Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[burakkucat]

It was late yesterday, Tuesday 16th June 2020, and I was connected to my LAN via a 2.4GHz WiFi link. The build of two different kernel package sets had been started using build systems in the USA. As is my normal technique I had two separate ssh sessions open and each build had been started in a logging screen session . . . just in case anything happened to the links.

Thinking that a quick check of the kitz forum would be worthwhile, I moved to the Firefox tab and left-clicked on the appropriate option. After a moments pause, I was greeted by the Firefox generic error of "not being able to connect to that site". Expecting it to be the result of some hoster's non-notified maintenance, I went to deal with some e-mail messages that had arrived earlier in the evening. An "unable to connect" message was displayed. Hmm . . .

Eventually my eyes focussed upon the absence of the "connected Wi-Fi" icon. Grumpily, my WAP was selected from the pull-down list and a left-click was actioned thereupon. It timed out as "failed". With a minor vocalisation, I moved across the distance of four feet to examine the ZyXEL VMG1312-B10A which was providing the WAP end of the WiFi link. I was greeted by a totally black rectilinear block. No LEDs were illuminated. "Now What?", along with other words, were muttered.

Turning the VMG1312-B10A off, waiting 60 seconds and turning it back on again had no effect. It was still a black rectangle. After collecting a DMM from the grotto, the PSU plug was removed from the device and the voltage presented was checked. 0V DC. After extracting the PSU from the mains socket it was closely examined and sniffed. Nothing untoward was observed (or smelt). Perhaps checking the resistance between the live - neutral, the neutral - earth and the earth - live pins might give a hint? The result was infinity when the earth pin was involved. Between the live and neutral pins asymmetry was seen . . . 1.81 MOhm with positive probe on the live pin & the negative probe on the neutral pin and 2.16 MOhm with the positive probe on the neutral pin & the negative probe on the live pin. Hmm . . .

[A Digression]
With the DMM set to the 20 MOhm range a consistent (& repeatable) 14.4 MOhm resistance was determined between both -- er -- "front paws".
[/A Digression}

The (obviously failed) PSU was the ZyXEL supplied original which came with the unit. A powerful magnifying was used to read the small label --

OEM_(R)
POWER ADAPTOR
PART NO: 30-132-1011648
MODEL: ADSO128-D120100
INPUT: 100 - 240V AC  50 - 60Hz  0.5A
OUTPUT: 12V DC  1.0A
FOR I.T.E. POWER SUPPLY LPS
EFFICIENCY LEVEL: V

Unfortunately the dimensions of the barrel-plug are not stated. The length and external diameter are easily determined but the diameter of the inner hole is the problem . . .

I concluded all of my work, late yesterday evening, by borrowing a little bandwidth from my neighbour's WAP. 

Finally, a few queries:

• Is the resistive asymmetry across the live and neutral mains input pins to be expected for a switch mode PSU?
• Does anyone have the means to measure (precision callipers) the diameter of the inner hole of the barrel plug (or the central pin of the socket) of a ZyXEL VMG1312-B10A?
• Does anyone have a suggestion as from whence I may source an appropriate replacement PSU, please?

[tubaman]

I'm not sure of the sizes, but do know that a BT Homehub 3 Type A PSU is the correct spec and fits just nicely, as that is what my (currently in the spares cupboard) VMG1312-B10A uses.
The 'Type A' bit will be important here.
There should be plentiful supply of these on a certain auction site for not much money as they are not a desirable device any more (if they ever were?).
The PSU I have is this model - https://www.ebay.co.uk/itm/BT-HOME-HUB-3-TYPE-A-and-BUSINESS-HUB-3-ORIGINAL-POWER-SUPPLY-UNIT/133436538865
 

[g3uiss]

Maplin used to supply a PSU of the correct voltage, with a variety of end plugs. Most combinations were catered for and an easy way to change polarity. I’ve one here as a spare and one of the ends fits my spare Zyxel

[banger]

PSU Plug for VMG-1312-B10A O/D: 5.5mm I/D: 2.0mm

[sevenlayermuddle]

I currently have two functional SMPSU bricks within arm's reach, as well as a DMM.

The first, a fairly aged Apple phone charger, fails to register a Live-Neutral DC resistance reading on the meter's highest 20MΩ range, ie it is 'infinity' as far as the DMM is concerned.

