Pentagon 128k debugging

Вид для печати

10.02.2013, 01:45
Shockwav3

Pentagon 128k debugging

Hi there,
I'm currently in the progress of building a Pentagon 128 (Revision 2011 produced by zorel). Good things first, it already boots, accepts keyboard input and even lists betadisk directories. Now comes the bad part, after a few seconds I get junk characters on the screen, but the machine remains operational for a few more seconds til it crashes (mostly after 30-45 seconds). Sometimes it reboots, but most of the time it just hangs requiring a hard reset. Any ideas what could be wrong?

Further info:
- IC type is all 74lsxxx except for dd02, dd11 & dd12 (used 74hct93 there). I used the translation table from here: http://sblive.narod.ru/ZX-Spectrum/P...ntagon128k.htm (except for dd100 which is wrong there and should be cd4069).
- RAM chips are 41256 and have decoupling capacitors
- random decoupling capacitors have been placed on the board
- resistor and capacitor values have been used according to pentagon.pdf from here http://zxspectrum.it.omegahg.com/doc/Pentagon.pdf

Here's also a picture:
http://img855.imageshack_.us/img855/...ntagon128k.jpg

So any ideas what to try next?
10.02.2013, 02:38
null_device

Are all the errors of the board have been eliminated?

http://zx.pk.ru/showpost.php?p=535392&postcount=371

Is the board AY \ YM, it also is an error.
10.02.2013, 03:07
Shockwav3

Thanks for letting me know about that one - didn't know about that one before, fixed that, didn't help much :(
I've also fixed the missing track on dd31 and added R50 (56kohm)
10.02.2013, 03:22
null_device

Sorry if this mу advice does not help.
In the next branch, discussed the assembly and adjustment of the board, try to ask a question there.

http://zx.pk.ru/showpost.php?p=574471&postcount=983
04.03.2013, 18:29
Shockwav3

Just to let you guys know: both of my Pentagons are up and running :)
Appears the first one was suffering from a flakey PSU (I was using my PC's) and C7 being soldered in the wrong place and the second one must have had a cold solderjoint, as it magically started working once I re-soldered every chip.

So ... what are good measures for testing that everything is working alright?
I've tried loading and running a demo from disk (Alienate by Hooy-Program), loading and saving a program from/to tape, testing the INT and RAM using Test 4.30 (which also confirms the tact at 171680 cycles).
One more thing that comes to my mind would be testing a demo that uses covox sound and that would be pretty much all.
05.03.2013, 15:26
solegstar

Цитата:

Сообщение от Shockwav3

testing the INT and RAM using Test 4.30 (which also confirms the tact at 171680 cycles).

How many INT cycles? "Too short", "Too long" or..?

Цитата:

One more thing that comes to my mind would be testing a demo that uses covox sound and that would be pretty much all.

If you connect DAC (R-2R or 8bit chip) on "printer port"... only after that you can hear covox sound.
08.03.2013, 00:08
Shockwav3

Test 4.30 gives me 32-28
TestInt reports 40-36
Hidden part of the Insult Megademo shows just fine.
08.03.2013, 00:58
solegstar

Цитата:

Сообщение от Shockwav3

Test 4.30 gives me 32-28
TestInt reports 40-36
Hidden part of the Insult Megademo shows just fine.

there are no more questions) pentagon work properly. my congratulations!!! :)
10.03.2013, 08:28
ZXMAK

Цитата:

Сообщение от solegstar

there are no more questions) pentagon work properly. my congratulations!!! :)

он жаловался на то что все работает, выводит каталог диска, но через несколько секунд появляются случайные символы и затем через 30 сек крешится..
10.03.2013, 12:05
solegstar

Цитата:

Сообщение от Alexander Makeev

он жаловался на то что все работает, выводит каталог диска, но через несколько секунд появляются случайные символы и затем через 30 сек крешится..

где? этот баг кажись устранен давно. он вроде отписался, что всё работает замечательно и он погонял разные тесты, демы и т.д....:v2_conf2:
10.03.2013, 14:32
Shockwav3

The computer crashing after a few seconds has been fixed indeed (was caused by a faulty connection of C7). But there's still 1 problem remaining.
KJoy does not seem to work at all - #1f reads all 255. I could get that semi-fixed by adjusting R11 to 1MOhm, which allows me to randomly use the joystick (#1f reads 0 or 255), but this also makes the machine a good bit more unstable, since it sometimes would give me a black screen with white border and random disk access when powering up.

