Testing memory/transistors etc with multimeter

Comments

• RST#08

  January 2011 edited January 2011

  Gosh, where to start... Google "How to use a multimeter", and you'll find lots of video and normal tutorials that'll help to get you started.
  
  The diode setting is probably the least useful setting when testing a Spectrum. The diode setting is most useful when testing discrete components.
  
  For basic Spectrum testing, start with the basic DC voltage setting. This will help you to identify the +5v, -5v and 12v which are so often a problem with the Spectrum.
  
  Testing transistors in situ is more problematic. On a live circuit, you can only test their effect - you do this by measuring voltages!
  
  For example, you can effectively "test" (for example) TR4 and TR5 by testing that the internal power supply is producing -5v and 12v when under load.
  
  Hope this is enough to get you started. I'd suggest doing a quick search on the WOS infoseek section for service manuals: these will help you identify the correct voltages etc.
  
  Of course, seasoned engineers will identify lots of more sophisticated tools you might use. But, for most common spectrum faults, a basic multimeter will probably do everything you want.
  
  Hope this helps
  
  Nx
• Macc

  January 2011 edited January 2011

  Thanks for the tip on the diode setting, it was basically the only setting I could get any reading off (probably due to testing unpowered as Im currently doing!)
  
  Ill get a few schem's and take a look.
  
  Thanks again and if anyone else wants to chip in with advice, feel free to do so, all help appreciated!
• orange

  January 2011 edited January 2011

  that diode setting can be useful. iirc, it gives beep when you shortcircuit it. so its used to test if some tracks are broken, bad joints, etc..
• Macc

  January 2011 edited January 2011

  mines quite a cheap one, doesnt beep!:-(
  
• Macc

  January 2011 edited January 2011

  Just finished testing the bank of 8 chips bottom left, black probe on heatsink, red probe touching bottom left corner pin, reads 12v, bottom right corner pin, reads 5v, well at least that looks promising! (This is powered obviously) on the tr4 im getting 10.8 and the tr5 4.9, z80, bottom left pin 3.84, bottom right 2.65, on the lm1889m im getting 3.92 bottom right but nothing at all bottom left (Im assuming youve got the speccy board facing the right way and you are looking from behine it here btw)
  
  Any of the above seem right to anyone?
• orange

  January 2011 edited January 2011

  you should test top left pin on RAM chips, too. it should be -5V.
  
  btw, no need to test voltages on each chip (sorry to state the obvious, just to be sure).
• hedge

  January 2011 edited January 2011

  I expect you've seen this but just in case http://www.worldofspectrum.org/SpectrumRepairGuide/
  
  I've used it a lot I think it's based on a v2 board. I get different voltages on my 3b -not a lot though.
  
  I've tried and indeed still try to fix em but unless it's lower ram or ula I've not succeeded. Very frustrating but I hope you get what you need from this.
  
  Cheers
• Macc

  January 2011 edited January 2011

  thanks hedge ill have a look mate
  
  (btw, ram chips read -5v correctly too)
• Macc

  January 2011 edited January 2011

  Done some more thorough testing and it appears, if results are correct that in at least three boards that ive tested the lower ram at least seems sound, went through the steps in the guide, all reporting within accepted values, what isnt ok however appears to be tr5, and this is in all of them pin 1 +6.7v, 2 +5.95 and 3 +5v on one (thats the one with the strange little black dots that appear) 5v across all pins in one that has the vertical "power on" lines permanently on screen, 0.8, 4.8 and 5 on another one that seems to be in a constant power on loop, im getting "somewhere", just not sure where!
• Pilsener

  January 2011 edited January 2011

  Remember the finger-test as well: Any RAM that's too hot to touch for more than a few seconds, are most likely defective.
  And differences in temperature among the chips are generally a bad sign.
• Macc

  January 2011 edited January 2011

  cool as cucumbers and thats after ten minutes of power
• roko

  January 2011 edited January 2011

  Macc wrote: »

  .... z80, bottom left pin 3.84, bottom right 2.65, on the lm1889m im getting 3.92 bottom right but nothing at all bottom left (Im assuming youve got the speccy board facing the right way and you are looking from behine it here btw).....

  Not all chips have the Voltage and Ground pins in the 'corner' position, please find yourself some datasheets/pinouts! As far as for the Z80 the 'bottomleft' could be the /IORQ signal, which for Spectrum will be rather 'hi' as that signal is only active (='lo') for very short moments. So 3.84 (not too far down from a 'hi') is not unlikely. Bottomright could be /RD, which signal goes lo for every call in ROM, and in general (very general!) you could say that 2.65 is not abnormal.
  The problem you meet here is that not all is solved by just measuring, interpretation of measured values is another thing! You did OK when measuring the (steady) supply voltages, that's a good start!
• Macc

  January 2011 edited January 2011

  thanks roko, ive had to walk away from it tonight as i seemed to be going round in circles, ill have another try tomorrow