4th Aug 2020, 9:44 am | #41 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Mine is still a work in progress, so I can easily make that change if necessary. (Welcome back, by the way).
|
4th Aug 2020, 6:55 pm | #42 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
When I drew the diagram way back when I didn't notice the 6116 had a seperate output enable, so I included the NRDS and NWDS in the chip enable, I really need to redraw that diagram. I'll get right on it since people are building it!!
|
4th Aug 2020, 7:11 pm | #43 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
I lashed up a preliminary version today using only the 74LS27 - NRDS and NWDS going only to _0E and _WR, and the 74LS27 providing _CE when...
Either A9 or A10 (or both) are high and A11 is low. However on first test I don't seem to be getting the _CE when I read from or write to 0200-07FF and the addresses are all reading 0x1F which is the usual value read from a memory hole. I've just checked all the physical connections and they are as I intended them to be. I have to admit my logic is terrible, though, so I may have mucked up the simplification of the _CE circuit. |
4th Aug 2020, 7:45 pm | #44 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
This is my modified version of the _CE circuit. What have I done wrong? Yes, I could have used a 74x02 but there was already a 74LS27 in place.
|
4th Aug 2020, 8:25 pm | #45 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
Quote:
Code:
ian@FX503VD:~/Projects/Test$ python2 ~/Bin/lsim mk14ram.sim .inputs a11 a10 a9 .outputs CE x: nor a10 a9 y: nor x a11 CE: nor y y a11 a10 a9 | CE -----------+---- 0 0 0 | 1 0 0 1 | 0 0 1 0 | 0 0 1 1 | 0 1 0 0 | 1 1 0 1 | 1 1 1 0 | 1 1 1 1 | 1 Last edited by Slothie; 4th Aug 2020 at 8:29 pm. Reason: Image is wrong |
|
4th Aug 2020, 8:34 pm | #46 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Hmm, strange. I'm just going to have to check it all again. I even went so far as to check that the A9-A10-A11 lines actually go to the right edge connector fingers on the issue VI PCB (...they do, of course).
The 74LS27 should be OK, came from RS - unfortunately I can't just unplug it and drop it into my device tester because I have soldered it pins-flat onto the copper side of a bit of veroboard. |
4th Aug 2020, 10:47 pm | #47 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Looked again more closely and found that my logic probe was actually seeing short _CE pulses (TTL mode : Pulse mode 'on') and only when stepping through the address range 0200-07FF. Outside that range, no CE pulses. And yet still no memory operation.
I've just found the reason why, my NRDS and NWDS pickup wires are crossed. Too tired to drag the soldering iron out to sort it tonight, I'll do it tomorrow and report back. Funnily enough Ian, it was the unintentional RD / WD 'twist' on your diagram in #45 which prompted me to check that, so thanks! |
4th Aug 2020, 11:03 pm | #48 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
Quote:
|
|
4th Aug 2020, 11:23 pm | #49 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Great minds flunk alike...
|
5th Aug 2020, 7:58 pm | #50 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Yes!
With the NRDS and NWDS signals actually being picked up from the right places, it works, I can enter data anywhere between 0200-07FF and it 'sticks'. So now I have a 'RAM Pack' which I can just plug in for an instant extra 1.5K of MK14 RAM. The chip enable circuit for the 6116 is as per my rough sketch using a 74LS27 in #44, Slothie's streamlined circuit using a 74x02 in #45 will obviously also work fine as long as you route NRDS to RAM _OE and NWDS to RAM _WR. The 6116 RAM in my case is a brand I haven't really heard of, an 'Elcap' EL6116-10. We had a tube at work. I haven't connected the VDU to the far end of the 'bridge PCB' since I added the RAM to it - I will, but when I do I'll continue in the VDU thread. |
5th Aug 2020, 9:08 pm | #51 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
This is how the 'Bridge' PCB has evolved, it's made from two back to back bits of stripboard with a connector for the MK14 on one end and a connector for the VDU on the other.
