TS1000 with no image :_(

Discussions about Sinclair ZX80 and ZX81 Hardware
User avatar
1024MAK
Posts: 2897
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: TS1000 with no image :_(

Post by 1024MAK »

bola_dor wrote:
Sun Mar 22, 2020 11:31 pm
Visually I only see some brownish substance smeared over the soldering of the channel switch near the 7805.. may be flux?
The brown stuff is very likely left over flux. If you use IPA and a toothbrush, it should come off eventually.
bola_dor wrote:
Sun Mar 22, 2020 11:31 pm
Of these the center point and the one near the regulator are soldered together and has continuitywith the 5v 8705 output lead.. is this normal ?
In a computer where the modulator can be switched between two channels (e.g. the TS1000), there should not be any solder linking the switch solder points. You will get continuity through the switch contacts depending on which position the switch is in. The centre pin of the switch goes to the modulator. One outside pin of the switch goes to 0V/GND. The other outside pin of the switch goes to +5V. So, on one position, the switch sends +5V to the modulator, in the other position it sends 0V to the modulator. If the modulator is not being used, the switch can be desoldered and removed.

More later...

Mark

bola_dor
Posts: 122
Joined: Wed Oct 02, 2019 5:32 am

Re: TS1000 with no image :_(

Post by bola_dor »

1024MAK wrote:
Mon Mar 23, 2020 9:20 am
bola_dor wrote:
Sun Mar 22, 2020 11:31 pm
Visually I only see some brownish substance smeared over the soldering of the channel switch near the 7805.. may be flux?
The brown stuff is very likely left over flux. If you use IPA and a toothbrush, it should come off eventually.
bola_dor wrote:
Sun Mar 22, 2020 11:31 pm
Of these the center point and the one near the regulator are soldered together and has continuitywith the 5v 8705 output lead.. is this normal ?
In a computer where the modulator can be switched between two channels (e.g. the TS1000), there should not be any solder linking the switch solder points. You will get continuity through the switch contacts depending on which position the switch is in. The centre pin of the switch goes to the modulator. One outside pin of the switch goes to 0V/GND. The other outside pin of the switch goes to +5V. So, on one position, the switch sends +5V to the modulator, in the other position it sends 0V to the modulator. If the modulator is not being used, the switch can be desoldered and removed.

More later...

Mark
Could it have been a failed previous repair attempt?
The soldering looks a bit messy.. I'll compare with the other machine and try to fix it. And let you know
Thank you !!!

bola_dor
Posts: 122
Joined: Wed Oct 02, 2019 5:32 am

Re: TS1000 with no image :_(

Post by bola_dor »

OK.
The switch soldering mess was solved. I desoldered, cleaned with IPA and resoldered more carefully than the previous one did.
Now regulator runs really much cooler than before :D
But I still haven't got any video output :oops:..
Sure I have a one less problem here but still on the search..
I may try the logic probe when I get one..
What else can be faulty there :roll: .. I count two electrolytic caps. And two. Transistors.. the green little resistor shape thing :?: and ..lots of lentils and resistors.. :roll: ...

User avatar
1024MAK
Posts: 2897
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: TS1000 with no image :_(

Post by 1024MAK »

With a ZX81 / TS1000 the normal ‘stock’ faults are (ranked most likely to less likely):
  • ULA
  • SRAM
  • Z80
Sometimes the mask ROM or EPROM fails.

Where sockets are fitted, sometimes someone who does not know what they are doing fiddles and either the ROM chip or the RAM chip end up fitted in the wrong holes of the socket (ROM socket a is 28 pin type, but the ROM only has 24 pins. It’s similar with a ZX81 when fitted with a 1k byte SRAM chip.

Occasionally the 7805 will fail (it will die very quickly if the incorrect polarity supply is connected or an AC supply is connected).

The transistors are normally reliable. Failure of TR2 will result in the Z80 not getting a clock signal, then the Z80 will just sit there doing nothing useful. Best to check with an oscilloscope, logic probe, frequency counter or logic analyser first though. As equivalent replacements are hard to find (it has to be a transistor with a very fast switch off, and I don’t know of any type still in production in a TO92 package).

The electrolytic capacitors are not critical, one C5 provides the reset pulse. The other(s) are just decoupling the +9V (nominal) power rail and are mainly provided to keep expansions happy.

