EEVblog Electronics Community Forum

EEVblog => EEVblog Specific => Topic started by: hans on April 18, 2012, 03:41:15 am

Title: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: hans on April 18, 2012, 03:41:15 am
Great vid dave, I like the older high power supplies. It seems like SMT can downsize everything so much, especially modern MOSFET and inductor technologies.

I've been thinking about the dual proccesor configuration.. It seems to me likely that they do the temperature control and cooling via proccesor. This may be the dual proccesor thing? Maybe the optional remote control is just a bill of materials expansion to the 'right side' of the PCB.

If you look closely at the bottom of the PCB, you can see in the middle there are no traces running across both sections. If you pause at 21:19 you can see optocoupler IC's. That would suggest that one motorola is doing the RS232 , and the other is doing the system monitoring and control. The RS232 port is isolated from the 300V supply, because well.. otherwise maybe have 300V on the PC controlling this thing (oops!).

I also looked up the bigger capacitors. One of the largest one from the list were $20 each. Ouch!
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: mikeselectricstuff on April 18, 2012, 04:08:42 am
Processor will have no role in protection (except maybe remote monitoring) - the whole PSU is analogue - thermal shutdown will be a simple stop if over a set temp on any sensor - possibly also fan speed control or on/off.
Serial ports on this sort of kit will always be isolated - I have a Xantex PSU that also has an option for isolated analogue remote control inputs - can't recall exactly how they did it but isolation was an additional option board.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: firewalker on April 18, 2012, 04:14:05 am
Today I took apart a similar device. I was using two Intel P8031AH (http://pdf1.alldatasheet.com/datasheet-pdf/view/107778/INTEL/P8031AH.html) processors. The one (as I could understand) was dedicated for temperature control. It was connected to 2~3 ADC0804LCN (http://www.national.com/ds/AD/ADC0804.pdf) ADCs. My guess for the dual processor is pin count (although there was a P8255A-5 (http://doc.chipfind.ru/pdf/intel/p8255a5.pdf) ic present).

Alexander.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: SeanB on April 18, 2012, 05:19:15 am
Looking at the temp sensors I would say they are PTC thermal switches, all in series, and connected to a trip line ( probably along with overvoltage, overcurrent not controlled by regulation, drive failure or switch breakdown, and fan failure. Probably all powered by a 24V rail generated from the rectified mains input ( I think all the control circuitry is at mains potential, it makes control easier only providing control inputs that are slow via an isolation barrier) by a small aux PSU somewhere on the main board. This all will cause the PSU to trip and operate a relay to disconnect the power from the input rectifier.

The remote inputs are optocoupled to the remote controller, as there are a row of isolators, with only a power rail and a high voltage cap across it. The other side is basically a DAC with digital output, and possibly also able to send the current status and values back across the barrier. This then feeds into the remote control analogue input, which is also likely to be isolated by another barrier. The links in the DIL socket probably allows selection of which input is connected, either digital board or rear panel input.

Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: nitro2k01 on April 18, 2012, 05:37:11 am
At about 5:40 into the video you can see how the black/blue twisted cable is pushing against the rectifier. That can't be healthy with all the potential heat coming out of the thing.
I'm wondering about the DaveCAD'ed layout. Could it be that the last film capacitor and pair of choke coils are actually placed after the rectifier?

As for the transformers, I'm not too sure they're all actually transformers. (Around 16:20) Out of the ones covered in yellow film, two of them appear to have only to wires coming out of them, and I would guess those two are actually filter chokes. Unless I'm just plain wrong and those wires are not for power transfer, but they are leads for flyback windings or something like that.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: T4P on April 18, 2012, 08:22:51 am
At about 5:40 into the video you can see how the black/blue twisted cable is pushing against the rectifier. That can't be healthy with all the potential heat coming out of the thing.
I'm wondering about the DaveCAD'ed layout. Could it be that the last film capacitor and pair of choke coils are actually placed after the rectifier?

As for the transformers, I'm not too sure they're all actually transformers. (Around 16:20) Out of the ones covered in yellow film, two of them appear to have only to wires coming out of them, and I would guess those two are actually filter chokes. Unless I'm just plain wrong and those wires are not for power transfer, but they are leads for flyback windings or something like that.

