30th May 2020, 6:51 pm | #461 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
Ordinarily if you just strap a few control lines into certain states power it up and you have, oh, provided it with a clock the VDU will just display whatever nonsense it is seeing on the unterminated data lines. Did you remember to pick up the clock from the underside of the issue VI PCB? If not, the clock input on the VDU will be shorted to 0V at present.
Of the control lines the most important would be b13 'VDU OFF', which disables the VDU when high and enables it when low according to the manual. On the PIC 'VDU driver' PCB I have b13 [Blank screen / VDU OFF] is just tied permanently low so the VDU is enabled, but when used with an actual system it will need to start off disabled because it has a very big effect on the system speed and none of the current loading methods (cassette, uploader etc) will work when it is enabled because the system is constantly being halted while the VDU reads data from its RAM. What's therefore needed is a modest pullup on B13 which will hold the VDU -off- until such time as the user intentionally switches the 8154 PA0 output from input to output and then drives it low to turn on the VDU. A reminder, attached, of how the MK14 top side and VDU 'b' row connections line up if they are connected together one-to one. b27 (Clock into VDU) and B15 (reverse pages) are the main ones which need sorting out. |
30th May 2020, 10:19 pm | #462 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
In what respect does the 8154 appear faulty?
-Stops the system from working? -Not storing data in memory at 0x0880-0x08FF? -Port pins not working? Bearing in mind this was a previously never tested board, I've just tried my (known good) second 8154 in my Issue VI PCB now, one thing I notice is that the machine doesn't always start up properly with it in - instead of the expected '0000 00' prompt I sometimes get '.... ..' even after manual reset. If it starts up in that state, usually power cycling will get it to the correct '0000 00' prompt and it's OK from there. I can write data to the address range 0x0880-0x08FF and it 'stays', proving that there is RAM there. I also ran this short test program which generates activity (a running binary count) on both ports, A and B, of the 8154, which you can check for with a scope. This and the memory check both work on my PCB, so I don't think there is a problem with the connections from the 8154 to the bus, for example. Quote:
Last edited by SiriusHardware; 30th May 2020 at 10:47 pm. |
|
30th May 2020, 11:18 pm | #463 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
I've just realised I'm chasing wild geese here - obviously you have one 8154 which works in that machine or you could not have run the larger programs which utilise the 128bytes of RAM. But you have another, which you maybe meant to use for the 'Martin' replica?
|
31st May 2020, 10:10 am | #464 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
Yes sorry I even tried the dead 8154 in my JMP just now just in case and I get the same effect RAM or IO space just shows as 0x1F and you cannot type anything in the monitor which seems to be what happens with RAM holes - I did not bother with the port test code - If I get time I may do that as well.
I do not get the startup problem with a working 8154 on either board though. As you suspected I have a few resistor differences. Initially because of a batch of bad 4.7K resistors which I found when building the SCRUMPI (and several of which had been liberally scattered to bags of bits for many kits) I thought to test the ones I had put aside for this and they were dead so I substituted 5K 1% blues ones - then I preferred the look on the board of all Blue so stuck with them. All my 1K2 ones are actually that value Blue 1% R7-10 and R12-R15 (Top) The reminder of the connection diagram and the IO line test on the V6 you have done are useful as I had not tried that, so I was running a bit blind on using the IO chip socket as a way to patch the VDU lines although I did tap them out for continuity as part of soldering up the connector on the VDU board. As in the VDU manual I have grounded b9-b12 (using the pins on the IO socket for simplicity with jumpers to a breadboard - a sort of 8154 simulator and left b14-b17 unconnected but, I do not get the expected bitmap of the monitor. |
31st May 2020, 10:35 am | #465 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
Quote:
As to the VDU, you still didn't mention the clock? Even if you did remember to switch it over from bottom side to top side it might be worth checking to see if the issue IV PCB is outputting a clock on the appropriate pin. It's almost the only thing I can think of which would make it seem so 'dead'. That, or something wrong with the 'VDU ON' enable path into the VDU via b13. I'm assuming you have power to the VDU - is that going to it via the connections on the underside of the edge connector? Last edited by SiriusHardware; 31st May 2020 at 10:45 am. |
|
31st May 2020, 10:44 am | #466 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 schematic revisions
Quote:
Roughly what % of times does it reset incorrectly? |
|
31st May 2020, 10:44 am | #467 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
I have just put the VDU board on the Scope (I really was too tired last night and had to tidy up the desk for half hour to see the screen!)
