UK101

Introduce yourself and reminisce
dave
Site Admin
Posts: 710
Joined: Tue Sep 09, 2008 5:24 am

Re: UK101

Post by dave »

Hi Denis,

Is the RAM socketed? If so, try removing the upper most 2K of RAM, and confirm that the 4K is detected correctly. Then, insert one pair of the new chips to see if 5K is detected, is should show about 4351 bytes free (I think, depending on any mods to the BASIC firmware). If that doesn't work, try the original chips and confirm proper behavior. If all is well, try the next pair of new chips in the next position. It should show 5475 bytes free. If all the new chips work, populate the last 2K, 1K at a time.

It's strange that BASIC reports "0E-04" for the memory size. I'd like to think the issue is a bad RAM chip, but I'm not sure that would fully account for this behavior. Let's see what happens.

The sync issues could be a matter of fine tuning the sync pulse widths and amplitudes. Unfortunately, I don't have a UK101 schematic. Although the 600 board should be nearly identical, I believe the UK101 output PAL instead of NTSC; this may mean a different clock, and there could potentially be some differences in where some signals are derived from the timing chain. However, there is likely a potentiometer (or perhaps wo) near the video output that may affect the amplitude of the sync pulses; try turning that one way or the other.

Best of luck,

Dave
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

Hi Dave,

I've done all the permutations/combinations with the ram chips - as you suggested.

It's now consistent in the sense that no matter what chips I put into the 6K slots, I get instability and hanging and occasional refusals to reset.

Any time I try to put more than 6K in , I gt a problem.
The RAM chips I'm using are MBM2148L-55. i have some concern about the access time on those.
If I use them in the video ram sockets , it will not boot at all.

From http://technology.niagarac.on.ca/people ... tific.html
QOUTE
It might be noted that the video RAM must be relatively fast: it must be rated at 400ns or faster (i.e. the chips in these sockets will not work if they have a "-5" suffix as this indicates a 500ns chip). Use of slower chips may not allow proper data read/write and so strange patterns may appear on the video screen.
/QUOTE

However, the spec sheet for those chips say the access time is 55ns - is that not faster then 400ns ?
Could they be too fast ?

I thought it might be the address decoding ICs . Swapped those and no progress.

I'm now inclined to the view that it might just be the sockets that are giving up the ghost and need to be changed.

I've noticed also that the one BASIC socket seems to be also rising up in the socket. Almost every time, I check, It seems to have migrated upwards and needs reseating
by pushing down on it.

I ran a simple print program tonite , looping for about two hours, with 6K installed. It ran flawlessly. No problem.. But with over 6K , It becomes unstable......

I've got the RS232 components install from way back. The output appears on J2 pin2 :) .
I'm wondering , I have a USB to TTL convertor/cable that I've used to Telnet to my media player (it's Linux) - Could I use it to connect directly to the J2 Rs232 Tx and Rx signal points
and get it to operate like that ? I did have Rs232 working way back but it was driving a modem at the time via a cable I made up for the process.

I've reverted it back to 300Baud, by undoing the switch of the C2 , C3 clocks, just to test if the cassette circuit was working. It is in the sense that it will save to cassette, but no success with reloading at this point. I'm sure it's time to move to more efficient methods to store /retrieve programs :)

With the end of Summer and long dark evenings ahead , I can see this as a project that will last all winter.

Regards

Denis












Intermittent faults are the worst scenario......
MK14HAK
Posts: 356
Joined: Wed Mar 16, 2011 1:49 am
Location: New Zealand

Re: UK101

Post by MK14HAK »

Denis
I think your best to fix that sync problem rather than go the terminal way. Have you got a monitor you could use or a TV with an AV input and bypass the modulator and use the direct video out on J1.You dont need an emitter follower it will drive directly. I dont think you have a sync problem if one TV is stable. Have you got a clean 5v rail as AC ripple here can upset modulators (and memory). Else check for dry joints around vert sync cct. Did you tweak the level pot as Dave suggested. Mine is set full clockwise.
As for your memory issue Im confused. You exchanged the top for the bottom pair and got it to boot. That should give you 7K not 6K. Have you found those top two chips faulty or was it broken legs? Did you have others faulty as well or has it just been contacts/broken legs.?
The best place to put suspect ram is in the first 1K row removing the second K or the rest of the ram entirely, as then pg 0 and stack get a good workout at boot up and running a small BASIC program with variables also stored in that first 1K. You could use a poke peek memory test program as a further test. As you found the system generally doesnt boot if theres a problem in the first 1K of ram.
You should test continuity to all pins on chips in those top two or four sockets. You could also try putting a know good pair up there and linking the CS ( isolate the pins )line from the first 1K up to that pair removing the first 1K mem pair and see if it boots !

