Recreating the Bywood Scrumpi

[Slothie]

Originally Posted by Timbucus

Excellent news - I am so pleased for you

Here is the latest edition of the Guide - give the IRQ one a try - just remember that SA and SB are reversed on your screen printing so SB will be IRQ (that should be corrected in the next one)...

The Soul of an Old Machine DRAFT v3

I've just read through and it's excellent however
the CONT pin of the SC/MP connected to 'x' seems to have been left floating on the assembled boards yet it's referred to as being tied high. Should 'x' be connected to +5v if using a SCMP 2?

[Timbucus]

Originally Posted by Slothie

I've just read through and it's excellent however
the CONT pin of the SC/MP connected to 'x' seems to have been left floating on the assembled boards yet it's referred to as being tied high. Should 'x' be connected to +5v if using a SCMP 2?

Ah yes - Phil and I had this discussion - in the end he ran a line under the board from the Via to the top of R11 on the basis it was easier to cut to be able to use the X input than have to strap it somewhere for normal running. It has to be be tied high for an SC/MP 1 as well.

[philoupat83]

exact x is connected to the vcc but can be cut for another use.
life thing is great

[SiriusHardware]

it occurred to me that when you have an SC/MP II fitted you might be able to use a SIL resistor package (with the common at one end taken to +5V via a flying insulated lead or link) and also make provision for one of those to be used on any future revision of the SCRUMPI I PCB, rather than having to tack all those pull up resistors on the back of the PCB. I assume that the Scrumpi II is SC/MP II from the start and will have provision on the PCB for pullups, probably individual resistors.

I'm glad you have decided to emphasise the use of a 'proper' +5V / -7V PSU, the original 'Use a 12V supply and a zener' comment by DJD is dangerously low on detail given that the purchase of an SC/MP I now doesn't seem to leave much change from £100 - even when the machine was contemporary they cost more than £10, which was a lot of money then. I wouldn't entrust anything so rare and expensive to the tender mercies of an ad-hoc power supply arrangement.

[philoupat83]

and yes before connecting the scrumpi well check
I first check without CI, with CI logic and finally
with the 8a/500d (gold and ceramics) I retained my soufle and the!
the thing life!!
what happiness

[Slothie]

I've been thinking about the zener diode arrangement and you'd need a 5w or 10w zener - the latter are very expensive (~£20) and the former would be operating close to their max power. So a pair of regulators is probably safer and cheaper.

[Timbucus]

Indeed and of course they will vary under load and temperature - thanks for answering that it was not something I was very familiar with as a main supply - I am quite surprised how warm my PSU is getting - next time I don't have a large program in memory I will lose I will insert a meter in the feed and see what the current draw is.

[Mark1960]

I have used a 15v laptop power supply with a -5v 7905 regulator to power an intel 4040 system, similar method might work with a 12 v psu for the scmp. The + from the psu to provide the +5v, - from psu to provide -7v, then the 7905 output provides 0v to the 5v chips.

[Timbucus]

Another good idea Mark - if I eventually locate a PMOS version of the Chip we can give it a go!

On the current draw it was just in-between the range my meter could read - it jumps from 200mA to 10A on ranges and the 10A will not power the board...

Anyway the desktop PSU reads about 260mA when paused and 320mA when running full tilt - no latch chips fitted, mostly LS chips (the data and flag buffers are CMOS) and AM9112 type memory.

[philoupat83]

Hello
good news the pcbs are in production
one for the closest scrumpi to the original and one where it is possible to choose sc/mp 1 or 2
I made a power supply -7V -12V +5V
it is necessary to use a regulator 7805 to3 because the current circuit breaker one to220 the transformer makes 2x12V 10 VA

[philoupat83]

see another picture

[Timbucus]

Wow - quick work - do you mean you have already ordered them or are working on the layout? I might have a few suggestions if the latter...

The second one would be handy for people who want to use a Crystal clock for a PMOS SCRUMPI as well I suspect. I will update the manual to reference the options board as a good choice.

I will also have some Serial BOOT code I have nearly finished to allow loading code from a PC over SIN and data out from SOUT soon. It does need a GND connection so which was going to be my suggestion to have somewhere to pick it up with a PIN...

Nice looking PSU - very retro with the to220

[philoupat83]

tim
yes possibility of a quartz for ins 8a/600d gives me the modifications you wish

[Timbucus]

Indeed - the other thing is an easy and obvious way to link -7v to +5v perhaps two extra pads on the input lines - I soldered a patch under the board but, an obvious space on the top would be better I think?

I was going to suggest the pullups between the switches but, I see you have thought of that! Are you sure the switch holes are big enough - they are pretty substantial...

[Timbucus]

This is the serial bootstrap code I am testing - you just assemble your program at address 0x27, enter these 38 bytes with the switches and then use the SIO to get the raw BINARY data for the remaining 218 bytes...