I did remember my XOUT patch - unusually for me I reread the thread we discussed it all in as it was that that reminded me to bend up b15 as well to take over to b8 when I am ready. My machine is a bit slow today as I am running a 1TB backup over USB2 to an external drive... The bent out b27 makes a handy earth clip point as well! So the clock is nice and strong 4V in and can be seem as nice clean square waves on Pin 3 of IC4 (no output on pin 5 of that latch so the feed from IC6 via the four gate NAND on IC13 and NOT on IC12 are doing nothing). It is also an even cleaner square on Pin 14 of IC3 from the IC4 Flip Flop. Just on the off chance the scope on the Composite (rather than the TV) shows a nice 1.5V p2p normally high with 10ms low pulses every 80ms so some counters are running. |
31st May 2020, 10:53 am | #468 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
I wouldn't worry about it too much for now because:
-Tim isn't having that problem. -My DC power socket, salvaged llike Tim's is a very slack fit, needs to be replaced. -I have a current limited (1.2A) 9V regulated PSU feeding a large (4700uF) input capacitor, which might be causing slow rise times - will disconnect or reduce the size of the input capacitor to see if that makes a difference. Plenty of things to try. The main event for today is whether Tim can get the VDU running. |
31st May 2020, 10:59 am | #469 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
No pressure then! Just on that point my smoothing CAP is only 220uF as it was bright yellow - from 1982 as I have a bag of 50 and the only axial I had handy. The PSU is a 9V noisy wall wart the one I use with everything except old Spectrums as it is the wrong polarity.
|
31st May 2020, 11:06 am | #470 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
and another thing I was interested with the decoupling point on CAPS as the original had even dropped one from the BOM. I put in only the original two C3/C4 but due to the fact I know my PSU produces ringing I did add the two by the PSU in 15 and 17 - didn't really check where they were on the circuit but, assumed close to the Regulator...
|
31st May 2020, 11:06 am | #471 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 schematic revisions
Funny enough I was thinking the capacitor might be too big as I was using a wall wart style 9v power supply and they're noisy (it's a switching type) and not known for having good switch on characteristics. I'm not too worried, maybe the power to the cpu has too much inductance (ie too long a path) but it didn't happen often enough to be troublesome. That's why I asked how often it happened.
|
31st May 2020, 11:13 am | #472 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
With the 8154 out, never, with the 8154 in, quite frequently, but it almost seems to depend on the angle at which I push the DC plug into the grotty socket I've fitted. That's why I'm saying don't churn over it too much for now.
The PCB has been a resounding success so far, amazing considering you put it out 'blind' without ever having been able to test it yourself. (I know you had one of your Rev 1.0 PCBs running, but you made a lot of changes after that). |
31st May 2020, 11:18 am | #473 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
I do use a 1000uF (edit from 100!) on my JMP with the same PSU - now we did have the reset problem on that occasionally although that seemed to be very infrequent now as the main culprit was a faulty 80C95 or its 74 series equivalent as Sirius will remember
Last edited by Timbucus; 31st May 2020 at 11:18 am. Reason: technical error |
31st May 2020, 11:20 am | #474 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
On the VDU front I did run every chip through a simple function tester before use... except the character generator obviously as that was too obscure for the £30 device to have in its database.
|
31st May 2020, 11:26 am | #475 | |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
Quote:
Once I have collected enough parts I will then complete the Martin replica V as close as I can to period looking as a museum piece... working of course... |
|
31st May 2020, 11:30 am | #476 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: MK14 schematic revisions
Why thank you. I'm not too worried, I'm actually relieved that it all seems to be basically functional. I ripped up and re-did most of the layout around the 8154, display and clock circuit and moved several tracks elsewhere as well as adding the extra chip for ROM decode, so the opportunity for errors was significant!
