Maplin N93CX (aka Manson NRP-2050) SMPS fault

Maplin N93CX (aka Manson NRP-2050) SMPS fault
Posted by xyberlink on 22 Feb, 2019 08:23
Hi, I have a Maplin N93CX (aka Manson NRP-2050) variable output SMPS that isn’t working. It’s designed to deliver 0-20v at 0-5A, i.e. 100W total.

I acquired it broken, hence the history isn’t known, but it’s in good condition inside and out. Symptoms were that it comes on, but no output.

Initial tests showed that R62 (0.05ohm 5W) near the +ve output was open, hence I’ve replaced that.

I now get output, but it’s severely limited, going into Constant Current mode almost immediately. Example being a 12v 5w bulb – with the settings at 12v and current max, it goes into CC mode with output of 5v at 0.2A. This is similar for other loads – while the voltage / current vary, it’s always delivering in the region of 1W.

Voltage – I can set it through the entire range (0-20v) and output appears accurate.

There is a preset current limiter, which when I attempt to set it (terminals shorted with voltage set to 3v) I can get to about 1.27A on the display before it stops going up. So that’s 1.27A at 0.1v displayed, i.e. again around 1W or so of output.

I’ve checked numerous components (resistors, capacitors, diodes etc.) throughout the circuit, but struggling to locate anything amiss.

I’m also not familiar with which element of a SMPS controls the output limit (something related to the feedback loop?). Not sure if the fault lies in the current sensing area, or if the supply just isn’t able to provide the output demanded.

Service Manual available here....
https://files.elv.com/Assets/Produkte/7/719/71969/Downloads/71969_nrp2050_data.pdf

Any thoughts much appreciated.

Thanks, Paul.

#1 Reply
Posted by CJay on 22 Feb, 2019 08:34
What did you replace the resistor with?

If you've not got 0.05R (or lower) the current limit will be wildly wrong.

Have you owned it since new?
(Edit, ugh, yes, you got it second hand broken, missed that bit.)

Check the current limit pot VR1 to make sure it's not faulty, if you've got an analogue meter they're great for checking potentiometers for smooth operation.

The current and voltage feedback are done by three optoisolators, you can trace back the current limit through U3B, PC3 and back to pin 16 of the TL494, it would be interesting to see what the voltage at the top of VR1 is with no output load and when it's in current limiting.

I'd be (as always, broken record) checking the capacitors as well as all the diodes after the secondary switching transformer.

#2 Reply
Posted by shakalnokturn on 22 Feb, 2019 10:38
+1 on the capacitors, mainly the electrolytics, don't forget the smaller ones (i.e: TL494 decoupling) and ones in hot-spots.

There is one mylar capacitor that is worth checking if any, that's C28 (1µF 250V)

#3 Reply
Posted by xyberlink on 22 Feb, 2019 11:14
All the electrolytics look fine (everything in this supply looks brand new). ESR meter suggests all ok (in circuit at least, but gut feel says they're all good).

R62 was replaced with a 0.05ohm 1% 5w wire wound. ESR meter reads it at 0.05ohm - so hopefully that's good.

C28 (in circuit) reads 0.04uF with capacitance meter. ESR meter shows no reading. Maybe this needs a ? on it?

Both pots (VR1 / VR2) tested and ok - smooth and correct range.

I think current limit is VR2 / PC2, which feed into TL494 Pin 16.

VR2 gives 0.02v to ground with output on (no load) and 0.11v to ground with output on (loaded).

Pins 2 and 15 of TL494 (reference voltage) are at 1.93v and 1.92v all the time.

Pin 1 of TL494 (voltage limiter?) is...
Output off - 1.93v
Output on (no load) - 1.93v
Output on (load) - 0v

Pin 16 of TL494 (current limiter?) is...
Output off - 0.348v
Output on (no load) - 0.348v
Output on (load) - 0.153v

#4 Reply
Posted by shakalnokturn on 22 Feb, 2019 12:15
Quote from: xyberlink on 22 Feb, 2019 11:14
C28 (in circuit) reads 0.04uF with capacitance meter. ESR meter shows no reading. Maybe this needs a ? on it?

Rather than "?" -ing it, test it off PCB, if bad replace, save yourself time searching around.
All the main PSU power goes through there, so if it is right off value you can expect just the symptoms described (limited power output).

#5 Reply
Posted by xyberlink on 22 Feb, 2019 17:50
C28 now removed and readings unchanged, so looks duff.

It's marked up 105k 250VPF.

It's a mylar / Polyester (PET) - got that.
105k = 1uF - got that.

Am I right to say it needs to be 250v AC rated, rather than DC rated?

#6 Reply
Posted by shakalnokturn on 23 Feb, 2019 10:25
Yes, 250V AC but if you have room for a higher voltage rating it should last longer.
Also make sure you go for a model specified for "pulse" use (high dV/dt).

#7 Reply
Posted by xyberlink on 23 Feb, 2019 13:39
Would be grateful for a little theory / education at this point.

I inferred that I needed an AC rated cap, and that's backed up by the last response from shakalnokturn, but I'm now trying to understand why that's the case?

I was under the impression that we're pulsing DC at this point in the circuit, i.e. taking the rectified output and pulsing it into TR4 via a pair of IRF840 MOSFET's, hence trying to understand where the AC element comes from?

Thanks.

#8 Reply
Posted by LateLesley on 23 Feb, 2019 16:00
Any time you start pulsing/changing DC - it becomes AC. Think about it. take an audio amplifier. it sits steady DC, until you feed an audio signal in, at which case you start getting an AC signal out. As soon as any signal starts changing, you start to have an AC component of some sort. So any pulse, is a change in level and this will get passed through the capacitor. ( remember caps block DC, but pass AC )

#9 Reply
Posted by xyberlink on 23 Feb, 2019 16:56
So for the sake of argument - picking the Kemet F461 series. These claim to be suitable for SMPS pulse operation.

Datasheet...
http://www.kemet.com/Lists/ProductCatalog/Attachments/310/KEM_F3028_F461-464.pdf

They do one rated at 250vdc / 160vac and another rated 630vdc / 250vac.

The circuit appears to be switching from the high point (240v) to the midpoint (120v), so 120v range (if I'm reading it right?).

It's primarily DC pulse with some induced AC.

Would the 250dc / 160vac variant therefore suffice (ignoring the benefits of overrating for a moment), or does it need to cover the full 250vac range?

#10 Reply
Posted by shakalnokturn on 23 Feb, 2019 18:43
If you're in the UK and assume 230V mains, once rectified you're at 330VDC.
C28 can be submitted to 330Vpp or 165Vpeak, of course this is only of it has time to fully charge/discharge, that itself depends on pulse width (load).
The dV/dt depends mainly on MOSFET switching times and transformer's inductance.

#11 Reply
Posted by xyberlink on 24 Feb, 2019 07:42
Good poing re the rectified voltage.

I think the queries in my previous post remain valid though. The high point becomes ~330vdc and the midpoint ~165vdc, hence the voltage across C28 should be around 165vdc?

#12 Reply
Posted by shakalnokturn on 24 Feb, 2019 08:01
Yes, that's the voltage thrown at C28, in practice the voltage across it may be less as it may not be given time to charge/discharge to the full voltage.
You probably could get away with a 250VDC/160VAC cap.

#13 Reply
Posted by xyberlink on 02 Mar, 2019 15:03
C28 now replaced with a Kemet R75, rated 400vdc / 220vac, which has resolved the issue, i.e. power supply will now provide the full voltage / current output.

Thanks for the assistance.