+2 grey rf out problem

edited January 2013 in Hardware
I got a kind of broken +2. it basically works, but with a annoying rf out problem.

When screen goes black (eg. after border 0: paper 0 command) it loses sync and picture starts to move upwards very fast. When bright parts are in, it stops moving, and clear picture returns. (I checked the tv with an amiga 500 and aerial broadcast, it works fine even in black screen).

Last night I opened the +2 up, and when I hover my hand over RF modulator tv image dissapeared. I moved my hand and image came back. I popped up RF unit's cover, and found the area of interest, which is a transistor right side of the rf unit. I don't have to touch it, but I found an exact spot where image is very clear (almost like RGB :) ) and steady, even in black screens. Picture sometimes stays black, until I put my finger in rf unit (again not have to touch the parts, 2-3mm close is enough, so it's not a broken solder joint). if i touch the transistor, or getting close to it with a screwdriver, image on the tv always disappears.

Here is the video of me fingering my spectrum: :P
https://vimeo.com/56600267

So is this a normal behaviour? Or is it fixable? I don't know if it's a common problem. I think broadcast channel is not steady on this zx spectrum, I don't know what to replace.
Post edited by Arda on

Comments

  • edited January 2013
    Certainly this is not normal behaviour, but why bother with the TV out on a +2? Get yourself an RGB-SCART cable and the world is much brighter :-)

    Just make sure you order the right one. Your Speccy looks like a grey +2, so no +2A/+2B, right?
  • edited January 2013
    Check here after Friday, get a cable from him and your RF woes will be behind you!
  • edited January 2013
    well, it's all about mobility. The color tv (as you see in the video) is very very small. 8 or 9 inch one. it's very easy to carry it to the parties, also it's a crt one, so it produces very nice image when you watch interlaced demos. Sadly it has no rgb or composite in, so I have to use RF :)

    Maybe it's better to fit a composite in socket into tv, but I'm not sure if I can find composite in wires in the tv.
  • edited January 2013
    Arda wrote: »
    Last night I opened the +2 up, and when I hover my hand over RF modulator tv image disappeared. I moved my hand and image came back. I popped up RF unit's cover, and found the area of interest, which is a transistor right side of the rf unit. I don't have to touch it, but I found an exact spot where image is very clear (almost like RGB :) ) and steady, even in black screens. Picture sometimes stays black, until I put my finger in rf unit (again not have to touch the parts, 2-3mm close is enough, so it's not a broken solder joint). if i touch the transistor, or getting close to it with a screwdriver, image on the tv always disappears.
    First try fine tuning on the TV. The modulators in home computers are not at all accurate in the frequency of the TV channel that they generate (they were designed to work with TV's that had analogue tuning circuits).

    Also if you have access to other TV's, try your Speccy and see if you get the same problem.

    The reason that your hand or finger affects the circuitry in the modulator, is because your hand or finger acts as an extra small value capacitor (remember humans are effectively a salt water filled bag :D ).

    Modulators are analogue in nature and over time, the value of the components drifts. I suggest that in your +2 this may be what has happened :sad:.

    If you get the same problem on other TV's and fine tuning does not help, if you wish to repair the UHF / RF TV output, I suggest you replace the modulator (maybe some one has a modulator from a non-working / scrap machine).

    Or get another Spectrum...

    Note that a lot of the GREY +2 machines have a manufacturing fault, so do not produce a proper composite video signal on the video out / RGB connector (this does not affect the UHF / RF TV output).

    Mark
  • edited January 2013
    What I noticed in your video is actually that touching the external shell of the modulator (which is ground) seems to make a difference.

    I dont think its actually the finger getting close to the 9018 transistor but the touching of ground.

    Long shot but check the two solder points of the modulator's case on the board...
  • edited January 2013
    gtsamour wrote: »
    What I noticed in your video is actually that touching the external shell of the modulator (which is ground) seems to make a difference.

    I dont think its actually the finger getting close to the 9018 transistor but the touching of ground.

    Long shot but check the two solder points of the modulator's case on the board...

    Umm, no, when I shooting video maybe I accidentaly touched to the ground, but no, if I touch only ground nothing changes. But if I touch the transistor (without touching the ground) image disappears. Again, I don't have to touch anything, but even holding my hand 5-6 cm over modulator changes the picture slightly (maybe this is expected). But under 4-5mm you figure out that it's the damn transistor :) I know, this is stupid, but there must be a weak point of rf resonance or something. I am not a spiritual man. I guess I will replace the modulator.
    Note that a lot of the GREY +2 machines have a manufacturing fault, so do not produce a proper composite video signal on the video out / RGB connector (this does not affect the UHF / RF TV output).

    Yes, I already turned the TR4 upside down to correct the comp signal. I'm not sure it changed anything. As I am testing the +2 with only RF out.

    I also tested it on different TV, behavior is the same. But, both of the tv's has digital tuner. So fine tunning is not manual.

    Ok, thank you everyone, as I understand it's possibly because my capacitance increates due to my diet and winter, and speccy's components got very old. Yet ~25 years is not a long time, maybe even last days of puberty :)
  • edited January 2013
    Arda wrote: »
    Last night I opened the +2 up, and when I hover my hand over RF modulator tv image dissapeared. I moved my hand and image came back. I popped up RF unit's cover, and found the area of interest, which is a transistor right side of the rf unit. I don't have to touch it, but I found an exact spot where image is very clear (almost like RGB :) ) and steady, even in black screens.

    So is this a normal behaviour?
    Yes it is. The circuitry around that transistor is an UHF oscillator. Moving your finger into this area tunes the oscillator. So what you have to do is fine tuning the TV set.
    The metal cover of the RF unit shall avoid any influence of external field variation.

