Repair: Click Surge-Protected & Sansai Switched 4-Outlet Powerboards

Powerboards are very simple and inexpensive devices used to convert one socket into multiple sockets on a (supposedly) temporary basis. Every house has a number of these devices, with basic devices simply bussing together the pins and more complex devices offering switching of outlets and/or surge protection.

Given that these items are so cheap, it might be wondered why I would even bother modifying or repairing such items which were never designed for repair. Ultimately, I take it as a challenge, but more importantly, fixing something saves you from buying something new and saves environmental resources as well.

Click Surge-Protected 4-Outlet Powerboard (CLKCPB104S)

Around the house, I’ve found a number of these boards doing duty. This is a fairly old Click branded product, and it turns out, there are a number of variations on it. This one has surge protection with an indicator LED on the board in-between the logo and the first socket. I also have another board with the LED just above the k in Click and without the moulded channels around each socket, which I suspect is a different design entirely.

The rear of the board shows it has the appropriate Australian approvals, and was Made in China. The “discard if supply cord is damaged” notice is likely to indicate that the unit is not designed for repair. The unit is held together by six security screws of the slotted-type with a post in the middle.

The cable and plug on it is intact and just fine, but what’s concerning was this burn mark on the top side of the board.

Generally speaking, most surge protected boards use metal oxide varistor (MOV) technology. This is frequently a blue disc-shaped component which drops its resistance sharply in response to voltages above a threshold, thus “crowbarring” or clipping the surge in amplitude. These tend to wear out over time, through exposure to accumulated surges, which can result in the voltage threshold of the MOV falling to the point where the device is conducting at the peaks of the normal mains voltage. This heats up the device, accelerating the failure of the device. Depending on the design, such devices can fail catastrophically and result in house fires. Understanding the internal implementation would be good to mitigate any risks.

Another interesting thing about surge protected boards is how the surge protection is implemented. Some boards have the ability to cut-off any connected loads when the surge protection fails, others have an annoying buzzer, while the cheapest just have one indicator neon or LED. A further design difference relates to the surge absorption capacity (related to the number and size of MOVs within the unit) and the type of protection offered (i.e. L-N only, or all three modes). I won’t be involved in an argument about which mode is best, although it does appear that three-mode protection does have potential issues of its own especially if the earth potential rises, whereas single-mode protection is likely to be quite sufficient in most cases, especially where double-insulated equipment is used.

As I’m planning an extended stay away from the home, I didn’t want to just keep using the unit without knowing what’s going on. As a result, the first step was to check the unit with power applied.

The unit reports its surge protection is still intact, after over a decade in service.

Connecting it to the power analyzer and a variac, it was determined that the draw of the unit varies between 0.67W at 220v up to 0.81W at 240v. At the extreme of the variac, I managed just shy of 1W. This doesn’t suggest the MOV is failing, although the power consumption is not trivial (in my opinion).

Consider that regulation exists that presently (in places, e.g. EU and certain states of the US) requires devices to have standby powers of 1W or less (depending on the function), and my own testing showed that some of the best phone chargers can manage idle powers of 0.01W – 0.1W, this adds up.

Lets say at 240V, the draw is 0.81342W. The yearly consumption will be 7.13kWh which means a cost of about AU$2 per surge protected strip just to sit plugged in and switched on. Lets say the average house might have six of these units, so AU$12 a year will be spent just keeping the indicator light on (in essence).

When I tried to get the waveform, there was some rather unexpected results – the current line was wavy (and the limited harmonic resolution of the analyzer didn’t help), and there was a phase shift as well. This suggests there is some capacitance (current leads the voltage) in the system.

Luckily for me, I had purchased a set of security bits a while back, just for this sort of occasion.

Looking inside, the unit is rather simple, with all the surge protection mounted on a short paper-type PCB above the plugs just near where the burn mark was. A second unit of the same design taken apart showed a PCB which bowed somewhat. The terminals are made of brass shaped into a “box” shape which provides good contact, and a circuit breaker is wired in series with the whole lot using crimp spade lugs.

The circuit breaker comes from Rong Feng RF-112B rated correctly for 10A/250V AC with line and load terminals correctly connected.