Mark

bola_dor
Posts: 122
Joined: Wed Oct 02, 2019 5:32 am

Re: TS1000 with no image :_(

Post by bola_dor »

ULA and Z80 were checked on the other machine as OK
RAM was ruled out by using a known working rampack.
ROM is soldered.
So the "doing nothing" theory sounds likely to me..
I think I saw somewhere some Transistor mentioned as capable to the ZX81 when looking for equivalents for my Spectrum.. I'll check what I've got..
Lab with instruments I can get access to may be closed by the quarantine as well.. and we spect this to last months as winter is still to come here.. :?
I have some Transistors that may serve to build the logic probe.. what characteristics are needed for them to work?
Thank you!!

User avatar
1024MAK
Posts: 2897
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: TS1000 with no image :_(

Post by 1024MAK »

With equivalents, keep in mind that in most applications, a general purpose transistor can be used to replace a specialist transistor used in a position where there are lesser requirements.

I’ll post up a schematic of a logic probe tonight.

Mark

bola_dor
Posts: 122
Joined: Wed Oct 02, 2019 5:32 am

Re: TS1000 with no image :_(

Post by bola_dor »

1024MAK wrote:
Tue Mar 24, 2020 5:21 pm


I’ll post up a schematic of a logic probe tonight.

Mark
That would be great!!!

I've read this topic viewtopic.php?t=1851&start=40
And I think I've made some proggress.

I've measured:
Z80pin6 to TR2b: OK
TR2b to R5: OK
R5: 330ohm OK
Z80pin6= 3.4v
TR2b= 3.1v
So I think that clock circuit and TR2 should be OK :D besides that tensions are may be a bit high.. (may be my multimeter has COVID-19)

ULA pin 14= 1v :twisted:
So is not oscilating properly :roll:

ULA pin14 to R6: OK
R6=2.4 K
R6 to TR2c: OK

And I don't know how to proggress.. I can't find components in the board :oops: .. I miss a marked board layout.. but I am almost sure has to do with ULA generating the CLK.. remember I have tested this ULA in another TS and works OK..
Any guide? Am I going OK so far? :roll:
And thank you for your help..
And mrtinb for the other topic :D

User avatar
1024MAK
Posts: 2897
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: TS1000 with no image :_(

Post by 1024MAK »

I’ve posted details of the simple transistor based logic probe here.

Due to its limitations, it has limited use on the CPU clock circuit. However by testing some of the other pins on the Z80 CPU, you can tell if the Z80 is doing something.

As the ULA on the board that you are working on is in a socket, we can perform some static tests. First disconnect the power and remove the ULA from its socket.

Now power up the board. Pin 6 on the Z80 CPU should be logic high (at or near +5V).
Use a thin piece of solid core insulated wire, with about 4 to 5 mm of the insulation stripped off (a wire from telephone or network cable is suitable) to connect ULA socket pin 14 to ULA socket pin 40. Now pin 6 on the Z80 CPU should be logic low (at or near 0V). Remove the wire. If this works, then TR2 is working at static DC voltages, so is likely to be okay.

By the way, what is the type number of the transistor that is fitted in position TR2?

Mark

bola_dor
Posts: 122
Joined: Wed Oct 02, 2019 5:32 am

Re: TS1000 with no image :_(

Post by bola_dor »

1024MAK wrote:
Wed Mar 25, 2020 1:04 pm
I’ve posted details of the simple transistor based logic probe here.

Due to its limitations, it has limited use on the CPU clock circuit. However by testing some of the other pins on the Z80 CPU, you can tell if the Z80 is doing something.

As the ULA on the board that you are working on is in a socket, we can perform some static tests. First disconnect the power and remove the ULA from its socket.

Now power up the board. Pin 6 on the Z80 CPU should be logic high (at or near +5V).
Use a thin piece of solid core insulated wire, with about 4 to 5 mm of the insulation stripped off (a wire from telephone or network cable is suitable) to connect ULA socket pin 14 to ULA socket pin 40. Now pin 6 on the Z80 CPU should be logic low (at or near 0V). Remove the wire. If this works, then TR2 is working at static DC voltages, so is likely to be okay.

By the way, what is the type number of the transistor that is fitted in position TR2?

Mark
I'll check that later
This is my TR2
Attachments
20200325_100944.jpg

User avatar
1024MAK
Posts: 2897
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: TS1000 with no image :_(

Post by 1024MAK »

Yeah, a MPS2369 is one of the few types that should be used, so it’s the correct transistor type.

Mark

Post Reply