On the underside, the side which mainly carries the bus signals, almost all signals are routed straight through one-to-one. (Image #1). The exception is the XOUT clock signal from the MK14 which starts out on the underside of the MK14 edge connector, but is jumped from there to the same track on the top side of the bridge board. On the top side there is an untidy looking array of socket connectors allowing routing of the VDU control connections with Arduino-style plug leads. It's not easy to see in the image (#2) but in all cases where 8154 port pins would be directly connected to VDU control pins there is a series resistor between the 8154 port pin and the VDU control pin / pluggable connector. This is so that the 8154 ports will not be damaged if they are inadvertently placed in output mode when there is a wire link pulling the VDU control line high or low. The TOP PAGE output from the VDU also has a resistor in series with it. Images #1 and #2 show the 'Bridge Board' as it was when it was basically just a 'joiner' which fitted between the MK14 issue VI and the VDU. In image #3 the underside suddenly looks very busy, as a RAM IC and address decoder IC have been added to a smaller square of stripboard which has been mounted on the underside of the Bridge PCB. The bus signals still pass straight through underneath the smaller memory board, but the memory board now picks up the signals it needs from the tracks on the main lower board. Note that this image was taken before minor issues like the routing of NRDS / NWDS were sorted out, but it still looks pretty much the same. |
5th Aug 2020, 10:18 pm | #52 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: MK14 memory upgrade
Wow nice well done glad it is working - I must have another go at mine now you have proved it CAN work... neat job on the unit looks really excellent. I think I have a 7402 somewhere so I will scrap my CMOS adaptation of the original circuit and rewire.
|
5th Aug 2020, 10:43 pm | #53 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
I just wish I could suddenly acquire Phil and Slothie's mastery of Kicad and make a proper pcb version of it. I really must at least try to get the hang of that software.
|
5th Aug 2020, 10:50 pm | #54 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: MK14 memory upgrade
Looks like an ideal project to learn with... I always find it best to have a definite goal to make it interesting enough.
|
5th Aug 2020, 11:01 pm | #55 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
Quote:
I've been checking out Kicad 5 and it looks like they've taken out some of the quirky things with libraries so it all seems a little less arcane. I've also heard good things about EasyEDA which was online based but now has a downloadable version. It also integrates with JLCPCB and chinese component supplier LCSC so symbol/footprint support is excellent. Of course I expect you'll have to roll your own retro components but that's going to be true of all of them. There's quite a thrill to holding a PCB you've designed yourself and looks professional! And its quite a relief when someone else builds it and it works! |
|
5th Aug 2020, 11:12 pm | #56 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,556
|
Re: MK14 memory upgrade
Yes, I still occasionally go down the side road of wondering whether any given problem I happen to have is down to some as yet undiscovered problem on the issue VI, but in the end it's always been my mistake, never yours. You did a miraculous job with it, really.
|
28th Oct 2020, 10:47 pm | #57 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: MK14 memory upgrade
Well as you may have seen in another thread my 6116 chips are not real - I should have noticed as they are a funny shape. Anyway I had a few 62256 which I thought should work and are the same pinout nearly. I just tied the extra A11,A12, A13 and A14 to GND
having taken note of the NWDS to _WE and NRDS to _OE - I initially worked off the diagram and got it the wrong way around... I get data now at 0x200 (although the lower nibble is always 0xF suspiciously but, I cannot write - normal memory F00,B00,0880 works fine read and write. So not sure what could be causing the issue. Pulled the 7402 (note it is that as well not LS) and it tests fine in the 866 and the quick tester. Any suggestions where to look - this is so simple it is annoying me. |
28th Oct 2020, 11:09 pm | #58 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
Quote:
https://www.futurlec.com/Memory/62256.shtml |
|
28th Oct 2020, 11:27 pm | #59 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: MK14 memory upgrade
That was the sheet I have been using and Pin 22 is marked as _OE - the truth table seemed the same that the _CS would leave it tristate until either NWDS or NRDS went low. Have I missed something...
|
28th Oct 2020, 11:33 pm | #60 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 memory upgrade
No, I misread the data sheet On closer inspection you're right, the signals should be the same... I'd look for an error in the CS signal given the lower 4 bits are always 1 - that indicates nothing driving the bus. Could be something as simple as a bent pin on the CS line?
|