The unit features single L-N mode protection only, as witnessed by the parallel connection of the PCB across the live and neutral with no reference to earth. A single MOV is used, suggesting the surge absorbance capabilities is relatively limited. The MOV is protected by a nearby thermal fuse, rated at 130 degrees C. The input has EMI/spike suppression provided by an X2 safety capacitor rated at 0.1uF. Aside from this, the remaining components are used to drive the LED – half-wave rectified AC is passed through a resistor to the indicator LED. The resistor value cannot be ascertained reliably as the resistor is the culprit, causing heat that has damaged the colour banding and caused burn marks around it.

This board does not disconnect the load upon failure of the surge protection element. In fact, the surge protection element is practically self-contained, and relies on the tripping of the thermal fuse to indicate its failure. Should the MOV (and or capacitor) not be present or be open circuit, the circuit will still erroneously indicate the surge protection is available. No other fuse is provided, so this means that unless the MOV gets really hot or the draw of the PCB exceeds the 10A rating of the overload breaker, the board can sip current and get warm.

Seeing as the burn mark is only caused by the LED resistor, it’s probably fairly safe to reassemble the unit and continue using it with the expectation that the resistor will probably fail open at some future date, or the LED would otherwise go pop if it went short circuit. However, I wasn’t pleased with the standby draw of the unit and I felt that the surge protection offered was limited, I decided to remove the surge protection element entirely – converting the board to a non-surge protected board. I suppose it could be possible to eliminate the LED “circuit” to lose the indication while maintaining surge protection, only to have it become a fire risk later on, but as I’ve got ample surge protection all over the house, it really was not necessary. Many more modern switching supplies are fairly resilient to minor surges, and some even have their own integrated MOVs anyway.

This gave me a chance to examine the components of two of these boards which I have converted.

The safety X2 capacitors filtering EMI disturbances have a hard life, and with their “self healing” capabilities, they tend to degrade by a reduction in capacitance as areas of insulation damage result in the internal metal film electrodes burning away around the insulation. As a result, what should be 100nF registered just 73.39nF on one powerboard, and just 31.74nF on the other, as measured by an Agilent U1733C. The filtering would not be as effective as they degrade, and seems to be indicative of the amount of use these powerboards may have seen, especially near inductive loads which cause spikes.

The LED resistance is a bit of a mystery. One recovered resistor managed to read 41.44kohm, whereas the other just 37.02kohm (~11% difference). This suggests that the resistors have likely drifted in resistance as the tolerance was likely either 5% or 1% at manufacture, and is a result of accumulated heat damage.

Assuming the resistance of 37kohm, the approximate power dissipation on the resistor is V^2/R or 237.3^2/37k*2 = 0.76W. Note that I’ve multiplied the denominator by two as the unit will be operating half wave, and 237.3v comes from 240v – 2v for the LED – 0.7v for the diode. The resistor looks like a 1W or maybe 2W unit, so it’s operating pretty toasty especially considering such products often see 24/7 duty.

I was interested in the fact that there was draw on both halves of the mains – this seems to be mostly the capacitor current with the half wave LED current quite small in comparison to this. The diode did not fail, it was measured and exhibited a regular characteristic.

The MOVs were tested on the Keysight U1461A, and their 1mA breakdown voltages were 481 and 478.1V respectively. As the units are rated for 470V breakdown, these MOVs are actually still healthy, which was an unexpected surprise to me. Good to know, but too late now since I’ve chopped them out.

Sansai Switchable 4-Outlet Powerboard (PAD-054SW)

This is a low-cost powerboard sold at a local variety shop, but sadly, it hasn’t fared well with respect to time. It has a power indicator light, which serves no purpose really, as each of the power switches has its own internal neon to indicate power. The board has no surge protection – the indicator light may just serve to mislead.

Because it was purchased locally, it does also come with approval numbers and markings to allow it to be sold here. It uses the same type and number of security screws as the Click board above.

Amongst its defects is this switch which has entirely split across its front. If it continues to break, the switch could fall apart and expose live mains connections.

Another is the fact it seemingly over-drives the neon bulbs, resulting in rapid loss of brightness as a function of the amount of use a given socket has. It’s also got very low retention force in respect to the plugs inserted into the sockets – the contact area feels small and the plugs easily slide out at the lightest wiggle. I can’t fix that necessarily, but at least I can replace the duff switch!

Power consumption varies between about 0.3W with no switches on, to 1.3W with all the switches on. This indicates each neon and LED draws about 1/4W. The draw is in phase with the mains, so nothing unusual is happening in this powerboard. Interestingly, the consumption of the power LED is about 1/3rd of that of the Click board, which is more in-line which what I would expect.

