Side Projects

[Cressy Snr]

And another.

[andrew Ivimey]

I'm just lurking....oops!

[Nick]

What I find very interesting is how complex it all is. I guess its because of the parts that were available to the designer at the time. Its interesting that they are regulating the -ve rail, no reason why not and I wonder if that was as much a result of the thinking of the time where most transistor designs were 0v and -ve rails. Today we are much more used to 0v and +ve rails. I have always assumed that the -ve rail thing was a combination of valve based thinking and PNP transistors. Its like a lot of the push pull audio amps from those times were close as possible to valve designs even down to the transformer coupling between stages.

[Mike H]

I'm pretty sure that's all it is, Nick. Early transistors era, therefore, positive ground rail. Of course I've gone and drawn it 'upside down' because it does my head in using pos as earth.

[Mike H]

Another crazy in the Not Andy’s Man Club. Not to put Andy’s Man Club down that is to say.
The first lesson to take away is take the lid off youre old desk power supply guys and renew youre electrolytic capacitors, or at least check visually. Better check for current flow.

Absolutely. I'm guessing it's from mid 1960's era, maybe a bit later. Don't think it's '70's, because I was starting to dabble in the hobby then, and the board, and the components on it, look like one of those 'scrap boards' you could get out of the bargain bin at the local electronics shop for 50p each.

So yeah the components are about 50 years old. That is pushing it for a 'lytic!

Dave Brooks (Danbury Electronics) used to work for Drake Transformers, then he was a salesman for a while, he said, which TBH kind of stretches my imagination a bit, having known him, I only mean that he seemed more of an engineering type of guy rather than a hard sell guy, then as I said he started his own winding business, I want to say about 1982, (I started with Maplin 1984 and I know he was supplying them before then). To start up he bought a couple of second-hand winding machines which very likely dated from the 1960's, and these 4 power supplies came with the machines, and doubtless other impedimenta. During the following years 2 of them drifted off to their own fates or futures, and I had one of the remaining two which literally were gathering dust on the floor under his work bench.

[Mike H]

Speaking of capacitors, the RS order came in today .....
-

Marconi DS 50_2 new caps-1.jpg (121.41 KiB) Viewed 5920 times

-

The cans are very nice - and not cheap tat, altho RS do stock much more expensive ones!
-

Marconi DS 50_2 new caps-2.jpg (79.02 KiB) Viewed 5920 times

-
There's only 2 because I only need 2, one will do the job of the original 2 x 1,000 µF. That will free up a bit more space and I think I could slot the small toroidal into the gap.

[Mike H]

Couple more drawings -

I finally managed to combine my earlier regulator schematic with the op-amp, with the 'real' differential amplifier schematic.

First, the original differential circuit on its own - it is starting to look a bit more like an op-amp - note CCS Q4, which may look a bit odd. 2 diodes to bias it, shirley? Ah yeah but no but, germanium transistor Vbe is typically quite low, all those I'm using here, less than 200 mV, D1 on the other hand is 0.64V, so there is a useful Voltage drop across R1. The current is 2.4 mA.
Vs1 is +9V, Vs2 is —9V. Remember these are relative to 'Bk10', and that 'Bk10' is also the negative or black output terminal, also the one that is being varied by the Voltage control(s). But here it's also being used as another positive supply rail for R3, R4 & C1. Yes it's confusing!
-

Marconi DS 50_2 regulator sub block real.png (24.21 KiB) Viewed 5908 times

[Mike H]

Now replacing the 'op-amp' in earlier drawing 'Marconi DS50-2 regulator-2.png' ...
-

Marconi DS 50_2 regulator-6.png (97.82 KiB) Viewed 5897 times

-

Yes again this simulation runs, with no mods. The previous version with the 'ideal o-amp', centred 'wipers' of VR1 & VR2 gave an 'o/p' of 26V. Here, the o/p dropped to 24. So I tweaked the preset VR3 for 26V again. But I didn't have to do anything else to make it work.

You may be thinking, as I did, that the o/p of the amplifier is a bit weedy, but no, it works fine. Turns out the base current into the BC161 (Q5) is very small, so no problem.

