@andy150677: it works "mostly fine" in 48k mode, but still crashes there occasionally. TRDOS usually crashes within seconds. I'll try to get a RAM Test ROM soon. Sadly I only have 41256 RAMs to test, but will try to order 4164 ones if the problem persists.

Scheme for connecting the Pentagon to my TV is just the ordinary one, R, G, B on the corresponding pins + Sync on pin16 (blanking), ground for r_ground, g_ground, b_ground, shield and blanking (pin18). Didn't use any amplification as shown in pentagon.pdf (470ohm -> kt315b/bc547 -> 75ohm -> scart socket for R, G, B & Sync), since it didn't make a difference on my TV

Shockwav3, And if you connect to another TV, the picture is the same?
There is an oscilloscope or frequency? Need to see what's on the input - output meter DD11, DD12,

If it's a manual for the assembly, transfer to translate.google.com.ua

zorel, нашел фотки, которые при сборке делал. Там мало что изменилось - только кондеры, которые навесным напаяны, по-человечески припаял, и для второй линейки ОЗУ блокировочных поставил на обратной стороне. В любом случае - не стало ни лучше, ни хуже.
http://i54.fastpic.ru/big/2013/0211/...30da421e2f.jpg
http://i54.fastpic.ru/big/2013/0211/...2ed41250a3.jpg
http://i54.fastpic.ru/big/2013/0211/...6ea8a3f0b8.jpg

сигнал на +5 посажен через кондер.

Оптическая иллюзия :)
На самом деле это блокировочный кондер, он впаян в шины питания, переходное отверстие посередине не затронуто.

Turned out I had C7 connected to an incorrect via (PCB print is slightly confusing there), which seems to have fixed the random RAM garbage appearing on screen (yay). Sadly the machine still seems to be slightly unstable (crashes in TRDOS after 5-10 seconds and after roughly a minute in 48k basic) and the SYNC still isn't working correctly, but we're getting there :)

Need to check the meter DD11, DD12, frequency counter or oscilloscope. And check out the Pentagon on a second TV.

Just as a little follow up: I've built my second Pentagon and it has exactly the same issue :(
I have pretty much tripple checked everything by now and used the Pentagon with another tv (even tho the tube TV I first used works fine with my Pentagon1024SL) ... I'm starting to get a little frustrated since I don't have access to an oscilloscope/frequency counter.

Could my failing RGB signal come due to me using a 74ls37 (555ЛА12) instead of a "real ttl" 7437 155ЛА12 for D72 ... wikipedia said they "should" be compatible?

try quartz on 14.3 MHz, with it the vertical sync has to be 50 Hz. if you use quartz on 14 MHz, vertical sync is 48.9 Hz. Probably in it a problem.

Okay, I got my RGB issue sorted out :)
Silly me used a Scart Female plug to connect a 1:1 scart cable which meant I had to connect RGB_Sync to Pin 19 instead of 20 :v2_rolley

And another one ...
I've started soldering my second Pentagon, sadly it won't even start. For the ease of debugging I'm not using a ROM. Various decoupling capacitors have been placed on the board already. I'm roughly seeing the bands I should get, but they're randomly changing color between red and blue character wise.
ICs 37, 39, 38, 40, 41, 42 & 46 have been checked already. RAM has been verified as good and the problem remains consistent between 2 different sets I'm using.

A video of the problem can be found here:
http://youtu.be/mipe27poaZY

Good news everyone :)

The Pentagon is working - it loads programs and plays music :)
I'm still having 2 minor issues with the video (minor "ghosting" on my LCD TV) and for some reason I can only access drive a: in TRDOS, but for now I'm happy it is working nicely.

short note about decoupling: apart from the RAM chips, I didn't have to decouple any additional chips.


---
EDIT: For some reason I can't find out what's wrong with the "ghosting" which seems to apply to both TVs I'm using. It always looks like some attributes/characters are embosed.
Here's a picture of how it looks like with the "bugs" being highlighted ... sadly the picture isn't every good since my camera sucks.
http://img40.imageshack.us/img40/637...buteembose.jpg
Scart is connected at a minimal approach: R, G, B & Sync on their corresponding pins + ground and 3V on Pin16, no 75Ohm resistors or transistors (tried that which only seems to make it worse)

Could you help me with the Kempston interface on the daughterboard?
I can't make out for sure where the pins for the directions and fire are.

http://img5.imageshack.us/img5/9717/vfghf.jpg

Part of the circuit which implements the interface joystick. There are two options. Fire - or 8 output chips, or 11.
Not rule out that wiring is faulty (or just very specific divorced).

Shockwav3, my advice, just load the game and run with selected kempston-joystick option, then try to connect unknown pins IDC10 to GND and see what happen on screen.

Yes, I've already checked the board against the scheme. While the Pins for left (13) and right (11) are obvious, the other pins connect to legs 4, 6 and 8 on the chip :(

it isn't important as inputs are connected, the main thing that was compliance an input output of AP3 and were connected to appropriate bits of the data bus. 1533AP3 (SN74ALS240) is 8-bit data bus tranciever with inverted outputs.

Небельмеса

I have checked wires on the PCB:

I've soldered a D-Sub 9 port as described, but for some reason it does not work.
When using "10 PRINT IN 31;:GO TO 10" in basic for testing I get a constant value of 255. It doesn't matter if I have a joystick connected or not.