The second, a PSU brick supplied for use with a Billion router, reads circa 3.5MΩ with probes one way around, circa 2.5MΩ the other.

I have never researched typical schematics for such devices, so I am unable to meaningfully comment on whether these observations surprise me.  Or even to comment on whether I actually understand the operation from inspecting the schematics.   

My gut feeling would be that neither of above results would be surprising to a currently (no pun intended) skilled and practicing electronics engineer.

[burakkucat]

Thank you all for your comments and suggestions.

PSU Plug for VMG-1312-B10A O/D: 5.5mm I/D: 2.0mm

That is very helpful.

[sevenlayermuddle]

Thank you all for your comments and suggestions.

That is very helpful.

Think nothing of it.

[Weaver]

Unfortunately I had only just read this post. I must have plenty of spare ZyXEL PSUs. Just shout if you still need one and I will ask Janet to dig in my spares box. I always have a lot of new or second-hand 1312-B10As for emergency lightning swap-out purposes.

[Alex Atkin UK]

They also seem to use the most common 12V PSU barrel jack size, unlike the stupid Home Hub 5A that uses an uncommon size.

[burakkucat]

Unfortunately I had only just read this post. I must have plenty of spare ZyXEL PSUs. Just shout if you still need one and I will ask Janet to dig in my spares box. I always have a lot of new or second-hand 1312-B10As for emergency lightning swap-out purposes.

Thank you. But I would not wish to deplete your stock-holding.

An update . . . Yesterday I checked three listings on a well-known auction site, all were appropriate for my requirement. Today I had an offer from one of the vendors, who had reduced the price of the PSU by one pound sterling -- which I accepted.

[burakkucat]

Early this afternoon a packet was delivered by the Royal Mail. It contained the PSU.

Checks were made . . .

Across the live and neutral pins a resistance of 1.7 MOhm was measured. It was a symmetrical measurement; swapping the DMM probes relative to the pins showed no change in the measurement.

Upon testing the output was seen to be 12.2 V DC, off-load.

All is good.

[tubaman]

Did you manage to find an original Zyxel unit or a same specification replacement?
 

[sevenlayermuddle]

I felt compelled to recheck my facts, and guess what?  My Billion PSU now measures 2.5MΩ symmetrically.  It is possible that the asymmetry may only occur on Wednesdays, or also possible I made a mistake. 

Giving it more thought however, I’m inclined to speculate that the symmetry of live-neutral resistance probably depends not just upon on the day of the week, but also whether the input rectifier is full wave, half wave, or broken full wave (with one or more diodes failed).

I’d personally be feeling a temptation to dismantle the faulty brick, and see if enough circuitry can be exposed to establish the reason for asymmetry.  But not on a Wednesday.

[burakkucat]

Did you manage to find an original Zyxel unit or a same specification replacement?
 

No, neither. I took your advice and sourced a PSU for a BT Home Hub 3, Type A.

I’d personally be feeling a temptation to dismantle the faulty brick, and see if enough circuitry can be exposed to establish the reason for asymmetry.  But not on a Wednesday.

If I was N years younger, I would definitely do so. That typed, I have not found any reliable schematic diagram of a switch-mode PSU . . . The Wikipedia article is devoid of anything useful in terms of a schematic diagram.

[sevenlayermuddle]

I found a few schematics, all of which seemed to begin with a half or full wave rectifier converting mains to DC, then smoothing it with a capacitor, which I understand.   

https://en.m.wikipedia.org/wiki/Rectifier

Then follows the harder bit...  the switch mode elements converting the derived DC voltage to another lower voltage by generating rapid Switching pulses that transfer energy by storing it in inductors, from which the energy is recovered to provide different voltages, with high efficiency.  Since I have a (long ago) degree in electronics I probably ought to understand that stuff too but tbh, not sure I ever did, it always was a bit of a mystery. 

I’ve never conducted a post mortem on a deceased switch mode PSU.   But I have seen a few low tech transformer-based linear PSUs - these also have the same rectifier arrangement, but on the low voltage side of a transformer.   As often as not I find the smoothing capacitor has died, as capacitors often do, taking with it at least one of the diodes in the bridge rectifier.  A full wave bridge with a broken diode would most probably look asymmetrical to an ohmmeter at its AC side...