Newbrain Expansion Port

PinSignal NameDescription / NotesEIM
1GNDGround50E
2Z80 CLOCK 4MHzSystem Clock (Directly for CPU/Bus)49E
3A14Address Bus Bit 1448E
4A13Address Bus Bit 1347E
5D5Data Bus Bit 546E
6RMSLRAM Select (connect to ground to switch 16k rams)45E
7D4Data Bus Bit 444E
8D3Data Bus Bit 343E
9D6Data Bus Bit 642E
10D7Data Bus Bit 741E
11A11Address Bus Bit 1140E
12A10Address Bus Bit 1039E
13A8Address Bus Bit 838E
14A9Address Bus Bit 937E
15A12Address Bus Bit 1236E
16A7Address Bus Bit 735E
17A3Address Bus Bit 334E
18A2Address Bus Bit 233E
19A1Address Bus Bit 132E
20A0Address Bus Bit 031E
21D0Data Bus Bit 030E
22D1Data Bus Bit 129E
23FCTRFunctional Counter / Clock?28E
24D2Data Bus Bit 227E
25A5Address Bus Bit 526E
26A6Address Bus Bit 625E
27RAMENBRAM Enable (Internal logic toggle)24E
28EXRM2Expansion RAM Page Bit 2 (Output to NB)23E
29EXRM1Expansion RAM Page Bit 1 (Output to NB)22E
30EXRM0Expansion RAM Page Bit 0 (Output to NB)21E
31nROMOVROM Override20E
32nBUSRQ​Bus Request19E
33nM1Machine Cycle 1 (Opcode Fetch)18E
34nRSTReset Signal17E
35RFRSHDRAM Refresh (Safe to ignore for SRAM)16E
36nWAITCPU Wait State15E
37A4Address Bus Bit 414E
38nBUSACKBus Acknowledge (CPU has released bus)13E
39A15Address Bus Bit 1512E
40nWRWrite Enable11E
41nINTInterrupt Request10E
42nRDRead Enable9E
43nNMINon-Maskable Interrupt8E
44nHALTCPU Halt State7E
45nMREQ​Memory Request6E
46nIOREQ​I/O Request5E
47PRTOVPort 0xFF Override / Select4E
48RAMINHRAM Inhibit3E
49+5VSystem Power2E
50BUSRQ Bus Request (TV)1E

7 thoughts on “Newbrain Expansion Port

  1. A quick observation on the connector image in your Expansion Port post.

    The two green circles are clear: pin 1 at the bottom-left, and pin 6 on the top row, third position. Together they fix the numbering scheme: odd pins (1, 3, … 49) along the bottom row and even pins (2, 4, … 50) along the top, running left to right as viewed from the rear with the keying notch at the bottom.

    The two red circles are harder to read: their callout arrows point in from outside the frame, so their labels seem to have been cropped out of the published image. As placed, one sits on the bottom row around the 22nd–23rd position and the other on the top row at the far right. If they’re meant to mark pins 46 and 49, that doesn’t fit the scheme the green circles establish – pin 46 is even and would be on the top row, pin 49 is odd and would be on the bottom row, and the top-rightmost pin would be 50. Could you clarify which pins the red circles mark, or re-upload the image at full resolution with the labels included, please?

    Also, out of interest: how did you establish the pin-1 position — continuity to ground, tracing on the main board, or another way? I’m doing some work against this connector and would like to know how firmly the orientation is established. Many thanks

    1. The two green circles are correct and if you jumper them you switch banks in ram low 16k becomes high and vice versa. The red ones was pointed to do that from another site and when i tried it it didnt work so i traced the problem and it was numbered incorrectly. The photo is from another post that i did to point out that issue.