But because that is a DC-DC converter any theories about filter chokes can be highly possible , remember , the power output is 1200watts !
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: LaurenceW on April 18, 2012, 08:35:40 am
Turn it on, Dave! we want to see sparks :) 1200W - beauty. This must a pretty EFFICIENT power supply - those tiny fans and even the large metal case are not going to be able to shift much more that 100W of raw heat - maybe much less, even.

Looks like the build is 2001

This one DC-DC switcher board PCB is probably supposed to cover several different output combinations. I expect the "apparently short circuited" second output capacitor would have been a second series-wired output cap (like the input ones are) for the 600V version of this PSU (since most electrolytic caps top out at about 450V DC). But it simply isn't needed in this 300V model.

It looks pretty clean and dust free. Has it seen no service, or is it recently serviced, I wonder.

The cardboard is a strange choice, but at least it doesn't melt! Probably there to direct the air flow over the muscle parts.

The main DC-DC switcher may have to be a combination of buck/boost topologies, so that it can produce both over and under the incoming DC voltage. Edit: Ah no - on second thoughts it is probably a flyback design - those are high frequency stepping  transformers, near the output end (and that's where the input/output isolation comes in, of course).

There does seem to be a lot of "optional" control electronics, if that board is indeed optional.  Like your own power supply design, Dave, no doubt the voltage and current feedback loops are analogue, and not relying on any of this fancy digital stuff.

Hmm, 5V TTL devices in the same box as 400V DC at several amps.... <shudder!>

Turn it on.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: retiredcaps on April 18, 2012, 09:02:15 am
O Canada lyrics. It would go nicely with eevblog #256 (Yamaha keyboard).  :D

http://en.wikipedia.org/wiki/File:O_Canada.svg (http://en.wikipedia.org/wiki/File:O_Canada.svg)
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: EEVblog on April 18, 2012, 09:21:48 am
Turn it on, Dave! we want to see sparks :) 1200W - beauty. This must a pretty EFFICIENT power supply - those tiny fans and even the large metal case are not going to be able to shift much more that 100W of raw heat - maybe much less, even.

Yes, these things are designed to be super reliable and run with full service life over rated temp ranges. So presumably very very efficient.

Quote
This one DC-DC switcher board PCB is probably supposed to cover several different output combinations. I expect the "apparently short circuited" second output capacitor would have been a second series-wired output cap (like the input ones are) for the 600V version of this PSU (since most electrolytic caps top out at about 450V DC). But it simply isn't needed in this 300V model.

Yep, almost certainly.

Quote
It looks pretty clean and dust free. Has it seen no service, or is it recently serviced, I wonder.

It has seen use in a lab rack environment, but has never been cleaned.

Quote
The cardboard is a strange choice, but at least it doesn't melt! Probably there to direct the air flow over the muscle parts.

Yep, it would be for airflow. Don't know why that didn't occur to me at the time.

Quote
There does seem to be a lot of "optional" control electronics, if that board is indeed optional.  Like your own power supply design, Dave, no doubt the voltage and current feedback loops are analogue, and not relying on any of this fancy digital stuff.

Yep, it would be all analog loops, including the temp sensors etc.
The digital is just for remote control, monitoring, and set.
It does work without that top processor board, you pay extra for that only if you want PC control.

Dave.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Time on April 19, 2012, 03:26:01 am
I wonder how it lines up with the 30V/200A Xantrex model currently sitting on my workbench. ;)
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Crumble on April 19, 2012, 04:23:46 am
Nice teardown Dave, I love power supplies! What age is it? It almost appears an 80s design due to all the through hole stuff. I would expect to find most daughter boards to have SMD devices on them, especially because it contains so much of them.

Furthermore, because it has no power factor correction its design must have preceded the regulations about it (in the EU these rules have been set in the 90s if I'm not mistaken, don't know how much the Australian authorities care about power factor), or be operated in a industrial environment where these regulations don't apply. That would explain the non-standard connector on it. Connecting so much capacitance directly to the grid causes some nasty interference, that is probably why there is so much filtering in it too. Even in an industrial environment an uninsulated application of this kind of configuration can cause serious degradation or even destruction of parts in other machinery.

The capacitors a probably series connected to make it possible to switch between 100/120 and 220/240V without having redundant capacitance at 240V. When set to 100/120V One terminal of the input power will be connected to the middle of the capacitors and the other one will be half rectified twice for both a positive and a negative voltage. It would explain the dual fuses as well, else only one would have done. It is actually quite clever, because when connected to 220/240V the fuse rating won't be unnecessarily high because they will be series connected, else they will be used intermittently. No separate fuse rating will be necessary for other grids.

