12V unloaded voltage

For experts to discuss very technical stuff and newbies to ask why the Spectrum they bought off ebay doesn't work.
Post Reply
jokkemokke
Drutt
Posts: 8
Joined: Tue Aug 28, 2018 7:10 pm

12V unloaded voltage

Post by jokkemokke »

Hello. I have a Spectrum ZX +2 (issue 4s) which that is not working, so I was trying to get it running again. There was no power to the board at all and it turned out to be a faulty 7805 regulator. I replaced the regulator, and there was now +5V on the board. I then tried measuring the +12V and the -5V supplies on the lower RAM, but there was no voltages on any of these.

I figured that probably TR4 and/or TR5 were faulty, so I replaced them both with same type as the ones already installed (TR4 = C2236 and TR5 = A733). I then tried powering up the Spectrum again with all RAM removed, because I wanted to check that the voltages were ok before installing the RAM to make sure that I would not blow anything in case my repair did not work.

Now when measuring the voltages I have a nice +5V and -5V supplies for the RAM. However, the +12V seems to measure 26V-28V for some reason. I was thinking if this behaviour could be because the +12V line is not loaded by the RAM circuits anymore and maybe therefore the internal power supply does not manage to regulate the voltage correctly to +12V.

Could anyone confirm if this is the case before I go ahead an install the RAM again? Or is there something else I should check first?
User avatar
1024MAK
Bugaboo
Posts: 3104
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: 12V unloaded voltage

Post by 1024MAK »

Can you please post a photo of your board.

The +12V supply is supposed to be regulated. It should not go more than about 5% above 12V.

It is TR5 that does the regulation. This PNP transistor compares the voltage at it’s base with the voltage at it’s emitter.

The emitter is connected to the +5V supply. This is the reference voltage.

The voltage at the base comes via a potential divider formed by R59 (1.8k, 2%) and R58 (1k, 2%). The potential divider drops the voltage on the +12V line. This is the feedback voltage. If the +12V line is actually at exactly +12V, then this voltage at the base of TR5 should be between 4.2V and 4.3V. Capacitor C74 smooths the voltage at TR5 base so that the regulation circuit does not oscillate.

If the voltage at the base of TR5 is lower, TR5 turns on harder and the resulting increased current flow causes the voltage at the collector to go up. This allows TR4 to work harder. TR4 working harder causes more energy to go into the magnetic field of ‘the coil’, which is a high frequency transformer. This causes the +12V supply voltage to rise.

If the regulation circuit fails, the voltage on the +12V supply can rise in an uncontrolled manner. Do not leave the computer in in this condition. As the colour encoding chip also uses the +12V supply.

Mark
:!: Standby alert :!:
“There are four lights!”
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :dance
Looking forward to summer later in the year.
jokkemokke
Drutt
Posts: 8
Joined: Tue Aug 28, 2018 7:10 pm

Re: 12V unloaded voltage

Post by jokkemokke »

Thanks for the explanation. I am thinking that I might remove the LM1889 temporarily so that I can leave the board powered without worries while I'm trying to find out what is wrong.

This is how it looks like:

Image
jokkemokke
Drutt
Posts: 8
Joined: Tue Aug 28, 2018 7:10 pm

Re: 12V unloaded voltage

Post by jokkemokke »

Hello again. Today I spent some more time investigating the issue. I tried measuring the voltage divider R58 and R59. On the DRAM supply I have 30V so I expected to have around 10.7V on the base of TR5. However, for some reason I measure exactly 5V on the base of TR5. I figured that this could happen if C74 is shorted so I tried to replace it, but the result was the same.

I then tried to google the A733 transistor which in my case is installed as TR5 and found this:

Image

When I tried to measure the board according to the pinout in this figure I got confused. According to the diagram, the transistor pinout in the figure and my measurements it seems that the following is the case:

Emitter is connected to 5V (ok)
Collector (middle pin) is connected to the R58/R59 voltage divider (not ok)
Base is connected to R60 (not ok)

It appears that somehow collector and base have been swapped.

I have a working Spectrum ZX+ also using a A733 as TR5 and I have verified that the transistor is installed in the same way in both boards.

Could it be that the old A733 in the working board has different pinout than the new A733 that I installed in the defect board? Could anyone please clarify?
jokkemokke
Drutt
Posts: 8
Joined: Tue Aug 28, 2018 7:10 pm

Re: 12V unloaded voltage

Post by jokkemokke »

Success. Another Speccy brought back to life. :D

It seems indeed that there are different versions of the A733 with different pinout. I tried to desolder the TR5, bend the legs so that base and collector became swapped, and soldering it back in place. After powering on again I had 5V, -5V and 12V as expected.

I then reinstalled all chips, powered on and it was working. I guess I was lucky not to fry anything with the high voltage on the 12V line.
User avatar
1024MAK
Bugaboo
Posts: 3104
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: 12V unloaded voltage

Post by 1024MAK »

I don’t know about the A733 transistor, but some European transistors were available in different pin-outs. For example the BC214 and BC214L are the same transistor in the same package, but with a different pin-out.

It may be that there is more than one transistor with a part number of A733. Or with part of a part number that ends A733.

Transistors are either numbered according to one of the principal industry ‘standard’ numbering systems, or with a manufacturer specific marking system. Further, often the first part of some ‘standard’ numbering systems is not printed on the smaller transistors.

The most common numbering systems are:
  • 2Nxxxx series, known as the Jedec series. The EIA in the USA maintains a register of all 2Nxxxx devices.
  • The European Proelectron numbering system. They have a two letter/three number or three letter/two number code. The code starts with an ‘A’ for germanium devices, or a ‘B’ for silicon devices. E.g. a BC548 is a common silicon transistor.
  • The Japanese 2Sxxxxx JIS numbering system. Nearly all transistors in Japan are part of the JIS numbering system. The number for all conventional transistors start with 2S. Then the next letter indicates if it is NPN or PNP and also divides the range into low frequency and high frequency groups. A = PNP HF, B = PNP LF, C = NPN HF, D = NPN LF. Then there are 3 or 4 numbers. So for example 2SB364.
The manufacturer specific markings can be anything.

The problem is that a part marked A733 could also be a 2SA733 (which is a silicon PNP type in a TO92 case, lead out BCE).

The transistor used in position TR5 can be a general purpose PNP type. Other types that can be used include ZTX750, BC307, BC557. Note that these have been suggested as replacements for the ZTX213 transistor used on issue 1, 2, 3, 3B, 4A, 4B, 5 and 6A boards. NOTE that these do NOT all have the same pin-out.

The pin-out of the pads on the issue 4S board may be different.

Edit: I see that you have sorted out your problem :D Well done 8-)

Mark
:!: Standby alert :!:
“There are four lights!”
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :dance
Looking forward to summer later in the year.
Post Reply