12th Oct 2020, 8:23 pm | #361 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
Quote:
Hold on. (Rustle Rustle) A quick look at the diagram for the original VDU also shows no connection to NWDS, so it was not deemed necessary for the original VDU to pull NWDS high. |
|
12th Oct 2020, 8:25 pm | #362 |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,294
|
Re: Mk14 vdu
NADS is not a tristate output so it shouldn’t need a pullup.
|
12th Oct 2020, 8:33 pm | #363 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
Quote:
|
|
12th Oct 2020, 8:38 pm | #364 | ||
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
Quote:
|
||
12th Oct 2020, 9:03 pm | #365 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
Good call, gents.
With a 4K7 pullup on NWDS, the PIC-VDU now correctly displays the contents of the screen RAM and the screen contents can be edited while the VDU is running, with the modified characters appearing on the screen as MEM is pressed. So Karen, it seems this is something you may have to add to your diagram (4K7 pullup to MK14 NWDS). |
12th Oct 2020, 9:28 pm | #366 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: Mk14 vdu
Oh that is good news - I have just spent all evening getting the bits together - I have everything (for a change) except a Composite socket and the DIN connector which I already put aside with some Vero to build it on - literally that last two hours looking. I am just going to order another one... then I will find it.
Funnily enough I have a pullup on NWDS on my SOC VDU as per the No 1 modification in the SOC VDU assembly instructions ( ) - I just did it on my VDU not on the MK14 as it seemed less destructive... |
12th Oct 2020, 9:30 pm | #367 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: Mk14 vdu
You can see it on my last photo in post https://www.vintage-radio.net/forum/...0&postcount=29
|
12th Oct 2020, 9:49 pm | #368 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
@!! All this time I have been thinking that pullup was on NRDS (and so I do have a pullup on NRDS on the bridge board), and yet that never actually made sense because the SOC VDU itself has a pullup (R12) on NRDS.
So the real mystery is, how has my SOC VDU been managing to work with the issue VI all this time without a pullup on NWDS? When I had the VDU connected to my original MK14 back in the day, I no doubt did have a pullup on NWDS because I used to read instructions properly in those days... |
12th Oct 2020, 10:21 pm | #369 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
Quote:
It seems strange to me that NADS is not tri-state too, I would have thought in a multi-processor system with shared busses each processor would have to drive the NADS signal but looking at the data sheet there is no mention of it disabling NADS when NENIN is high. However in the context of the MK14 its a moot point since the NADS signal is not used. |
|
12th Oct 2020, 10:25 pm | #370 | |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,294
|
Re: Mk14 vdu
Quote:
|
|
12th Oct 2020, 10:32 pm | #371 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
One last play before I put it away for the night.
These are offscreen shots of Karen's PIC-VDU displaying the two test images from a 'Slothie' issue VI MK14 replica, not from an EPROM. The images are much cleaner / lacking in artefacts than any that I have seen rendered by my real SOC VDU recently. One thing I noticed about the shots posted by Karen is that they seemed quite vertically stretched but when displayed on my monitor the aspect is as it should be. Apologies for the slight smearing on the images, they don't do the display justice at all. I didn't realise there was a fingerprint on the phone camera lens. There is a bright vertical line up the left hand side of the image, yes, ignore that for the moment. |
12th Oct 2020, 10:53 pm | #372 |
Rest in Peace
Join Date: Jul 2011
Location: Bridgnorth, Shropshire, UK.