One of the relays will probably be involved in this scheme too because a manual switch can cause considerable trouble when set wrong. (Can you imagine the carnage when overloading all of these caps simultaneously when hitting the on switch?) The voltage selector will probably be automated, especially because I don't see any room for a switch near the power input.

I think the 'paper' inside is not as paperish as you might expect. If you try burning it you will probably find the are hardly combustible and the are usually treated to repel water. I also think it might very well be unexpectedly hard to tear a piece off, they are usually treated with some kind of polymer, similar to paper money but thicker and fire retardant.

The rest of the unit appears to be a full bridge resonant mode converter. You noted the controller, and two of the magnetic units only have two terminals, so they are probably inductors. The capacitors you looked up are probably series connected to them. In this location in a converter one would need capacitors with considerable reliability and peak current capability. That is why they used such exotic types. Only very few capacitors are rated for sustained high-frequency, high-current use and have a suitable (low) ESR.

I am very surprised though about the apparent symmetry all over the unit, a single inductor for example would have done fine, especially because there is only one controller. Maybe lower powered units use the same board (0.5U rack mount power supplies appear to have limited appeal so far). Another user noted he had a 600VA unit on his workbench. I wouldn't be surprised when it would have a similar layout.

The dual processors might be to provide galvanic seperation between the unit ground and the serial connector grounding. Another user already noted the optocouplers. If this unit is as high spec as you suggest it is probably possible to have the output float from all interfaces and the power terminals. It does surprise me though it has no routed clearance gaps on the PCB. Especially because the capacitor on the processor board is rated at 1 kV. Most PCBs however can withstand higher voltages than air when built well.

[edit]: mikeselectricstuff already noted most of this stuff. Sorry!

The eight disc ceramics are probably to supply the same (capacitive) seperation between the serial connector and the LAN interface. They probably have low values and high voltage capability, I am not an expert on LAN interface though. Usually they use small signal transformers (the black boxes on LAN cards for your computer).
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: G7PSK on April 19, 2012, 05:57:21 pm
As the main capacitance is after the bridge rectifier the load would surely be seen as unity or near unity by the mains supply and therefore not need power factor correction.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: nitro2k01 on April 19, 2012, 06:12:09 pm
As the main capacitance is after the bridge rectifier the load would surely be seen as unity or near unity by the mains supply and therefore not need power factor correction.
Hmm, correct me if I'm wrong, but doesn't a full bridge rectifier + a capacitor bank use more power during the peaks of the sine waves, and so flatten them and introduce harmonic distortion to the main voltage?
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: mikeselectricstuff on April 19, 2012, 07:22:12 pm
As the main capacitance is after the bridge rectifier the load would surely be seen as unity or near unity by the mains supply and therefore not need power factor correction.
No - it draws power in gulps at the peaks of the sinewave. More modern PSUs power-factor correct using a boost converter on the unsmoothed rectified AC so it draws a steady current throughout the cycle.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: gazza666 on April 19, 2012, 08:04:27 pm
Dave
You missed giving this piece of equipment a good old smell
“Lovely”
 :)
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: G7PSK on April 20, 2012, 03:24:20 am
But is the load from a bridge rectifier leading or lagging.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: ejeffrey on April 20, 2012, 05:35:37 am
Neither.  Leading or lagging are what you get from linear reactive loads.  A rectifier presents a non-linear load that draws a non-sinusoidal waveform.  It is slightly leading due to the filter caps but the bigger problem is the harmonic content due to the diode switching. 
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: G7PSK on April 20, 2012, 06:25:33 am
So if its neither leading or lagging then its unity surely it might pulsing but it's still unity. 
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: mikeselectricstuff on April 20, 2012, 06:51:24 am
Surely a measure of leading/lagging is only meaningful where both voltage and current  are sinusoidal. The current waveform will be far from sinusoidal,  and measuring phase at zero-crossing will not convey much useful information. The power factor will effectively be varying significantly throughout each cycle.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: G7PSK on April 20, 2012, 07:03:57 am
What I was trying to ascertain was if power factor correction was really required on 1200 watts of rectified power supply, it's not quite the same sort of thing as 150KVA spot welder or 200HP motor both of which will give the supply company nightmares. I have supplied alternator sets to people in the past due to the excessive charges the supply company's wanted to make due to poor power factor equipment. 
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: ejeffrey on April 20, 2012, 07:33:45 am
So if its neither leading or lagging then its unity surely it might pulsing but it's still unity.