Opening it up, we see that the unit is made with a bunch of solder connections, a PCB with a lot of wire and brass contacts. The incoming cable is physically restrained, and the circuit breaker is in series with the live. Earth goes straight into the busbar. The live is then run along the top trace of the PCB busbar, with a wire soldered at each switch-point to increase the current carrying capacity of the PCB (or reduce its resistance). The lower trace of the PCB is the neutral busbar, built in the same way, but using brown coloured cable. I suppose it’s no major issue, but it isn’t what I’d expect from a compliant product. The switches are soldered to the PCB directly, which is good as it avoids the possibility of the whole switch-cube being pried out of the front of the unit. Then the switched outputs are wired to the brass contacts which are crimp-terminated.

The switches (model HSK-1) claim to be rated at 10A, which is good to see as many aren’t correctly rated. However, I do have my doubts as to the contact life at that load.

Each of the brass contacts isn’t a nice brass-shaped “box” grabbing the pins as in the Click board. Instead, it’s just a single vertical piece with a slot cut out of it, which is designed to “rub” against the incoming pin along its edge. I don’t see much in the way of sustained springy-pressure against the pin, so a thicker pin can splay the single edge apart, meaning thinner plugs won’t have good contact. The small contact area, which is along the axis of insertion/removal, also results in low contact friction meaning that the plugs can easily slide out. A bad design if you ask me.

Repair involved getting the soldering iron out, desoldering the switch, cutting the insulation between the blades of the replacement switch, inserting it and soldering it in. Unfortunately, as the allocation of the “energized” and de-energized side of the replacement switch is the opposite, the switch rocks the opposite way to the rest of them. The neon also remains on regardless of whether the outlet is powered or not. Still better than a safety hazard, but in terms of power consumption, not the most optimal outcome.

Since the old switch isn’t of much use, I decided it’s probably nice to see what’s inside.

The construction seems pretty similar to the single-pole dual-throw wallplate switch with the exception that this is dual-pole, single throw, with a ball bearing at the end of the spring. As this is a dual pole switch, each pole is also connected to the neon and resistor combination at one side to act as the indicator. If wired with the mains coming in from one side, the indicator is off when the output is off, and on when the output is on. Otherwise, wiring with the mains coming in from the other side, the neon remains continually energized.

Owing to the limited rating of the resistor and the confinement inside the plastic, there was some discolouration of the plastic former which holds the neon. The resistor value appears to be brown-green-yellow or 150k, which seems a little low. Most neon bulbs strike around ~70-90v, so about 150-180V is dropped over the resistor, thus the current is conveniently 1mA or a little higher. The neon bulb would be operating at about 70-90mW, with the resistor dropping roughly 150-180mW of power. This is within the estimated 250mW rating of the resistor (although it could even be an 1/8th watt resistor due to its small size), but it means that it would be operating at relatively high temperatures.

It seems from this reference that normal glow is achieved from about 1uA through to 8mA. A regular NE2 designation bulb has a design current of 0.3mA with a lifetime of 6,000 hours. Running it at twice the design current cuts the life expectancy by eight-fold, so from that, it seems that the indicators in the switches are pushed quite hard and the neon is not expected to last even 750 hours.

There is a wide range of design currents for T2 miniature neon lamps with other designations, with currents ranging from 0.1mA through to 10mA, with generally longer lifetime (25,000h) achieved for the lower current models and shorter lifetime (1,000h) for the higher current models. However, significant variance is seen depending on the manufacturer’s design of the bulb.

Conclusion

Powerboards are pretty simple and cheap items, but that doesn’t mean they can’t be serviced, modified or repaired. It’s good sense to just avoid these things going to landfill if they don’t have to.

The Click unit’s burn mark was traced to a failing resistor that regulates the current to the surge protection indication LED, and not the MOV. The MOV is a single mode protection from L-N, with an EMI filter capacitor that has degraded in service. The MOV is protected by a 130 degree C thermal fuse, which is the primary means of failure protection and indication by breaking the circuit to the LED. The connected loads remain connected as the surge protection device is merely in parallel with the incoming mains. Due to the risk of future failure, and the needless use of power, the board was converted to a non-surge-protected board by removal of the surge protection element.