Posts: 787
|
Re: Mk14 vdu
Hi guys,
Glad you've had some success with my PIC VDU. My expectation is that bus masters drive NRDS and NWDS using bipolar outputs when granted the bus. When not driving the bus, these drivers should go high impedance so as not to fight with another master. The role of the pull-ups, as far as I can discern, is to hold the strobes in the inactive state when no master is using the bus. As for spurious writes: the edges on PIC signals are rather faster than Mk14 era parts, and this could lead to cross-talk on the main board. The bright line down the left edge of the picture is unavoidable. When the serial port is enabled, the output immediately goes high causing light-up. No matter how quickly you try to feed the port zeros to suppress the light-up signal, some white gets through. Sorry, but it's a limitation. |
12th Oct 2020, 10:57 pm | #373 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
I think what Karen has achieved is incredible. That picture is excellent and clearly will allow software to be written for the MK14 that is compatible with the original hardware as well. It's going to be a must-have for any MK14 enthusiast!
|
12th Oct 2020, 10:59 pm | #374 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
Quote:
|
|
12th Oct 2020, 11:15 pm | #375 |
Rest in Peace
Join Date: Jul 2011
Location: Bridgnorth, Shropshire, UK.
Posts: 787
|
Re: Mk14 vdu
You could try slugging NRDS by connecting 100p to 1000p from this signal to ground.
|
12th Oct 2020, 11:32 pm | #376 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
A 4K7 pullup on NWDS - which as Tim has pointed out is the 'official' fix - is fine. Yes, I think Slothie should provide a position for one on the issue IV rev.1.3 as it is required if a VDU is used. It will be a lot nicer than having to clag one on the back of the PCB.
When I said this way back in post #21 of this thread... Quote:
And yet, miraculously, here it is. Thanks for all the incredible work you have put into this in such a short time and in such difficult circumstances. Don't go away just yet though, we'll continue playing with it and having a lot of fun with it, and if we do come across any quirks we'll need you here to sort them out. |
|
13th Oct 2020, 4:16 pm | #377 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
It would be interesting to compare how much the "Ortenview" slows the MK14 compared to the SoC original, given that the way in which the memory is accessed is slightly different (Karen buffers the memory whereas the SoC is direct-to-shift-register). From what I understand of the code I would imagine there isn't much difference because although the memory accesses are buffered they are still tied to the line rate, just happening technically earlier in the frame?
|
13th Oct 2020, 4:44 pm | #378 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
I'll try writing a simple loop of code to toggle a flag output and measure the frequency generated.
a) With VDU disabled b) With original SOC VDU c) With 'Ortonview'. That should give a relative idea of the degree of slowdown the two different VDU systems cause. |
13th Oct 2020, 7:54 pm | #379 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Mk14 vdu
OK, I had a go at that.
I'm lazy, so I just re-used the code below, which on my original 4.43MHz MK14, running without the VDU, generates about 500Hz (499.6Hz) on Flag 0. On my 'Slothie' Issue VI which has a 4.00 MHz crystal in it for the benefit of the SOC VDU, the same code generates 450Hz when the VDU (either version) is inactive With the SOC VDU attached and active (in both graphics and character modes) this falls to 415Hz. With the PIC VDU attached and active the frequency falls to a surprising and suspiciously exact 50Hz. I was conscious that this was also the mains frequency / frame rate so I scoped the flag 0 output as well - sure enough, output frequency on Flag 0 really was 50Hz (20mS between leading edges of the pulse waveform on Flag 0). I then decreased the delay values in the code for a higher initial frequency when the PIC-VDU is inactive and again measured the generated frequencies: PIC VDU disabled: 888.1Hz PIC VDU enabled: 75.1Hz Incidentally, the Flag outputs don't go anywhere on my 'bridge board' so they are not connected to any of the VDU control inputs. Code:
0F12 C4 0F13 11 ; * 0F14 8F 0F15 02 ; * 0F16 C4 0F17 01 0F18 07 0F19 08 0F1A C4 0F1B 11 ; * 0F1C 8F 0F1D 02 ; * 0F1E C4 0F1F 00 0F20 07 0F21 90 0F22 EF ; * Delay value lowered to 0x01 for second test |
13th Oct 2020, 8:02 pm | #380 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Mk14 vdu
Looks like the NENIN is being asserted a lot more then. Thats a little disappointing. Thats a much larger difference than I would have expected. Perhaps 'scoping the signals might reveal what is so different - but that could be a rabbithole
So its only getting a few instructions in per frame.... |