  2. One other point on the signal table. Your pin 23 entry – FCTR, “Functional Counter / Clock?” – caught my eye. The Grundy Software Technical Manual Appendix E lists pin 23 as unused, but with an interesting note: early machines do carry a signal on it, and Appendix E advises a 10 kΩ pull-up to 5 V and warns against driving the pin. The Expansion Board documentation agrees. Do you know where the FCTR identification came from: was it traced on a physical NewBrain, or from another source? If it names the signal the early machines carried, that would be a real find, since the Grundy material never identifies it.

    Two smaller naming points, mainly for anyone cross-referencing against the Grundy documents: pin 48 appears there as RAMINB (“logic one inhibits Z80A request to internal RAM”) rather than RAMINH – I think RAMINH may be MAME’s name for it – and PRTOV on pin 47 is described more generally as inhibiting the machine’s internal I/O port decode, rather than as a port 0xFF select. Appendix F’s port 255 paging/status-register mechanism appears to be a separate thing, so I’d also be interested to know whether your 0xFF description comes from tracing the hardware or another source. Many thanks.

    1. The comments are from Gemini when i was trying to replace the drams with a sram so some of them could be just guessing. FCTR is the name on the schematics and so is the prtov. Now fctr i do not know what it does but prtov is port override probably. Raminh is also how the schematics has it and it is probably correct.

  3. Thanks, Chris. That clears it up nicely, and the provenance of the annotations is particularly helpful.

    For anyone else using the photograph as a connector reference, my understanding from your explanation is:

    * the two green circles are the correct positions. They are pins 1 (GND) and 6 (RMSL), so linking them grounds RMSL and swaps the low/high 16K RAM banks;
    * the red circles are deliberately showing the incorrect positions from another published source which you subsequently found did not work and traced as a numbering error. So they shouldn’t be read as additional pin-number callouts on this image;
    * the explanatory descriptions in the table were partly generated with Gemini during your SRAM-replacement work, so those descriptions shouldn’t necessarily be treated as source documentation;
    * however, the signal names FCTR, PRTOV and RAMINH came from the schematics rather than from Gemini.

    The FCTR point is especially interesting. The Grundy Software Technical Manual Appendix E leaves pin 23 unnamed/unused for expansion purposes, but notes that early machines carry a signal on it and advises a 10K pull-up to +5V and that the pin should not be driven. I’ve since also found a NewBrain On-Line magazine description which says pin 23 goes into the video circuitry; pulling it low shifts the picture two positions to the left, and noise on it can cause picture jitter; again recommending a 10K pull-up. So FCTR may well be the actual schematic name of that otherwise rather mysterious early-machine signal.

    If you happen to know which schematic or NewBrain board revision shows the name FCTR, that would be very useful to pin down.

    Likewise, for anyone cross-referencing sources, there seems simply to be some nomenclature variation around pin 48: the Grundy Software Technical Manual uses RAMINB, whereas the schematic material you have uses RAMINH; both describe the same RAM-inhibit function.

    Many thanks again for taking the time to check this and explain how the photograph came about – it makes the page much more useful as a reference.

  4. Thanks Chris, found it. You were quite right: the Grundy main-board schematic does explicitly label the expansion pin 23 signal FCTR, and I can also see RAMINH and PRTOV there.

    That clears up the apparent discrepancy nicely. Appendix E of the Software Technical Manual doesn’t name FCTR, but its warning about pin 23 now makes sense alongside the schematic: it is an internal signal brought out on early machines, rather than a signal intended for use by expansion hardware.

    The NewBrain On-Line material adds the behavioural piece: pin 23 goes into the video circuitry; pulling it low shifts the picture two positions and interference can cause picture jitter, hence the recommendation for a 10K pull-up to +5V.

    So for anyone coming across this later: FCTR is a genuine Grundy schematic signal name; “Functional Counter / Clock?” was only the tentative description and shouldn’t be taken as the meaning of the acronym. I haven’t yet found a Grundy document that actually expands FCTR.

    Likewise, RAMINH is indeed the schematic nomenclature, while the Software Technical Manual uses RAMINB for the same RAM-inhibit function.

    Many thanks for pointing me back to the schematic set – that resolves it.

Leave a Reply