ZX Spectrum +2 booting problem

For experts to discuss very technical stuff and newbies to ask why the Spectrum they bought off ebay doesn't work.
CapSmasher
Berk
Posts: 16
Joined: Mon Oct 22, 2018 8:53 am

Re: ZX Spectrum +2 booting problem

Post by CapSmasher » Sun Oct 28, 2018 7:41 pm

Hey Mark, thank you for your replies

I checked the R79 and R78 outside circuit and indeed they are good, thank you for the tip!
R78 is correct at least on my board - and is indeed connected with C27

TR9 can be burned if is reversed (sorry my noobines)? I initially intalled it with the emitter on the 9v... Saw that on a video on youtube that the emitter had to be connected to 9v - I'm confused now... I guess if the diodes and resistors are OK the coil may be damaged? Is there the possibility that the caps to be to blame?
I will try to investigate that R52 because it seems is tied with D19 like in the power schematic so it must be that.

You said that TR9 is mainly feeding the TEA2000 - that means if the TR8 is working and in charge with the +5V, I still should need to have +5V on the chips, no? But I have only 4.7V ... Let's assume that all circuit is OK except the coil and TR9, I still should have sound when pressing the keys throught the headphone jack, if the rest of the board is OK?

Do you think I can replace the coil with one from a 48k ZX Spectrum? Does it have a specific name/specifications, or some commercialy alternatives?

Alternatively (and maybe more extreme), can I bypass all this power circuit and feed the chips directly to see at least that the rest of the components are good?

Thank you
0 x

User avatar
1024MAK
Manic Miner
Posts: 491
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: ZX Spectrum +2 booting problem

Post by 1024MAK » Sun Oct 28, 2018 10:39 pm

The 7805 voltage regulator supplies the +5V Rail. If the voltage on the input to the 7805 voltage regulator is at least 7.5V, then the output should be within +5V +/-5% at worst. So 4.75V to 5.25V. Most 7805 voltage regulators manage better than +/-5% though.

If the regulator and it’s heatsink are hot, and the output voltage is lower than expected, measure the current on the +9V input.
It may be that if the +5V line went above 6V during the time that the old regulator was faulty, it may have damaged some of the other chips. So see if any are hot (apart from the ULA which runs hot anyway). The Z80 may be warm, but should not be hot. The other chips should be cool or slightly warm.

Whatever the actual voltage from the 7805 voltage regulator, there will be a gradual loss across the board. So a little bit of voltage drop on the +5V rail is to be expected.

I’m confused by your reference to D19. The 2.2k ohm resistor for the -5V line should be between diode D37 (1N4148) / capacitor C22 (47uF) and diode D18 / capacitor C24 (10uF) according to the schematic I’m looking at. Or is this another error on the schematic?

Normally, the transistor (TR9 in this case) dies before the coil becomes damaged. But occasionally the coil may overheat. There are no know suppliers of new coils. As I said earlier, it is believed that it is the same as those used on 16K / 48K boards. A very limited number of second hand coils may be available from a repairer. But the normal fix for a damaged coil, is to remove it from the board. Then very carefully make notes and take photos as you unwind the old enamelled wire. You have to record where the wire starts, which direction it is wound in, and the number of turns, and where it finishes. For both coils of wire. Then obtain some new enamelled wire of the exact same diameter. Then wind the new wire on the former following the notes that were made earlier to the letter. Before soldering the enamelled wire, scrape the enamel off with a sharp knife. Then tin it.

For test purposes, you can remove / leave out TR9. Then using a current limited, but regulated +12V power supply, you can feed +12V to the cathode of diode D20. Remember to also connect the 0V line to the Spectrums 0V/GND line.

It is also possible to use an external current limited, but regulated +5V power supply. I recommend that you remove the 7805 voltage regulator and then feed in via the appropriate solder pads. The 7805 voltage regulators don’t like voltage on their outputs if there is no voltage (or a lower voltage) on their inputs.

