Thinking about this yet again...It is interesting and unusual that the problem is the power supplies won't turn off. It seems most of the time power problems are missing power. From your last revelation of using a jumper to attempt to turn 1TR21 off (by connecting 1TR21 collector to emitter) it tells me the problem (as Ralph-Marcus also noted) is on the power side of the circuit and not on the microcomputer control side. 

The other interesting thing is that whatever the problem is, it affects both the +15V and -15V supplies. Those two supplies have separate transistor controls though (0TR1 and 1TR19 respectively).  A problem in and around 1TR19 would affect both plus and minus supplies though. 

I looked back at some of your measurements when you were trying to match up with the simulation probes. At least one of them doesn't look like you got the right point measured. For example PR14 has to be around -24V and you showed 0V. But that is the source of the -15V and you showed that you had -15V.

Focus in on 1TR19 base and emitter voltages for the Beogram ON and OFF states.  There should be a 0.6V to 0.7V difference in those when the -15V is on and near zero difference when the Beogram is supposed to be off. Your earlier measurements showed the 1TR19 emitter at -22.5V and the base at -17.4V. That means 1TR19 would be on and -15V is on (which is the case with your circuit). So why is 1TR19 not turning off?  If 1TR20 has its emitter side grounded (per your jumper) then it will be off and the 1TR19 base and emitter voltages should measure close to the same value (keeping 1TR19 turned off). That apparently is not happening in your circuit. 

The answer is NO 😕 to zero volts at E - TR20

next focus was on TR19 -E,B 

I can only confirm the previous readings at E and B.

not changing whether I have TR21 in place or out and play only with that jumper to simulate on and off

absolutely no impact that I can see in the readings ??

this is seriously odd ?

I do however like the diagnose putting the uP into the clear as this leaves at least some hope OFF will be achieved in the end 🙏

ALF

Okay, so with the jumper in place that removes the microcomputer from this problem. Its function is to turn 1TR21 on and off. Using the jumper you still cannot get the 15V supplies to switch off so while that is a mystery it does narrow things down to the main board and an issue with the controlling components there. So that is a bit of positive progress. Yes, I would classify this as seriously odd but it isn't magic. There has to be a logical culprit you just haven't found yet. It is starting to sound like some sort of ground (bad ground) problem. 1TR21 - E connects to ground. If you jumper 1TR21 - C to E then 1TR20 - E has to be ground and 1TR20 has to be off.  

Continuing with that... 1TR20 being on causes current to flow through R82 so 1TR19 - B is greater than 1TR19 - E turning it and the 15V supplies on.

Leave the jumper to take care of 1TR21 and see what happens when you lift off R82's lead that connects to 1TR20 - C.
Doing that should force both 15V supplies to turn off because no current should be taking the 1TR19 - B voltage higher than its emitter. At the same time, please report the voltage levels of 1TR20 - E and C.  

That should also shed some more light on the problem. I hope this is another step closer as I'm sure you are too :-).

NO............😡

with TR21 removed, jumper connected C and E/ground of TR21 terminals,  R82 leg connecting with C- TR20 out..... power is still on........🍷🍷🍷

This drive is getting power.... unless I take out the transformer ——— but from where ?

I simply don’t believe it - do we have to take out the complete ON/OFF circuit ?

ALF

Wow!  That  shouldn't be able to happen.  So what did 1TR20 - E, B and C measure under that test condition?  Do you still get -1.4V where you have the jumper?  Also, what to you measure on the lifted R82 leg?

It kind of sounds like an undesired short to ground somewhere. It almost sounds like the node between R82 and 1TR20 - C are shorted to ground while the connection between 1TR20 - E and D26 is not connected. I am not guessing at that...I am just playing with scenarios on the simulator that end up with similar voltages you are seeing and result in the 15V supplies still on. 

I don't know about all of the On/Off - 15 V supply circuit but I feel like the suspect area involves 1TR20, 1R82, 1C30, 1R81, 1R80, 1TR19, 1D26.  Not necessarily bad components but maybe. The interconnecting wire paths (traces as well) of those components is also suspect.

If you would, report on the measurements I requested in the first paragraph above. Then we might have an idea what to look at next. 

Ok,

with the jumper connecting the C and E terminal of missing TR21, deck in OFF mode, platter spinning reverse

TR20 showed the following:

E = -1.9V,  B = 0V,  C = jumping between -0.150V and -0.900V

lifted leg of R82 = -17.8V

ALF

So 1TR20 should be a solid 0V with the jumper to ground (1TR21 - E to C). Since that is not a solid ground it appears that is one problem.
With R82 lifted there shouldn't be any current flowing on that 1TR19 circuit but there is some path there because you have -17.8V instead of close to -24V (or whatever 1TR19 has at its emitter). That is a second problem. Maybe they are related but either way they shouldn't be the voltages at those locations. 

I'm too tired to think any more on it for now. 

Hi ALF,

John is correct, you can't have voltage at the emitter of 1TR20 if the collector of 1TR21 is properly grounded. 

Instead of just using a link between C&E of 1TR21 why don't you try a link between C of 1TR21 and another earth point like the negative terminal of C40.

I'd try measuring the +/-15V supply with R82 added back, 1TR21 removed and adding a link from C 1TR21 to C40 earth. 

Regards

Lonnie

Interesting suggestion from Lonnie (thanks Lonnie) connecting the jumper to C40 ground instead of using TR21-E ground.

looking again at TR20 voltages (thanks John):

with R82 in place: E = B = about 0V, C = -16.6V

with R82 leg lifted: E = B = about 0V, C = -0.62V.    R82 leg = -16.6V 

+/-15V supply still present !

ALF

Okay, good...there is a problem on th 1TR19 side. That's why we wanted you to check that.

You should also see that 1TR19 - E is in the -22V to -24V range. But with the 1TR19 - B side being at -16.6V that transistor (TR19) is always on...causing the 15V supplies to always be on. 

There must be a path from 1TR19 - B to ground besides 1TR20. That is causing 1TR20 to not have any control over 1TR19 that it should. 
It is also looking like you have some bad ground connections as the new ground you used put the 1TR20 - E, D26, 1TR21 - C, R83 node at zero volts as expected (and intended).  So the 1TR21 - E ground seems suspect and there appears to be some sneak-circuit ground that is affecting 1TR19's base.

With the Beogram unplugged what resistance/continuity do you measure between the 1C40 ground and the power supply circuit grounds - C24, C27, C29 and the transistor grounds of 1TR21 - E, 1TR20 - B, 1TR18 - B. Keep one DMM probe on the 1C40 ground and measure the other grounds to it. The resistance between all of those grounds should be pretty close to 0 ohms (although you might measure 0.2 ohms depending on your meter).