Maplin Precision Gold XG95D PSU

[Daphne_g7ena]

Hi All,

Someone gave me a XG95D but it is in a bit of a sorry state. I am looking for any information on this unit, particularly a schematic.

Can anyone help please?

Daphne

[SiriusHardware]

Can't find a circuit, sorry, but some snippets I found online suggest it is based on the well known 723 regulator ICs and uses LM324 quad op-amp ICs for the current sensing circuits. Both of these ICs are still easily available. The output transistors are probably relatively commonly available as well.

What exactly is the problem? If one half works but the other does not you can make comparisons between the 'good' half and 'bad' half, and if the whole thing is not working that suggests a problem with something common to both halves, ie, mains input transformer, etc.

[Daphne_g7ena]

Hi,
Thank you for your thoughts.
The left hand board was not plugged into the control cables or mounted to the current meter.
Voltage on the LH terminals was 13.7 volts constant despite adjusting the controls. The volt meter showed a negative deflection.
Voltage on the RH terminals was -16 to -38 volts with the volt meter showing up to 20V

With the LH board plugged in but not connected to the current meter the voltage measured was 0.9V and the RH side behaved the same as above.

Daphne

[SiriusHardware]

As we don't currently have the diagram available, some more information from you, especially some photos of the unit, would be helpful. From what you say it sounds as though there are two identical boards, one per channel, and they are removable?

-With the PCBs fitted and connected, please take a couple of photos of the interior general layout and interconnections between the PCBs and the power transformer and meters and output transistors (presumably mounted on a heatsink at the rear).

-Take full-size photos (as clear as possible) of the the upper and lower side of one of the PCBs.

If you can get at the output transistors on the heatsinks to read the numbers off them can you say what type, and how many, there are associated with each channel of the power supply?

[SiriusHardware]

After a lot of surfing I have managed to find images of this PSU badged as the 'Lodestar PS-303D'. That in turn led me to what appears to be the website of Taiwanese company 'Lodestar'. A search there on PS-303D didn't bring anything up but when I dropped the 'D' I got this:

http://www.lodestarelec.com/01-dc_power_supply.html

...Which bears a very distinct resemblance to the left hand side of your dual PSU. The 'D' in your model presumably stands for 'Dual'. It looks as though they've dropped the dual version in favour of more up to date versions with digital meters, but the single version is still there.

There's a contact email address on that page. Nothing ventured, nothing gained?

[Daphne_g7ena]

Thank you for your help, the lodestar does look similar and I will fire off an email to them.
I will upload some photos tomorrow.

Daphne

[Daphne_g7ena]

Hi, I have taken one of the boards out, the LH one which had been unplugged.

There is a 5 watt resistor that has been removed and relocated. The transisitors are 2x per side.

Both boards look identical.

Daphne

[Daphne_g7ena]

A few more pics

[SiriusHardware]

The photos seem to confirm that the main active devices in each half are the 723 regulator IC, 324 quad op-amp IC and 2N3055 output transistors. The little extra PCB may be something to do with the dual tracking feature - not sure if that 8-pin JRC IC is just a 741 op-amp or something a bit more exotic.

We could just about reconstruct the circuit diagram of one of the main PCBs from your photos given enough time - I assume you won't have heard back from Lodestar yet. We should probably give them a chance to respond, as the official diagrams would obviously be a great help and save a lot of time.

In the meantime can you check each of the output transistors separately and do resistance / diode checks c-e, c-b, e-b both ways, looking especially for collector-to-emitter shorts as that is the most likely failure mode for the output transistors.

[buggies]

Originally Posted by Daphne_g7ena

There is a 5 watt resistor that has been removed and relocated.

Looks like a standard factory "adjustment".
I think they use the same PCB in my XG90X.
If I ever had a user manual it is hiding well.

[SiriusHardware]

Ah! Maplin's version of the single version of the Lodestar PS-303 (Without the 'D').

I take it yours works, Buggies, that could come in useful. I suppose that flying (but very low value) chunky resistor must be the current sensing resistor since it seems to be in series with the feed to the positive output terminal.

[Daphne_g7ena]

Hi,

I will check the transistors once I get the chance, might be the weekend now.

There was a C1384 transistor which i was able to remove by wobbling it, so that wouldn't help

Daphne

[SiriusHardware]

If you can also check the on board transistor(s) and everything which looks 'diodey' that would be useful too.

[Diabolical Artificer]

They don't look too complex and an approximate circuit could be found by looking at the 723 datasheet and application notes as well as 723 based PSU's, it may be quicker to just draw the circuit yourself.

In the absence of a schematic , starting from the AC in, bridge rectifier, check Q's/diodes for shorts, zeners for correct V, swap opamps, check sense resistors and give any preset pots a good few turns each way keeping note of original position. Hopefully the 723's will be ok, again check he datasheet for an idea of what's what.

Nice Toshiba 3055's BTW, Andy.

[buggies]

Originally Posted by SiriusHardware

I take it yours works, Buggies, that could come in useful. I suppose that flying (but very low value) chunky resistor must be the current sensing resistor since it seems to be in series with the feed to the positive output terminal.

Yes - the ex-resistor was the current sensing device (before it fell apart during board handling )
The current limiting function of the 723 is not used - the function being performed by one of the LM324 op amps. The other switches the relay on when the output voltage is increased to select the full secondary winding. The remaining two op amps are unused.

The attached circuit is about the best I can get at the moment as supposed to be preparing living room for new carpet. Should be enough to go on - apologies for missing values (and any errors).

[SiriusHardware]

Wow, a great job in such a short time. What's the circuit drawing package? I usually have to resort to Windows 'Paint', which means creating all the symbols by hand of course.

[buggies]

Its an old windows version of Orcad. I used the DOS version for a couple of decades as part of my job and inherited this version as I wound down for retirement about 10+ years ago. Library component creation is always a chore - best delegated to someone else! These days the freeware Kicad seems popular but too much of a learning curve for me.

[Daphne_g7ena]

Thanks everyone for your input.
I have given the board a thorough checking over and several issues came to light.
On the Left hand board there were signs of overheating on one of the resistors, although it measured okay. One of the 2200uF caps had such a bad dry joint the capacitor wobbled.
The transistor marked Q3 on buggies excellent diagram was loose.
Capacitor C6 appears to be orientated the wrong way according to the PCB silkscreen but looks okay on buggies drawing.
The pass transistors and diodes checked out okay.

On the Right hand board transistor Q3 checked out as faulty. Diode D10 was burnt and open circuit. The silkscreen for C6 is the same as above.
Some of the ic socket pins were corroded (Verdigris).

I have replaced the ic sockets and will replace both Q3 transistors once they arrive. I have replaced D10 and re-soldered all the joints on the PCB
I haven’t got round to checking the tracking circuit yet.

Daphne

[buggies]

Originally Posted by Daphne_g7ena

Capacitor C6 appears to be orientated the wrong way according to the PCB silkscreen
Daphne

I can't blame the pcb layout guy for that - I had to triple check it - hence the Vout+ and Vout- net labels placed near C6.
If you identify any of the unmarked resistor values (I have trouble with some of the colours especially on the 1% types) post here and I can re-annotate.
I should get my 0R2 resistor from CPC tomorrow so will be experimenting with the presets...

[buggies]

Replaced broken 0R2 resistor and played with presets.
Discovered error on version A circuit - I had renumbered all the components forgetting that the pcb had idents for the presets.
Attached version B has corrections and additional notes on the five preset functions (empirically determined...)