If you use the RGB video outputs and the display you are using can display the RGB video inputs without needing active function switching inputs (RGB monitor or TV with a SCART socket that can be switched to RGB inputs manually), then the missing +12V is not a problem.

I can’t remember if the keys make a keyboard click when the ZX Spectrum +2 powers up, because it goes to the menu. But certainly once in BASIC the keys should produce the keyboard click. How are you listening to the sound? Only if the video output is not functioning due to the +12V line not being operational, there will be no sound on the UHF TV output, and no sound out on the composite video output (on the RGB/video DIN socket).

Mark
0 x

User avatar
Ast A. Moore
Dynamite Dan
Posts: 1167
Joined: Mon Nov 13, 2017 3:16 pm

Re: ZX Spectrum +2 booting problem

Post by Ast A. Moore » Sun Oct 28, 2018 10:50 pm

1024MAK wrote:
Sun Oct 28, 2018 10:39 pm
I can’t remember if the keys make a keyboard click when the ZX Spectrum +2 powers up, because it goes to the menu.
Yup, they do.
0 x
Every man should plant a tree, build a house, and write a ZX Spectrum game.

Author of A Yankee in Iraq, a 50 fps shoot-’em-up—the first game to utilize the floating bus on the +2A/+3,
and zasm Z80 Assembler syntax highlighter.

CapSmasher
Berk
Posts: 16
Joined: Mon Oct 22, 2018 8:53 am

Re: ZX Spectrum +2 booting problem

Post by CapSmasher » Mon Oct 29, 2018 1:22 pm

1024MAK wrote:
Sun Oct 28, 2018 10:39 pm
The 7805 voltage regulator supplies the +5V Rail. If the voltage on the input to the 7805 voltage regulator is at least 7.5V, then the output should be within +5V +/-5% at worst. So 4.75V to 5.25V. Most 7805 voltage regulators manage better than +/-5% though.

If the regulator and it’s heatsink are hot, and the output voltage is lower than expected, measure the current on the +9V input.
It may be that if the +5V line went above 6V during the time that the old regulator was faulty, it may have damaged some of the other chips. So see if any are hot (apart from the ULA which runs hot anyway). The Z80 may be warm, but should not be hot. The other chips should be cool or slightly warm.

Whatever the actual voltage from the 7805 voltage regulator, there will be a gradual loss across the board. So a little bit of voltage drop on the +5V rail is to be expected.

I’m confused by your reference to D19. The 2.2k ohm resistor for the -5V line should be between diode D37 (1N4148) / capacitor C22 (47uF) and diode D18 / capacitor C24 (10uF) according to the schematic I’m looking at. Or is this another error on the schematic?

Normally, the transistor (TR9 in this case) dies before the coil becomes damaged. But occasionally the coil may overheat. There are no know suppliers of new coils. As I said earlier, it is believed that it is the same as those used on 16K / 48K boards. A very limited number of second hand coils may be available from a repairer. But the normal fix for a damaged coil, is to remove it from the board. Then very carefully make notes and take photos as you unwind the old enamelled wire. You have to record where the wire starts, which direction it is wound in, and the number of turns, and where it finishes. For both coils of wire. Then obtain some new enamelled wire of the exact same diameter. Then wind the new wire on the former following the notes that were made earlier to the letter. Before soldering the enamelled wire, scrape the enamel off with a sharp knife. Then tin it.

For test purposes, you can remove / leave out TR9. Then using a current limited, but regulated +12V power supply, you can feed +12V to the cathode of diode D20. Remember to also connect the 0V line to the Spectrums 0V/GND line.

It is also possible to use an external current limited, but regulated +5V power supply. I recommend that you remove the 7805 voltage regulator and then feed in via the appropriate solder pads. The 7805 voltage regulators don’t like voltage on their outputs if there is no voltage (or a lower voltage) on their inputs.