I'm really enjoying seeing these boards come together. Really hope tim gets the VDU working since that was the main point of the update (other than fixing the back to front display!) . |
31st May 2020, 11:33 am | #477 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
Well its not just me Sirius has an original (known working) VDU as well so once the correct connector is added he will be able to test - or are you waiting for me to prove I can get it working before risking yours...
|
31st May 2020, 11:38 am | #478 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,483
|
Re: MK14 schematic revisions
Whatever can you mean?
Anyway, I am now unavoidably sidetracked for the better part of the day so I expect to see it all working by the time I surface again. Good luck! |
31st May 2020, 2:24 pm | #479 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
Well made a bit of progress - I had to ground b13 as it is not hooked to Flag1 as on the original VDU (so would be GND on power on) but to PA0 so in its floating state at 2v or so I think it was turning off the VDU - does that just stop the DMA as the previous pulses were still there? I assume so as there is no reason the frame and line pulses could not continue to ensure just a black picture when it is OFF - it just needs to stop interrupting the CPU a bit similar to the FAST and SLOW mode on the ZX81... how history repeats itself.
I now get the below two traces at 0.5us per div on x5 Mag and the built in TV Trigger (seems more like 5v p2p now). I would swear that is likely line blanks and data on the tv lines - setting the b16 invert video gets the right hand trace... seems currect So I am missing a key part of the VDU signal it seems or overdrive the TV - I will try a resistor - probably just omitting the Modulator as on the Spectrum is not suitable composite mod for the MK14 |
31st May 2020, 6:39 pm | #480 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: MK14 schematic revisions
Well success at last!!!!
After a whole afternoon chasing ghosts and going down rabbit holes of reading - all of the cz forum MK14 threads (which was useful for several things) - I was right it was the output overdriving the TV but, a 330ohm resistor just created a very distorted signal, it needed a divider really but, I tried just a resistor as Martin's photo of his replica had a resistor on it feeding the composite socket. I later found reference in the forum he used a 150ohm (as you can't see the bands in his photo) but he also telling says that modern monitors can handle a lot... he has an LCD... https://translate.google.co.uk/trans...78640%23p78640 There is a very good part of the thread where MrStaff (the originator of the PONG game video) describes building his Video card and discovering the lack of back porch (which is visible if I try to invert the screen where it looks like a ZX81 with very poor letters... that is another topic to discuss moving forward). He also simulates the output circuit: http://www.falstad.com/circuit/circuitjs.html Concluding: It is tuned, R7 = 1k5, "R9" = 47, R6 = 1k2 So I swapped miy R7 and R9 out and added R6 which is omitted (in the VDU manual as well) and bingo... the scope shows: So I thought what the hell just plug in the TV again... Then I ran around screaming for a while - putting b14 high allowed me to see graphics as well. After a bit of playing I discovered (back in Text mode) and connecting B11 to the b17 Top Page connection while Pulling b10 and b12 high through resistors and keeping b9 Low (not quite what the manual says so I was confused for a while) I get a character display of F00 at top and B00 at bottom... After trying to work out the character set by trial and error I remembered Sirius posted it once... to save people searching: Code:
@ 00 A 01 B 02 C 03 D 04 E 05 F 06 G 07 H 08 I 09 J 0A K 0B L 0C M 0D N 0E O 0F P 10 Q 11 R 12 S 13 T 14 U 15 V 16 W 17 X 18 Y 19 Z 1A [ 1B \ 1C ] 1D ^ 1E _ 1F 20 ! 21 " 22 # 23 $ 24 % 25 & 26 ' 27 ( 28 ) 29 * 2A + 2B , 2C - 2D . 2E / 2F 0 30 1 31 2 32 3 33 4 34 5 35 6 36 7 37 8 38 9 39 : 3A ; 3B < 3C = 3D > 3E ? 3F |