Good luck Mike
600RevB:16K,2MHz,64x32,470,CEGMON
SuperKit:502,540B,542B,CEGMON, 8" and 5" FDDs
Cards:PE IO,6522 D-A-D, AY3-8910,ProgramGraphics,Color,UK101
WIP:HexDOS,FDD Emulator
dave
Site Admin
Posts: 710
Joined: Tue Sep 09, 2008 5:24 am

Re: UK101

Post by dave »

Denis,

Your findings are promising.

With respect to the memory, I think you've established that that the memory chips should be fine.

When you insert chips in the higher 2K, have you tried limiting the memory size to 6K (6144 bytes) on cold start, to see if your system is stable? Also, is the system stable when running the OS65V monitor ("m" option)? If so, then you're looking at most likely a bad connection for one of those chips, likely one of the two chips mapping right above the 6K you have installed now. Something like a bad socket, cold solder joint, or broken trace.

If, however, adding RAM in those sockets makes the whole system unstable, even in the 65V monitor, and with BASIC memory size limited to 6K or less, then I'd be more suspicious of something like a /CS line shorted low.

I also second Mikes suggestion to check to make sure the +5V rail is good quality.

With respect to the serial connection, if you hook the USB<>TTL converter directly to the UART inputs, you should be good as gold. You should be able to handle speeds up to 9600 baud easily; just make sure, when loading BASIC programs, that you add lots of padding characters at the beginning of each line, to allow the screen output code to scroll (OR, disable screen echoing of the incoming code with crtl-o).

Best of luck. Keep us posted!

Dave
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

Hi Dave, Mike,
Did you tweak the level pot as Dave suggested. Mine is set full clockwise.
Ah ! The UK101 did not have the variable resistors for adjustment that the Challenger had (No R57,R58).

It's working on my very small TV - so the sync problem is not an issue for the moment.
With respect to the serial connection, if you hook the USB<>TTL converter directly to the UART inputs, you should be good as gold. You should be able to handle speeds up to 9600 baud easily; just make sure, when loading BASIC programs, that you add lots of padding characters at the beginning of each line, to allow the screen output code to scroll (OR, disable screen echoing of the incoming code with crtl-o).
Got that working perfectly :) . wired directly to the ACIA Rx Tx and Ground. I can transfer text files from Hyperterm no bother. The software file I do have were downloaded from the web as a zip file. Unfortunately, all the files seem to lack CR/LF after each line. With a bit of editing, it works fine - of course I had to do the delay after character and line fix as you suggested. Maybe OSIWEB has more software correctly formatted sofrware to download ?
You should test continuity to all pins on chips in those top two or four sockets. You could also try putting a know good pair up there and linking the CS ( isolate the pins )line from the first 1K up to that pair removing the first 1K mem pair and see if it boots !
I've done that and put good ram in all possible places - It comes down to a definite predictable reproducable situation wher ANY IC in the 7 th pair will cause a problem.
I did think I could swap the /CS lines , just to see it the fault would follow the /CS line :), but it involves a hack of the tracks or bending out of pins.
If, however, adding RAM in those sockets makes the whole system unstable, even in the 65V monitor, and with BASIC memory size limited to 6K or less, then I'd be more suspicious of something like a /CS line shorted low.
I wish I could report some consistency here :( . Sometimes, the monitor ('M' ) will respond and sometimes not. Reseating (some/all) the 24 pins ICs seems to always bring it back to life.

I have considered all your suggestions about the Ram issue and decided that I will go ahead and pull upper 6 the Ram sockets and replace them. I will also replace the 6 24 pin sockets for the ACIA, Monitor, + 4 BAs Ics. They seem to need reseating every time I touch the PCB and clearly many of the failures are due to the BAS chips not responding - (No response to 'C' , but will enter Machine Monitor.

I did clean all the IC legs and spray contact cleaner into the sockets, but I reckon they are gone weak and are probably not up to the task of securely holding the pins any more.
I also second Mikes suggestion to check to make sure the +5V rail is good quality.
I replaced the elec Capacitors and all the decoupling caps. (I've not replaced all the caps - some timing components in the video/cassette areas. Maybe I should do that also.)
The 5V seems stable.

Anyway , I'm pleased to have got so for to date.............. more to do

Thanks again.


Denis
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

Ah! The good news :)

I've replaced 6 Ram sockets (upper 6)
Also replaced the 24 pin sockets for the ACIA, Monitor, and Basic ROMS.

I installed 8 K and it works perfectly :)

I removed the sockets by breaking the frame and lifting them off , leaving the pins to be desoldered individually.
I did notice that some of the pins were very loose and may not have been contacting very well in use. (where does all the dirt come from ?)
The ROMs are now all firmly held in place and that must be helping.