If you use the RGB video outputs and the display you are using can display the RGB video inputs without needing active function switching inputs (RGB monitor or TV with a SCART socket that can be switched to RGB inputs manually), then the missing +12V is not a problem.

I can’t remember if the keys make a keyboard click when the ZX Spectrum +2 powers up, because it goes to the menu. But certainly once in BASIC the keys should produce the keyboard click. How are you listening to the sound? Only if the video output is not functioning due to the +12V line not being operational, there will be no sound on the UHF TV output, and no sound out on the composite video output (on the RGB/video DIN socket).

Mark
Hi Mark,

Thank you for the info!

The voltage regulator last time I checked had correct values of 9v input and 5v output. The problem is that even with the 5v output I had only 4.7v on the RAM chips.

My board is issue1 and the schematics are for Issue3, maybe some resistors/capacitors changed a bit. Didn't find any schematic for Issue1..

OK so in theory I should be able to power the ZX spectrum +2 like this - I've re-written the info from you from what I understood so to be sure I will not mess up:

1 - desolder the TR9 (just to prevent frying it because is my last piece)

2 - connect +12v to the negative (cathode) pole of the diode D20 and the GND from this power source to the Spectrums' GND.

Here I will need just a bit of clarification... Do I need to desolder the D20 cathode then wiring it with 12v, or I'll wire it directly on the solder pad of that cathode? The GND from the +12v power supply will connect directly to the ground of the power supply jack solder pad? Will I need this +12v if I will use only the analog TV cable?

3 - remove the 7805 voltage regulator

4 - connect +5v and GND (from external power supply) to the +5v and GND pin where the 7805 was attached

5 - check ram voltages and see what chips get warm

6 - try to search for some signal on TV

7 - if no image on screen, I will connect some headphones and check if there is some sound while pressing keys (but only if I will have the +12v line functional?)

Thank you
0 x

CapSmasher
Berk
Posts: 16
Joined: Mon Oct 22, 2018 8:53 am

Re: ZX Spectrum +2 booting problem

Post by CapSmasher » Tue Oct 30, 2018 10:08 pm

In the meantime I tested the coil resistance:

Primary winding (connected to the transistor collector) 0.05 ohms.
Secondary winding (connected to the transistor base) 0.05 ohms.
Primary winding to secondary winding: greater than 20M ohms (it kept growing, 20M is maximum on my multimeter)

Do you think are OK these values?
I also checked the resistors, seems all OK - didn't find the R52 but don't think is that important since is on -5V

Thank you

LATER EDIT:
So while I was looking at the wonderful speccy board that I have (since couldn't power it on) I just found this capacitor (C41) which paradoxical it has some connection with +12.. It looks a bit bulky and like there is some green leaking on its wire? Could this been causing the TR9 (ZTX653) sudden death?


Image

Image
0 x

User avatar
1024MAK
Manic Miner
Posts: 491
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: ZX Spectrum +2 booting problem

Post by 1024MAK » Wed Oct 31, 2018 3:55 pm

Electrolytic capacitors can cause all sorts of problems. If you are good at soldering and desoldering, you should indeed replace any that look like they are leaking, or may leak.

The resistance tests on the coil look good. Resistance tests alone can’t prove that it is good, but it is unlikely to be faulty.

Mark
0 x

CapSmasher
Berk
Posts: 16
Joined: Mon Oct 22, 2018 8:53 am

Re: ZX Spectrum +2 booting problem

Post by CapSmasher » Wed Oct 31, 2018 3:59 pm

1024MAK wrote:
Wed Oct 31, 2018 3:55 pm
Electrolytic capacitors can cause all sorts of problems. If you are good at soldering and desoldering, you should indeed replace any that look like they are leaking, or may leak.