No.  Power factor is real power / apparent power.  Apparent power is Vrms*Irms while real power is the time average of V*I.  Anything but a straight resistive load has a less-than-unit power factor.  You can prove it as a fun homework exercise :)

You can even calculate power factor for DC.  If you have 1 volt / 1 amp, you have 1 watt real power.  If the current has a 50/50 duty cycle and switches between 0 amps and 2 amps the real power is the same, but the RMS current is sqrt(2) amps.  The apparent power is now ~1.41 watts and the power factor is ~0.707.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Crumble on April 20, 2012, 11:10:26 pm
Neither.  Leading or lagging are what you get from linear reactive loads.  A rectifier presents a non-linear load that draws a non-sinusoidal waveform.  It is slightly leading due to the filter caps but the bigger problem is the harmonic content due to the diode switching.

Very true, the amount of lag of diode bridge rectified current is negligible and hardly influences power factor, the pulsed shape however does. However, the harmonic content of the resultant signal might not be filtered out very effectively and is therefore transmitted through the power grid and that can cause problems in other machines. Due to the extensive EMC filter the high frequency content will probably be rather limited. The low frequency content is usually less problematic for most devices. Some can even function on rough square waves without malfunctioning.

This power supply was apparently in an industrial environment where special agreements can be made with the supplier of the electricity or where local filters can isolate the distortion from the grid. Some rack mount housings have built-in power conditioners and sometimes even UPSs and the devices inside need not be especially 'nice' for the grid. (Older types of these housings were notoriously heavy because of the huge - often passive - power factor correction in them.) The only thing that has to be done is make sure the other devices in the same grid can handle this kind of distortion. Usually this is not a problem.

This kind of construction is used quite often because it is very reliable. The number of parts that can easily fail is rather small, and the ones in use are quite robust. A bridge rectifier contains a large chunk of silicon that can handle huge surge currents without failing (a small Vishay 1.9A bridge rectifier* is specified to handle a 50A (!) surge current and the one in here is probably a much larger 15 or 25A unit). Good electrolytic capacitors are rated for a specific maximum ripple current capability which can be controlled trough good design of the preceding filter. That is probably why is it has so many stages. The inductors are also very unlikely to fail so they can be added without compromising reliability.

*http://www.vishay.com/docs/93561/vs-2kbbseries.pdf (http://www.vishay.com/docs/93561/vs-2kbbseries.pdf)
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Crumble on April 20, 2012, 11:23:56 pm
Dave, don't you think the MOVs are connected to the middle of the capacitors? It would make sense to protect those when one shorts out, especially because over-voltage of even a small capacitor causes quite some carnage.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: mikeselectricstuff on April 20, 2012, 11:37:22 pm
MOVs are not generally included to prevent physical damage, just to suppress transients that might affect the output.
You dreally don't want a 300v-capable PSU suddenly deciding to output its max value unexpectedly.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: G7PSK on April 21, 2012, 01:18:15 am
It is funny that you should say that Bridge rectifiers are robust and don't fail very often. The chargers on my wife's wheel chairs (she has two identical units) were failing at a high rate on average lasting less than a year, I ended up buying them in tens, went through 15 in seven years with the two chairs. The chargers are rated at 24 volt 10 amps the rectifier is a GBPC3504 rated at 35 amps 400 volts. I was replacing with the original parts. I then ordered a batch from CPC Farnell and since then there has been no failure. But now the chars have been replaced and the new chargers don't seem to have the same problem. 
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: SeanB on April 21, 2012, 04:05:28 pm
I have seen passive PFC where it is a massive LC resonant circuit in series with the one mains lead, where the LC is tunes to 3 times the mains, to trap the third harmonic, and as a bonus it increases the rectifier on time while reducing the distortion in the mains.