I downloaded a MC program (memory test) and was able to load it via the TTL/RS232 cable from Dell notebook to Uk101.
It took a while to get it to load. I was getting corruption a while i realised that the errors were the same every time. I examined the downloaded file - no errors there !
Then I realised that I had CEGMON and the free memory is different to the standard Superboard/UK101. Free mem starts at .0235 - not .0200 where the mem test was written to load. I edited the mem test start address to .0235 and it loaded perfectly.

However it will not respond to .0235G - This must be due to the altered vectors in Cegmon. i wish I had a way to disassemble , so i could edit to run.

In the good old days I had the extended monitor on tape - think i still have it but at it's age -no way could it be loaded :(

Anyway, I now reasonably certain that it is now functioning correctly.
Attachments
bad_pins.jpg
bad_pins.jpg (71.52 KiB) Viewed 16554 times
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

Forgot the most important bit :)
Tadaaa !!

Hmmm.....Difficult to get a good screen pic. Must study photography :)

D.
Attachments
UK101_IntroScreen.jpg
UK101_IntroScreen.jpg (84.81 KiB) Viewed 16551 times
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

I've solved the video problem I was having -the bad pic on my larger TV.
I'm now driving it from the video output, rather then through the modulator.
I changed the cable connecting the video o/p to the TV and it made all the difference
- bad cable was the problem ! Previously the video output was missing, which I suppose had to be a bad cable, now
that I think about it :)

The video signal is now steady. I occasionally get a jitter at the top of the first video line. It jitters left/right.
I still need to adjust the resistor/capacitor combination that affects the display - to move it down
a bit - currently the 1st line is partly off screen. That should be no problem now.
Maybe that will affect the jitter too ?

(I should now be able to get a better photo of the intro screen too :)

And then , while I can load from a PC, I do want to get the tape loading/saving working too , just cause it's there...

Denis
dave
Site Admin
Posts: 710
Joined: Tue Sep 09, 2008 5:24 am

Re: UK101

Post by dave »

Dennis,

That's fantastic progress! Some pictures would be awesome.

So, you're driving the TV through a built-in RF modulator? I think you'll get a MUCH better result bypassing the modulator and using the video out directly, if you can find a TV that has a video input. Even better, if you can get an old Apple or Commodore monitor, inexpensive security monitor, etc., you should have excellent results. The SBII provided the video out directly, so I'd bet the UK101 brings it out to a pin somewhere. I will have to hunt down some schematics.

Best regards,

Dave
DenisM
Posts: 12
Joined: Mon Aug 19, 2013 7:44 pm
Location: Cork, Ireland
Contact:

Re: UK101

Post by DenisM »

Hi Dave,

I'm now driving the AV input of my TV through pin 12 of J2 and the picture is rock solid. I hand picked a resistor to move the pic down slightly (the top line was partly hidden).
The Uk101 does not have a variable resistor R57 to do this and it's a little tricky to fit a variable in the confined space when the track layout do not support it.

I was delighted with the steady video and moved on to getting the rs232 working -I'm now using Martin Ward's PERL script to convert my old cassette tapes to text and can load them on using Hyperterm. I've fitted a DB9 connector to the TTL/USB convertor cable and it works perfectly.

http://www.cse.dmu.ac.uk/~mward/martin/software/#UK101


I'm in process of making a new back panel, so the case top is not yet back in place......... more to do.....

However I'm really pleased to be able to access code that I thought had been lost forever - none of it significant - it's value is purely sentimental :)

The pics are not hires. Must try to put them on flicker or some such. I'm still not sure about uploading images here - in terms of size etc.

Denis
Attachments
The case is made of ABS plastic, and only gets it 's rigidity when the top is bolted on. This makes the whole unit fragile, as the case bottom and thus the pcb can flex when moved.
The case is made of ABS plastic, and only gets it 's rigidity when the top is bolted on. This makes the whole unit fragile, as the case bottom and thus the pcb can flex when moved.
Uk_Case.jpg (19.2 KiB) Viewed 16480 times
I got the convertor on ebay , some time ago - intended to access a port in my media player, not knowing it would come in handy for this task :)
I got the convertor on ebay , some time ago - intended to access a port in my media player, not knowing it would come in handy for this task :)
TTL_USB_cable.jpg (19.09 KiB) Viewed 16480 times
This is the screen taken with my wife's iPhone. Much better then I could get on a Cannon D400 SLR. It seems to manage the bounced back flash a bit better :)
This is the screen taken with my wife's iPhone. Much better then I could get on a Cannon D400 SLR. It seems to manage the bounced back flash a bit better :)
Intro_UK101.jpg (18.46 KiB) Viewed 16480 times
Post Reply