41:35 PST 2025
Total time taken to generate the page: 0.00936 seconds
|