I have seen bridges fail, 4 diodes are generally more reliable. I have a ferrroresonant ups that draws a near constant current, but which has a near unity input power factor, though the output is slightly distorted, though a sine wave, as it has a 100uF output capacitor.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Crumble on April 24, 2012, 12:10:54 am
It is funny that you should say that Bridge rectifiers are robust and don't fail very often. The chargers on my wife's wheel chairs (she has two identical units) were failing at a high rate on average lasting less than a year, I ended up buying them in tens, went through 15 in seven years with the two chairs. The chargers are rated at 24 volt 10 amps the rectifier is a GBPC3504 rated at 35 amps 400 volts. I was replacing with the original parts. I then ordered a batch from CPC Farnell and since then there has been no failure. But now the chars have been replaced and the new chargers don't seem to have the same problem.

That is a massive failure rate indeed! :o This does sound like poor cooling design to me because they fail so consistently. The specs of the part itself certainly appear to be sufficient. It will dissipate roughly 20 Watts in a configuration like this.* The design might have been dependent on heat-tolerant parts. Some products are more tolerant of over-temperature than others.

*I usually estimate the power by multiplying the average current by 2 assuming a 1 Volt per diode voltage drop, the exact calculation depends heavily on the voltage source impedance and the capacitor ESR. It certainly is a quite rough estimate and is likely to be on the high side.

I have seen passive PFC where it is a massive LC resonant circuit in series with the one mains lead, where the LC is tunes to 3 times the mains, to trap the third harmonic, and as a bonus it increases the rectifier on time while reducing the distortion in the mains.

I have seen bridges fail, 4 diodes are generally more reliable. I have a ferrroresonant ups that draws a near constant current, but which has a near unity input power factor, though the output is slightly distorted, though a sine wave, as it has a 100uF output capacitor.

Love these "spare-no-expense" designs! I usually only see the cheap-ass consumer goods.

Separate diodes can be more reliable, but usually only due to the lower volumetric power density and greater thermal inertia, but there are exceptions:

Electric design
Some designs are just poor, but others can evolve very unpredictable problems that can only be discovered when tested thoroughly or when they fail "in the field". Problems with inductive kick-back from the transformer does exist and can be found when load currents rise and drop fast and only very limited storage capacity is available. Electro-mechanical equipment like DC motors and switches (these need not be mechanical) can cause significant voltage peaks too.

Over-current is usually much less of a problem because a bridge can handle huge surge currents and a fuse will usually blow before the bridge gets destroyed. Over dimensioning bridge rectifiers is also commonly seen as they are quite cheap.

Thermal design
Bridges, especially in high current environments can dissipate substantial amounts of power. Bridges are usually built in housings that can dissipate a lot of heat on their own. When convection is obstructed this dissipation mechanism stops and can kill the bridge. Sometimes the bridge is screwed onto a metal part of the housing it is in, but these metals often have very poor heat conducting capability so the cooling of the bridge is substantially less than immediately obvious. Absence of heat conducting paste can also cause a bridge to overheat and is surprisingly common.

Cheap suppliers
Some no-name or fake semiconductors get built to very poor construction standards and are sometimes incapable meeting their specs. When buying from respected sources like Digi-Key or Farnell it is hardly likely you will ever encounter one. Most are perfectly reliable, but have alternative issues like a rough finish decreasing the contact area and subsequently the heat transfer.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: Crumble on April 24, 2012, 01:02:49 am
MOVs are not generally included to prevent physical damage, just to suppress transients that might affect the output.
You dreally don't want a 300v-capable PSU suddenly deciding to output its max value unexpectedly.
That might be a somewhat too simplified perspective. With this kind of capacity a surge must be really substantial to get through the EMI filter and charge the capacitors enough to affect the output. It is a switching regulator, so the output voltage is well controlled and should not be dependent on the input voltage at all.

Protecting relatively voltage sensitive capacitors can substantially increase reliability. The ESR in large capacitors can cause the voltage about them to peak quite easily when a surge is applied. Usually only a small part of physical capacitance is exposed to the actual surge, with the rest of the device separated by their ESR and ESL.* This leads to an exponential breakdown mechanism.

*Well-built capacitors only have a low ESL, but a high ESR is quite hard to prevent in a practical way.
Title: Re: EEVBlog #268 - Xantrex 300V 4A PSU Teardown
Post by: king.oslo on October 08, 2012, 04:12:50 am
Hello everyone,

I am looking to buy a PSU with similar specs. Perhaps 150V - 8A or so. 230VAC single phase input.

- What should I pay?
- Which eBay search keywords springs to mind?

Kind regards,
Marius