Using /IORQ instead of /OIRQ to drive the joystick interface allows me to use the joystick without any issue (#1f all 0 and reacting to input) but eliminates TRDOS (listing a disk gives me a *no disk* error)

Any ideas what could be wrong there?
10.03.2013, 15:31
solegstar

try to replace silicon diodes on germanium diodes or Schottky diodes with small voltage drop on P-N transition. The R11 resistor shall be 10 kOhms. On scheme (http://sblive.narod.ru/ZX-Spectrum/P...28VMGPage3.jpg) used /OIRQ signal and germanium diodes Д9.
10.03.2013, 15:36
Shockwav3

Цитата:

Сообщение от solegstar

try to replace silicon diodes on germanium diodes or Schottky diodes with small voltage drop on P-N transition. The R11 resistor shall be 10 kOhms. On scheme (http://sblive.narod.ru/ZX-Spectrum/P...28VMGPage3.jpg) used /OIRQ signal and germanium diodes Д9.

I'll try that with adjusting R11 to 10kOhms, in case that alone won't help, I'm going to drive /RD, /OIRQ and A5 through OR-gates as shown here: http://8bit.yarek.pl/hardware/zx.joystick/kjschem.gif
10.03.2013, 17:56
Shockwav3

Good news, disabling the diodes and instead using OR-Gates (74ls32/555ЛЛ1) the kempston interface works reliable. #1f outputs all 0 and reacts to joystick inputs, while floppy access works nicely :)
30.01.2014, 15:36
Shockwav3

After almost a year it's me again.

I tried to power up my Pentagon once again, sadly the video output seems to be extremely butchered. Otherwise the computer seems to work (keyboard, menu, sound, trdos).
Here's a picture of the output (3MB): http://imageshack.com/a/img11/2371/ojpv.jpg

The problem is as follows: all characters shown on screen are just very delicate shadows of what they're supposed to be, there's no color - except when entering the tape loader which causes the whole screen to flash in red and cyan.

The scart cable is correct (connecting R, G and B - SYNC on Pin20, +5V on Pin16)

Any idea where to start? I've already tried counterswapping most of the Pentagon logic chips and replaced the VT1 & VT2 transistors.
31.01.2014, 03:40
Shockwav3

Seems like this was related to the PSU. Using another one I got a proper picture and managed to load half a demo before the machine crashed.

Time to add some more decoupling caps and a properly adjusted PSU :)
04.02.2014, 07:29
zorel

The essence of the question is not very clear. To facilitate the understanding of fault recommend to make a video of what is happening on the screen to make it clearer to others, in what appears a fault.

---------- Post added at 07:58 ---------- Previous post was at 07:57 ----------

Shockwav3 Recommend to continue the discussion here http://zx.pk.ru/showthread.php?t=14277&page=43

---------- Post added at 07:59 ---------- Previous post was at 07:58 ----------

topic is more relevant here.
16.02.2014, 17:26
Shockwav3

And another patient, this time a Pentagon 1024SL 2.2

For some reason I can't get the RGB Video working, as I just get a white screen (second photo on the first image). Composite Video seems to work (upper picture). Connecting SYNC via Composite, I get a slight tint of the image as I should. Since I get a fully white picture when using the SCART cable I suppose the mode switching (pin 16) works as it should.
The second image is the pinout I used to connect a SCART connector.
I already doublechecked R42 ... R53 and VT13 ... VT15.

Any ideas what could be wrong?

---

И еще один пациент, на этот раз Пентагон 1024SL 2.2

По некоторым причинам я не могу RGB Видео работает, как я только получить белый экран (второе фото на первом изображении). Композитный видео, кажется, работает (верхний рисунок). Подключение SYNC через Composite, я получаю небольшой оттенок изображения, как я должен. Так как я получить полностью белое изображение при использовании SCART кабель Я предполагаю, что переключение режимов (контактный 16) работает как надо.
Второе изображение является распиновка я использовал для подключения разъем SCART.
Я уже перепроверили R42 ... R53 и VT13 ... VT15.

Любые идеи, что может быть не так?

https://imagizer.imageshack_.us/v2/3...90/31/csbe.jpg
https://imagizer.imageshack_.us/v2/4...0/534/n10k.jpg
11.03.2014, 14:49
Roger Wilco

For switch TV to RGB mode, you need connect pin 16 of SCART to 5 V through 180 Ohm resistor.