    Greets Ingo.
  • edited January 2013
    @Arda

    My friend, have you corrected the factory mistake of TR4's orientation. Originally the board is silk masked the wrong way round. This reduces the RF and Audio output.

    You will find TR4 on the bottom right of the UHF modulator.

    pic_001.jpg


    Here is TR4, just below LNK2 correctly installed.
    pic_002.jpg

    As you can see the TR4 transistor is now reveresed compared to its silk mask. Amstrad made an error in the inking of the PCB which places the emmiter on the collect line and vice versa.


    This may help you problem with RF. good luck my friend =D
  • edited January 2013
    Hi ZetrO,

    you are rigth, that Amstrad did this mistake with TR4. But TR4 does not process the signal that goes to the RF unit. TR1...TR3 processing that signal and there is no such error around that transistors.

    Greets Ingo.
  • edited January 2013
  • edited January 2013
    So... what exactly does the tr4 reversal fix guys? Does it only have to do with composite on the RGB socket as was mentioned here? Because I had a +2 gray that could not hold a stable RGB on a Sony CRT (the tv was switching in and out RGB mode). Could that be the problem?

    And is this fault present on most or just some grays?
  • edited January 2013
    gtsamour wrote: »
    So... what exactly does the tr4 reversal fix guys? Does it only have to do with composite on the RGB socket as was mentioned here? Because I had a +2 gray that could not hold a stable RGB on a Sony CRT (the tv was switching in and out RGB mode). Could that be the problem?

    And is this fault present on most or just some grays?
    Yes and yes, all gray +2 computers I saw had that error. The reversed TR4 has too low amplification. After reversing it into the right direction the composite signal at the video connector that is also used for synchronizing the RGB signal has the right level.
    If you use the RGB output you should reverse TR4 as well as desoldering the capacitor that is leading the sound carrier signal to the TEA2000. Even if you use RGB the interferrence of the sound carrier and the color carrier leads to a noticeable jitter in the synchonization. Vertical lines may be frayed, synchronization at modern TV sets maybe instable.

    Greets Ingo.
  • edited January 2013
    Amstrad in some versions of the GREY +2 made a manufacturing mistake.
    Computers with this error do not produce a proper composite video signal on the relevant pin of the RGB (video) out connector (this does not affect the UHF / RF TV output). The problem is that three transistors (all type 2N3904) have the incorrect marking on the PCB, so have been fitted incorrectly (they have been fitted the wrong way round).

    TR4 is one of these and is the "buffer" transistor for the composite video feed to the RGB (video) out socket (DIN socket pin 1).

    It does not affect the RGB video lines, but some SCART leads try to use the composite video signal to feed to SCART pin 20 ("video in") which is used as composite sync in RGB mode. If the composite video signal was good, this would work okay...

    This is the fix I used on my RGB lead (but it may not work on all TV's):-

    In the SCART lead I found there is already a composite sync signal (that is not affected) routed to the SCART plug on pin 16 ("Blanking"). It is from the DIN plug, pin 4 ("CSync").

    I removed the wire for the composite video ("CVBS") from SCART pin 20, insulated this wire so the conductor could not touch another connection.

    I added a short link wire from SCART pin 16 to SCART pin 20, thereby routing the composite sync signal "CSync" to SCART pin 20 (as well as pin 16).

    The supplier of this lead suggested another variation. Apparently on some TV's, the CSync signal voltage is dragged down too much when feeding both SCART pin 16 and pin 20. So route the faulty composite video ("CVBS") to SCART pin 16 and don't bother with a link between SCART pin 16 and pin 20. Have not tried this last bit myself yet.

    Hope this helps (yes I know it's a bit confusing :D ).

    Mark

  • edited January 2013
    1024MAK wrote: »
    Amstrad in some versions of the GREY +2 made a manufacturing mistake.
    Computers with this error do not produce a proper composite video signal on the relevant pin of the RGB (video) out connector (this does not affect the UHF / RF TV output). The problem is that three transistors (all type 2N3904) have the incorrect marking on the PCB, so have been fitted incorrectly (they have been fitted the wrong way round).
    Mark

    Three wrong positioned transistors??????? Which are the other two and what do they cause?

    Yes using csync (wherever it is available) at pin20 gives better picture in general compared to composite video. In gaming consoles such as the Sega megadrive, PCEngine etc, by using csync instead of composite at pin20, you dont get the known checkerboard patterns.
    For the +2 gray you shouldn't see any difference though, but you can use csync yes... to bypass the weak (due to reversed transistor) composite problem.

    @Ingo
    Thanx Ingo...
    As for the capacitor.... can you tell me which number is it? And after I remove it would anything else be affected? Will the machine work 100% without it? Will there be sound from RF without it?

    :Edit

    I have three +2 grays.. all three ISS3 were with wrong positioned tr4. Fixed them and I get stable RGB now. There is a very faint verical kind of waving in the letters for example. Probably thats what Ingo is talking about.
    If I use csync on pin20 of scart and also connect composite to pin16, there is no effect like that BUT I get stable RGB on two grays (out of three) on a sony CRT but on all on an LG LCD.... Weird since all 3 grays are identical (Issue 3)
  • edited January 2013
    gtsamour wrote: »
    Three wrong positioned transistors??????? Which are the other two and what do they cause?
    I have put the information in a new thread, see here...

    Mark
  • edited January 2013
    Can anyone else confirm that all three transistors (TR4, TR5 and TR7 (all type 2N3904)) need to be reversed before I start doing it? :)
Sign In or Register to comment.