Post by Vyper68 on Aug 8, 2019 10:18:09 GMT
I have recently picked up a battered and bruised Spectrum 128K +2, it was sold as "untested but powers on fine", so I was hopeful it would be OK. This turned out to not be the case for this poor old Speccy, there was no UHF output and the RGB/SCART output was so dark it was virtually Black, you could only just make out the 128K boot menu.
I first thought that the SCART signal was not switching the TV to RGB and that was the cause, this happened with my +2B and needed a resistor value changing to bring the signal back into range for the TV to switch to RGB. So I changed a couple of links on the PCB to set pin 1 on the RGB socket to 12V from the TEAC video chip, this combined with a 470R - 600R resistor in position R9 acts a voltage divider working with the TV to produce the correct volatage to switch a TV into RGB mode.
So I was a bit peeved to find this didn't work, but on the other hand not totally surprised and so I dug out my multimeter to do some tests, but first off I saw this this was a PCB issue 3 ( hey Baker another issue rather than version ) and TR4 which is a 2N3904 transistor was soldered in as per the silkscreen, unfortunately for Amstrad someone made a mistake and TR4 is actually the wrong way round!, this problem affects all issue 3 PCB's, incidentally the other 2N3904's on the PCB are also the wrong way round but do not affect the operation so are generally advised to be left alone. The Amstrad manufactured Spectrum PCB's themselves are poor qualitly which seems at odds with their CPC range as they are very good to work on.It is very easy to damage the Spectrums when desoldering components. So TR4 was removed and replaced with another 2N3904 the correct way around ( the opposite way to the silkscreen ), this again did not allow a RGB picture but was slightly improved.
So testing the voltages around the power input I was getting a good 5v reading but there was not 12V reading anywhere, I was only getting around 1.3V. So the two candidates for this are TR8 ( ZTX213 ) and TR9 ( ZTX650 ) these are both obsolete now but the alternatives are TR8 ( ZTX751 ) and TR9 ( ZTX651 or ZTX653 ). So I placed an order from Phil Ruston ( Retroleum ) and they arrived a day later. Quickly replaced both transistors and reverted the links back to the factory settings.
I powered on the Speccy and Tah-Dah!! a nice clear RGB picture with no dot crawl!
So in this instance the fault lay in the transistors at TR8 and TR9 and to get the 12V rail working again which is needed for the TEAC video chip and the main reason the video was not working. They were both replaced and also fitted a new TR4 the opposite way to the PCB silkscreen, without this the RGB socket would not work properly either!
So top top if you ever pick up a Grey +2 and get an almost Black/Negative video output check for the 12V! and look at TR4, TR8 and TR9
Another Speccy +2 has been saved from the scrap heap and after a good clean looks pretty darn good. I have also fitted an external EAR socket modification so I can now load .TAP and .TZX files through that socket and feed that into the Spectrum and bypass the cassette unit which means you don't need to buy another tape again apart from collecting them or if any new software comes out like the physical copy of the new ZX Spectrum game Quadron, which I just purchased from Cronosoft.