NB: BC161 isn't the original, it's one that I put into the empty space on the board, as someone before me had removed the original. It's silicon, with a modest Hfe of 90. I got datasheets for all of them, and BC161 has the greatest gain of the lot. The power transistors I have set at 45, all the other small ones are between that and 60 (using datasheet Hfe info).

The two missing transistors I replaced back in 2014(?), I don't think they had been removed with a view to replacing because they had failed, it had all got damp at some time so that their legs had corroded through. A third transistor still on the board, also had a detached leg, due to corrosion. I was able to reattach it to the board using a bit of added wire as an extension.I'm supposing the previous repair stalled because they couldn't source replacements, or thought there was another problem.(?)

Not sure why we need the 5.1V Zener D2 (BZY88C5V1). It actually prevents the drop across VR3 exceeding 5.1V. Because ...?

Ignore R10, that's simply a way of joining the two separate wires 'Bl11' & 'Bk10', this point is actually a toggle switch on the front panel, a drawing of which is coming shortly...

Don't forget, 'VR9-F' are Zeners, and +9V & —9V are coming from the low power secondary and bridge rectifier etc.

[Mike H]

Some of the front panel wiring, excluding the pots (we've done those) ...
-

Marconi DS 50-2 schematics-4.jpg (132.81 KiB) Viewed 5887 times

-

'O/p switch' does what it sounds like, it turns the output to the black neg socket on or off. NB: there are two '[Bk8]' wires (black with '8' cable marker), because when I was in here last time I thought it would be a good idea if it was possible to set the current limit while the switch was off, i.e. no connection to any external circuit. Bit late fiddling with the current limit if that's already smoking ...

The ammeter is in series with the supply to the red terminal. '[Rd18]' (red wire, '1' & '8' cable markers) comes from the positive end of the 2,500 µF smoothing cap (the same end that's fed from the smoothing choke).

'[Bl11]' (blue wire, '1' & '1' cable markers) is the wire that goes back to inverting input of the differential amplifier.

I don't know how the other end of the Voltmeter is wired up yet, but a preset is involved. I remember last time I had to replace a burnt out adjustable wirewound resistor (value changed by a moveable metal collar), which set the FSD for the meter, I used a couple of 1% metla film resistors. Dread to think how it got burnt out in the fisrt place.

[Nick]

I wonder if D2 is to protect the transistor base from voltage injected into the black terminal from the outside world.

[Mike H]

Good thinking Batman!

I also didn't fail to notice D1, which is the kind of thing you sometimes see in op-amps too.

[Nick]

Yep generally on differential input stages it's to prevent the input differential voltage from saturating the input pair.

[Mike H]

And finally for this evening, one mystery solved - which was, why the blinking flip does this metal rectifier even exist ... 'cause the way it's connected up seems weird and pointless ...
-

1S420 diode.jpg (19.06 KiB) Viewed 5856 times

-

It's an 1S420, silicon, 100V 10A ...

It's connected between the negative end of 2,500 µF smoothing cap (C4, via a 1 Ω resistor & wire '[Bl38]'), and the positive side of the reservoir caps (C1 & 2, aka positive of bridge rectifier). So I plugged it into the block diagram ...
-

Block Diagram original values.png (100.51 KiB) Viewed 5856 times

-



... and ran the sim. And this is what happened ...


-

Function of diode 1S420.png (55.29 KiB) Viewed 5856 times

-

Red 'trace' is the reversoir Voltage and you can see it charging up at the lefthand end (power start up).
Green is the red output terminal Voltage, aka positive side of the 2,500 µF, and there's a big pulse and then some ringing.
Dark blue is the diode conducting, therefore the neg side of C4 went more positive than the pos side of C1 & 2. The pulse is much bigger without the diode (first version of the schematic, I had to cheat by adding 100 Ω or less parallel to the choke to damp it out).
Light blue is the Voltage across C4, and seems normal.

What I'd like to know is, was someone farsighted enough to include it in the initial design, or did they have a problem with the protoype and have to find a fix.


More later...

[Nick]

It looks like it's intended to allow the cap to discharge on shutdown. Also I would see this sort of thing to prevent exceeded the reverse voltage limit of the power transistor.

[Mike H]

Yep does that too. The discharge thing.

The nice post lady brought the transformers today, so I should really get on with it. But you know how it is ..