*****************************************************************************
0000-                  5        ;
0000-                  6        ; SC/MP Serial IN adapted from one in the SoC Manual for the MK14
0000-                  7        ; Receive remainder of 256 bytes via Serial IN then execute program.
0000-                  8        ; Shortest version with no ECHO back - note you cannot drop below 1200 baud
0000-                  9        ; due to the trick we use to pass the Half time wait from Start in Extension
0000-                 10        ;
0000-                 11        ; Typed in for SBASM by T.Gilberts June 2020
0000-                 12        ;
0000-                 13        ; To use minicom to send a binary file...
0000-                 14        ;  Add a line to minicom config -> File transfer protocols (CTRL A + O inside)
0000-                 15        ;  J  binary     /bin/cat                        Y    U    N       Y       N
0000-                 16        ;
0000-                 17        ; *****************************************************************************
0000-                 18
0000-                 19                .OR     0x0000
0000-                 20
0000-                 21        ;Note that this is 2Mhz SCRUMPI at 4Mhz Mk14 will use next line up so 300=600
0000-                 22        ;Baud   HBTF    BTF     BTC
0000-                 23        ;4800   0x05    0x16    0x00
0000-                 24        ;2400   0x1F    0x4A    0x00
0000-                 25        ;1200   0x52    0xB2    0x00
0000-                 26
0052-                 27        HBTF    .EQ     0x52            ;1200 baud SCRUMPI = 2400 baud MK14
00B2-                 28        BTF     .EQ     0xB2
0000-                 29        BTC     .EQ     0x00
0000-                 30
000A-                 31        BITCNT  .EQ     10              ;1 start (half anyway), 8 data and 1 stop
0000-                 32
0000-00               33        LOOPCNT:        .DB     0x00    ;Use unused zero address on SCRUMPI for variable
0001-                 34
0001-                 35        ENTER:
0001-C4 27            36 (  10)         LDI     #DEST&255       ;Reduce by ONE if using to repeat JUMP
0003-33               37 (   8)         XPAL    P3
0004-                 38
0004-                 39        GETCHR:
0004-C4 00            40 (  10)         LDI     /DEST           ;We set this because we destroy it at end of loop
0006-37               41 (   8)         XPAH    P3              ;and it means high byte which is indeterminte is set for MK14
0007-                 42
0007-08               43 (   5)         NOP                     ;Will be patched by user with 0x3F for XPPC P3 to start code
0008-                 44
0008-C4 0A            45 (  10)         LDI     BITCNT
000A-C8 F5            46 (  18)         ST      LOOPCNT-$-1(P0)
000C-                 47
000C-                 48        AGAIN:
000C-C4 A5            49 (  10)         LDI     HBTF*2|1        ;Wait until we see a low pulse sampling
000E-01               50 (   7)         XAE                     ;SIN but, keep outgoing line high
000F-19               51 (   5)         SIO
0010-40               52 (   6)         LDE
0011-94 05            53 (9/11)         JP      WAIT            ;If Bit 7 is 0 then +ve
0013-90 F7            54 (  11)         JMP     AGAIN           ;Enter 00 as offset before STEP after LOAD
0015-                 55
0015-                 56        ;We should really check at the half a pulse again that it was still low but that
0015-                 57        ;adds a lot of bytes so we hope to get away with it.
0015-                 58
0015-                 59        LOOPB:
0015-19               60 (   5)         SIO                     ;Sample the bit
0016-C4 B2            61 (  10)         LDI     BTF             ;Wait a full pulse to center of next one now
0018-                 62
0018-                 63        WAIT:                           ;If we enter here we are waiting only half a pulse
0018-8F 00            64 ( 13+)         DLY     BTC
001A-B8 E5            65 (  22)         DLD     LOOPCNT-$-1(P0)
001C-9C F7            66 (9/11)         JNZ     LOOPB
001E-                 67
001E-40               68 (   6)         LDE                     ;Get the byte
001F-                 69
001F-CF 01            70 (  18)         ST      @1(P3)
0021-37               71 (   8)         XPAH    P3              ;Is it end of 256 byte block
0022-E4 01            72 (  10)         XRI     /DEST+0x0100
0024-9C DE            73 (9/11)         JNZ     GETCHR
0026-                 74
0026-                 75        ;Make sure to allow user to check send was OK before running, they can step now
0026-                 76
0026-                 77        HALT:
0026-00               78 (   8)         HALT

[circuitryboy]

Hi, I've been catching up with the latest SC/MP threads.
Cricklewood Electronics have been listing "Museum Piece"
SP-8A/500D for years. Currently £30.

[Timbucus]

Originally Posted by circuitryboy

Hi, I've been catching up with the latest SC/MP threads.
Cricklewood Electronics have been listing "Museum Piece"
SP-8A/500D for years. Currently £30.

Nice one thanks for the heads up ordered! Now I need another PCB when Phil finishes them, and more switches and LED's with built in resistance and ... and ... yay

[philoupat83]

hello tim
the pcbs are in production
for switches I think the holes are big enough
I’ll save you a surprise for scrumpi 2 which is the next step
I will meanwhile the pcbs build an original mm5204 eprom programmer
a read the kirkpatrick articles on the S68 and the ETIBUG monitor

[Mark1960]

Originally Posted by philoupat83

for switches I think the holes are big enough

For the same style switch but with solder tag type termination I used 2mm hole diameter. The switches were then a tight push fit, but not too hard to fit. I didn’t need to solder the switches, though they have started to loosen up a bit now and I may end up soldering them soon.

[philoupat83]

hello mark
yes the holes are 2mm
the space between them is enough for the version with ins8060 with pullup as TIM experimented on it model