Symbols of diodes VD3 - VD5 are misdirected on the picture of this PCB for 1N4148 diodes. Cathodes of all diodes must be connected to pins 1, 19 of DD7:

I think I read that somewhere before, which is why they're connected opposite of the pcb prints on my board. 1n4148

EDIT3: We're getting there, adjusting the R11 with the faulty value of 1k-Ohm to 1m-Ohm allows me to access TRDOS and _randomly_ the Joystick (Test program returns values like 00000255255255, etc.)
EDIT4: It's working :) Bypassing the diodes and sending /OIRQ, /RD and A5 through OR-Gates as shown here (http://8bit.yarek.pl/hardware/zx.joystick/kjschem.gif) the kempston Joystick works along with TRDOS access.
Maybe that's something that could be fixed on the next Revision of the board?

In case anyone of you wants to build the Turbo extension on breadboard, I've made a soldering plan for myself, which I thought I could share with everyone.
Leftmost ICs are TM2/74 while the right one is ЛА3/00, other values are according to the scheme (1n4148, 1k Ohm, 100pF). The board itself is designed to be as compact as possible with the wires going over the top layer being as short as possible

Could anyone of you also be so nice and translate what the other text says?

EDIT: appears the soldering plan I uploaded was resized ... here's a bigger version: http://img854.imageshack.us/img854/3409/penturbo.png

Мля! НЕБЕЛЬМЕСА, В НАТУРЕ! :)

Shockwav3 , whether you can attached the original schematic diagram?

Sure ... it's included in pentagon.pdf tho.

http://img543.imageshack.us/img543/9972/turbos.png

Shockwav3 попросил перевести текст со схемы Турбо доработки Пентагона 128.
Modifications on the computer's PCB.
1. Disconnect DD5-2 and DD13-28 from their circuits.
2. Disconnect DD15-1 from the +5V circuit. Connect DD15-1 with the ~RFS circuit (DD13-28).
3. Disconnect DD13-6 from DD1-8, but keep the connection of DD13-6 with the pull-up resistor.

Опять здравствуйте, теперь вот вопрос по контроллеру дисковода. Напомню, плата от Zorel`а, крайней серии. Такая картина. Есть три 3.5 дисковода, два новых, один начала 90хх, все проверены на 286'м. На прямом шлейфе каждый определяется диском "С", т.е. A> *"C", затем
O.K.
C>
На этом признаки жизни заканчиваются, при команде FORMAT короткое движение башки и полное зависание, только двигатель крутится. При команде LIST - стандартная ошибка, abort, ignore и т.п.
На перевернутом шлейфе определяется, как диск B, точнее только жужжит башкой, O.K. нету.
ВГ93 менял, ЛП9, ЛП11 рабочие. Куда курить надо, где методики настройки почитать? BDI впервые налаживаю.
Спасибо =)

UA3GBV, если есть в наличии дисковод 5.25 (всё равно какой, хоть 1.2Мб лишь бы у него была установлена резисторная матрица), для начала экспериментов попробуйте его посадить на один шлейф с дисководом 3.5, только выставьте его (5.25), как А, если всё заработает, то нужна подтяжка к +5В, через 1кОм резисторную матрицу, выходных буферов с "открытым коллектором" контроллера дисковода.

Помоему 3.5 дисководы изначально подтянуты к +5в...

В перую очередь - заклеить "окошко" плотности на дискете (если дискеты НD).

Это странно. С ИД4(D98) и ТМ9D88) всё в порядке? Не попутал местами ЛН3 и ЛП9?

У меня такая же история была со старой платой пентагона. Имела место ошибка разводки - шины "выбора" дисководов "В" и "С" перепутаны местами.

Спасибо, за ответы, но как-то дело не продвинулось. Дисковод определяется нормально, любой, при попытке чтения - закономерная ошибка, глохнет опять-таки при форматировании. Отсюда вопрос - а на каких микросхемах можно отследить процесс форматирования? Он аналогичен процессу записи, т.е. задействется Д96(ИР16)? На н10D96 вечный ноль, на 2нД88 (ТМ9) - 1 при бездействии, 0 при любом обращении к дисководу, на 5,7,10нД88 1 всегда. Перед запайкой все микрухи проверялись, естессно )

Форматирование это запись разметки диска, так что D96 должна работать и при этом процессе на 10н D96 должен быть сигнал.
Просмотри сигналы до D96 и проверь D94 (может замкнуть выход D96) Так же просмотр этой дороги на непротрав.

UA3GBV, логику работы бета диска, и методику проверки - можно почитать в книге периферия своими руками. Довольно подробно и детально описана проверка работы узлов типового контроллера BDI.

На Д96 приходят синхра на 9-ю ногу и сразу после ввода команды формат - две пачки импульсов на 6-ю ногу. Но на выходе тишина... Может и дохлая, сил выпаивать уже нету...:v2_dizzy_messed:

А может и выход глушит ЛН3 или замыкание?

---------- Post added at 21:01 ---------- Previous post was at 20:59 ----------

Попробуй аккуратненько 10ю ногу D96 поднять над платой и проверить выход.

lisica, ненене, все буферные микросхемы на панельках стоят, ЛН3 в том числе. Проверял и без нее. Замыканий нет, выпаивания не избежать.