Mark
Hey Mark,

Thank you for the reply, can you please tell me if the steps from this previous post:
viewtopic.php?f=22&t=1062&start=10#p15314

..are correct? I may try first to power the Spectrum from an external source, just to check the status of the chips

Thank you!
0 x

User avatar
1024MAK
Manic Miner
Posts: 491
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: ZX Spectrum +2 booting problem

Post by 1024MAK » Wed Oct 31, 2018 4:17 pm

CapSmasher wrote:
Mon Oct 29, 2018 1:22 pm
1 - desolder the TR9 (just to prevent frying it because is my last piece)

2 - connect +12v to the negative (cathode) pole of the diode D20 and the GND from this power source to the Spectrums' GND.

Here I will need just a bit of clarification... Do I need to desolder the D20 cathode then wiring it with 12v, or I'll wire it directly on the solder pad of that cathode? The GND from the +12v power supply will connect directly to the ground of the power supply jack solder pad? Will I need this +12v if I will use only the analog TV cable?

3 - remove the 7805 voltage regulator

4 - connect +5v and GND (from external power supply) to the +5v and GND pin where the 7805 was attached

5 - check ram voltages and see what chips get warm

6 - try to search for some signal on TV

7 - if no image on screen, I will connect some headphones and check if there is some sound while pressing keys (but only if I will have the +12v line functional?)
Yes, but note the following:-
Solder the +12V to the pad of the cathode of D20, or any other convenient point on that track. So you could also connect to the positive (+) pad of capacitor C41 (100uF).

If the +12V AND the +5V are being supplied by the same PSU and already share a common 0V/GND, you don’t need two different 0V/GND connections. If however the +12V and +5V do not share a common 0V/GND, OR you are using different PSUs, then two 0V/GND leads (wires) are needed.

You must use PSUs where they limit the current in the event of a fault. Otherwise you risk damaging the PCB tracks if there is a short circuit somewhere on the board. If possible, I recommend no more than 1.2A on the +5V line. And no more than 100mA on the +12V line.

If different PSUs are used, power up the +5V supply first, then the +12V supply.

The audio output from the ZX Spectrum +2 will need to be amplified if you are going to use headphones. This does not use the +12V rail. Only the composite video (on the RGB/monitor DIN socket) and the TV UHF outputs need the +12V for the TV audio to work (it is combined with the video signal hence the need for the +12V to be working).

Good luck.

Mark
0 x

User avatar
1024MAK
Manic Miner
Posts: 491
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: ZX Spectrum +2 booting problem

Post by 1024MAK » Wed Oct 31, 2018 4:21 pm

Oh, by the way, to reduce stress on the PCB, when removing a component where you are not intending on reusing it (such as an electrolytic capacitor), it’s better to note down the holes it uses and the polarity (or take photos), then cut it’s legs, then desolder the legs, then fit the new one.

Mark
0 x

CapSmasher
Berk
Posts: 16
Joined: Mon Oct 22, 2018 8:53 am

Re: ZX Spectrum +2 booting problem

Post by CapSmasher » Mon Nov 05, 2018 10:53 am

1024MAK wrote:
Wed Oct 31, 2018 4:21 pm
Oh, by the way, to reduce stress on the PCB, when removing a component where you are not intending on reusing it (such as an electrolytic capacitor), it’s better to note down the holes it uses and the polarity (or take photos), then cut it’s legs, then desolder the legs, then fit the new one.

Mark
Hello,

I managed to desolder and check the C41 capacitor and it was in good parameters despite its looks (I will change it anyway just to be sure). Unfortunately I don't have a power supply that I can limit to a certain current value, I will receive one in a month or so.

In the meantime...

...is there any other way to detect where is this short happening? Do I really need to desolder any component and check? I read something that the TR9 instant burning can be triggered by a short in at least one lower RAM on the 48k Spectrums. In my case the only chip that uses +12 would be TEA2000? Could I test if this is the fault if I will desolder the PIN11 from the board and if the TR9 transistor will not burn that means this chip was the fault?

Thank you
0 x

Post Reply