The Sansai board has pretty primitive contacts, and a broken switch which was desoldered and replaced. The switch was torn down, and the neon appeared to be driven quite hard, with the resistor causing some discolouration of the switch plastic. The indicator LED operated at much lower power compared to the Click unit, although the unit does not have any other “features” in regards to surge protection.

On the whole, the indicators-on-powerboards concept seems to be superfluous, and I dislike it in general, as it results in needless power wastage. In an era where many devices are forced to work below 0.5W-1W standby power, it seems silly to plug them into a power-board with an indicator or a switch for each socket that consumes 0.25-1W per position. While the burden is not big financially, it’s undoing all of the progress being made by the designers of mobile phone chargers, power supplies, etc. It’s also one of the reasons why many modern power supplies no longer have a power-on indicator LED, as the power consumed by that is significant compared to the unloaded power consumption of the supply.

Posted in Electronics | Tagged , , , , | 2 Comments

US Fax Test Services: HP, Brother and Canon

The death of voiceband modem communications is slowly but surely happening. I started Project Fax in recognition of this, as a way to remember one of the technologies, but sadly from the few faxes I have received, it seems I might be too late.

Regardless, we exist in an awkward period where there are still users of voiceband modem technologies, and somehow in an IP-centric world, we have to maintain backwards compatibility in a way that is somewhat reliable (enough) and hopefully not too costly.

To that end, fax diagnostic services are important for us to verify the proper functioning of our equipment. In Australia, Telstra operates FOLDS and FOLDS-B which provide technical information about the quality of your fax transmission. The only other commonly cited phone numbers are those in the USA, so I sought to try them out before they eventually go extinct.

Connectivity

As I reside in Australia, and I’m not travelling to the US anytime soon, I had to devise connectivity to the US to both fax the service and obtain the reply fax without having a physical land-line in my possession.

As a result, VoIP technology (i.e. not optimal for fax) was leveraged to perform this. All the test fax numbers are toll free and calls can be placed to these numbers through free online SIP toll free termination. The one I chose to use was Alcazar Networks, as they forward the CID as set and claim T.38 support. Note that calls in pass-through mode use G.711u in the USA as compared to G.711a in Australia, and setting the incorrect mode can result in call quality degradation due to transcoding.

The incoming line was provided by Callcentric, using their “free” inbound US DID program – the same number that is used for the US incoming line for Project Fax. Incoming calls are auto-answered by Digium Free Fax for Asterisk running on my FreePBX server, converted to PDF and e-mailed to me. This is a “virtual” fax machine to receive the faxes.

While I could place outgoing fax calls using the FreePBX server, I found the process to be problematic and unreliable. I tried using WinFax PRO under Windows 2000 Professional in a VM with a USB-to-Serial adapter with my Netcomm modem of choice, but it wasn’t operating properly today for some silly reason, so I resorted to an old Dell Latitude C600 laptop with Windows 2000 Professional and WinFax PRO using its internal mini-PCI 3com modem. In essence, a real physical computer to do the faxing. This was connected to the Grandstream HT802 ATA to convert the call to IP, sending it directly to Alcazar Networks by configuring it for “no-registration” operation.

To ensure the fax would be sent back to the correct number, the DID was programmed to the incoming Callcentric number, and the CSID was set in the US “domestic” 10-digit format – i.e. xxx-xxx-xxxx without the hyphens.

Hewlett Packard (1-888-HP-FAX-ME)

The most often cited number is this one, and you receive a return fax if your fax to them completes successfully. No technical information is provided about your call.

Brother (1-877-268-9575)

The next most often cited one is Brother’s test line. Again, no technical information is provided.

Canon (1-855-FX-CANON)

This number is operated by Canon, and I only recently discovered it. This one seems to send you back a fax that they have scanned rather crudely, but there is no technical information provided either.

Conclusion

I have managed to successfully, and without cost to myself, test and demonstrate the three toll-free test fax numbers in the US that I am aware of. The reply fax was successfully received and documented. Unfortunately, unlike the Australian services, none of the US services above provide any technical information or guidance as to the quality of your fax transmission – only providing an indication that your fax is successful with a reply even if the call was marginal.

Posted in Computing, Telecommunications | Tagged , , | Leave a comment

Data Recovery: Kingmax Pro 32Gb microSDHC – A Sign of TLC Weakness?

If you think I’m not having much luck with microSD cards lately, you’d be right. The first issue for the year was the Sandisk 128Gb microSDXC card having some unreadable sectors causing the loss of some files. Now, I’ve had this Kingmax Pro 32Gb microSDHC I reviewed in the past cause issues as well.

The Issue

This particular Kingmax Pro 32Gb card was used in a Raspberry Pi, which was kept in the loft area of the house (warm temperatures) along with another Raspberry Pi with a Sandisk 4Gb Class 4 card. Both units were hooked up to a single power supply and were running various radio-based experiments.

Around a year ago, for power saving and because I didn’t require the units to keep operating, power was disconnected and both units remained untouched in the loft.

After salvaging the units, I decided to read out the home folders of both units, archive the data, and reuse the cards for another application.

The older Sandisk 4Gb Class 4 card completed its read-out successfully with no errors. The Kingmax card initially refused to be detected, and required three insertions just to get it to detect. The partition table was read but the rest of the unit could not be accessed.

I decided to let HD Tune Pro try each sector sequentially – it took over 7 hours to have the program “overflow” with errors and abort itself with nothing but the first few megabytes readable.

Not feeling defeated, I removed the card and then plugged it back in and tried again. Miraculously, everything now was able to read out – but only very slowly, taking 1 hour 41 minutes and 42 seconds for 32Gb – an average rate of 5.24MB/s, well below the Class 10 rating of the card. Feeling a little more adventurous, I cycled the card again …

This time it took just 6 minutes 28 seconds to read and has recovered its full speed. No data was lost, and the read-out data appeared to be correct.

Discussion

It seems to me that this behaviour seems to suggest a similar phenomenon to the Samsung 840 Evo SSD problem with slow data access for “old data” stored on the planar TLC NAND flash. It seems that low-cost memory cards are made with planar TLC, and aside from having weaker write cycle endurance, it seems to have lower retention endurance as well.

The issue is with TLC recording, more voltage levels are used, and the read-out is more sensitive to loss of cell voltage. The leakiness of flash seems to have increased with smaller lithography – which is increasingly used to reduce costs of flash memory by reducing silicon usage/increasing yield per wafer. Combined together, retention times of about 10-years formerly claimed are now realistically probably about 2-3 years based on my own experience.

This card was only unpowered for a year, but the difficulty in reading out the data suggests it was close to losing its data. The fact the data “returned” suggests the card’s internal controller was busy retrying and recovering the data, and internally rewriting the worst eroded data first. By the time I had the first complete successful read, it seems likely that the worst eroded data had been fixed, but slowdowns remained due to some weak data that could be recovered “in time” before the card was to issue an error. Post-recovery, speeds increased to normal as the data was likely all rewritten – this consumes a write cycle of its own and with increasing wear on the flash, increases the leakage rate.

An exacerbating factor may have been the slightly warmer environment of the loft area – however, it’s good to note that the Sandisk 4Gb card also used in the same way and stored in the same way did not lose enough of its charge to cause readability issues. This may be controller related in some way, or it may be because it was an older card based on MLC or larger pitch TLC flash.

Regardless, it seems inadvisable to use “modern” consumer grade flash memory for anything more than as a “temporary” storage vehicle for shuttling data around. Long term retention is not its strong point and the price-sensitive TLC units appear to make situations worse.

Conclusion

At least this Kingmax Pro card ended up with a happy ending in that no data was lost. However, it seems to confirm a glaring weakness in cheap, consumer-grade TLC memory – especially those destined for low cost microSD cards or USB sticks. Not only do they have a limited write cycle endurance, they also seem to have data retention endurance problems. Leaving them unpowered for a year seemed to be enough to cause data recall difficulties – chances are, if left unpowered for about two years, maybe the data would all have been lost and the card itself “damaged” due to the loss of internal configuration data.

It points to the whole issue that flash memory is not good for long term data retention. While formerly, retention would commonly be rated for 10-years, whether this is possible under TLC and smaller lithography processes is debatable. The industry seems to point to one year retention at “end of life” as being the norm.

It’s a good time to stop and think about all those special event photos and videos being delivered on USB flash disks and how long they will remain readable for. It’s probably best to back them up to a diverse range of media, and continually “transfer” them to newer media as time permits to stop the “erosion” of time.

Posted in Computing | Tagged , , | 1 Comment