-
Hi everybody,
I am working on a power supply based on 8 smaller power supplies that are either put in series or in parallel (or a mix of both). The total output power is 10 kVA and the maximum voltage is 2000 V. Each of the smaller supplies can deliver up to 250 V and up to 5 A. Before I get a lot of answers regarding the problems of putting supplies in parallel, the work contains a CAN bus based control circuit with master controller and slaves on every unit, and while that may be an interesting discussion topic it is not my concern right now. At the moment I try to find the best solution to switch these supplies either in series or in parallel. Here are the basics:- There is no need for hot switching. All switching can be done while the power is off.
- The power can be slowly raised. No need for a rapid turn on.
- The output power needs to stay on for several hours.
- The 8 supplies shall allow the following configurations:
- 2000 V, 5 A
- 1000 V, 10 A
- 500 V, 20 A
- 250 V, 40 A
" alt="" class="bbc_img" />
My solution so far involves two kind of switches. The serial switches (orange) between the power supplies and the parallel switches (green) from power supplies to bus bars.
For the serial switches I am currently testing a solution based on MOSFETs. 300 V MOSFET with Rdson < 10 mOhm are available around the $20 level. At the maximum current of 5 A the power dissipation is 1/4 W. I would prefer a higher voltage rating but this is not hot switched and I don't expect any large transients.
For the parallel switches the situation is different. They need to block 2000 V and they need to carry 5 A. I have considered:- MOSFET: Not really a great option for 2000 V. I have considered serial MOSFETs but I prefer not to go that route for safety reasons.
- IGBT: The units with Vce > 2500 V typically have saturation voltages of 3 - 4 V, especially at higher temperatures. That could easily lead to 20 W heat dissipation for each switch.
In a fully parallel configuration 14 switches are involved ==> 280 W heat dissipation. Now compared to 10 Kw output power that is not too bad, but I still need to deal with the heat. - Reed relays: I found types that carry 4A and have 50 mOhm on resistance max. I would have to use two in parallel for each switch. The benefit is the low power loss with .625 W/switch. The drawback is the price of >$40/unit. That would be $80 per switch or $1100 just for these switches.
Currently I prefer the IGBT solution but I would like to hear your opinion on best practices here. There must be other applications out there that have similar requirements. Any hints on literature or parts are highly welcome. Also, I have never used reed relays before. Are there any issues that would make these less reliable? I am thinking about vibration for example.
Many thanks
-
If you put supplies in series, have a big diode in parrallel with each of them. In case they loose power, all the other supplies will drive full negative voltage on their output, and that blows up their output capacitors. Oh and check the voltage isolation rating, because this is not their intended use case. IMHO, you want relays, everything else is just asking for problems.
-
IMHO this setup is better build using an adjustable power supply (or several in series) suitable to work up to 2kV. One of the major problems I see is sourcing relays that can deal with breaking 2000V DC at several amps. Those are not going to be cheap or small. Or you have to be absolutely sure the relays won't open while there is current going through them. But then there still is the 2kV DC isolation requirement. Reed relays are probably the worst choice here; they are likely to blow up.
-
Ordinary relays are rated for a few kV transient, in that they have to deal with mains voltages, surge included. But you won't ever find such a DC rating, and only see it attested rarely as such, with a "250VAC" or whatever rating being often the most you will ever see (digging into datasheets or appnotes, you may uncover category rating or the like, which determines surge rating).
So, one could make such an assumption, and proceed. Hopefully just for a one-off sort of thing, with adequate shielding and protection in case switching occurs under load, and arcing and ignition occurs in all relevant relays.
But that's not a great assumption to make.
Better is to get proper relays. They will be quite specialty. Notice you aren't shopping for DC relays, necessarily: those involve special ratings to open and close at up to the ratings. The voltage and current are not so much intrinsic limits of the contacts themselves, but a type rating based on underlying assumptions involving a range of load characteristics (inductive, resistive, inrush, etc.) and the switching thereof. All the design, testing and certification that goes into a given relay, is justified by its mass-market production: make a few hundred thousand, or billion, of something and you can pay off that NRE pretty well. Only making some thousands of some specialty part, though, the sale prices go waaaay up, even if the certifications and such go down.
So, you could ask mfg reps if they can recommend a part, or a part under alternate use given the specific operating conditions of your system, or perhaps even have them come up with something, not necessarily a customized part as such, but just that it's had additional back-end engineering done to check it out for such a purpose only -- but you will likely find few results with the former, and great expense in the latter, if any at all.
It's probably not quite as hard these days, as medium voltage DC applications become mass-market: solar arrays have comparable ratings, for example. (Is this perhaps even a solar panel simulator application?) And, keep in mind there are minor manufacturers, and traditional sales channels, offering somewhat-specialty industrial products in low to modest quantities (thousands/yr ballpark). Usually regional, so, I can't exactly make a suggestion, but they are out there. You might be able to fish around some of those suppliers -- their salespeople are quite knowledgeable about their suppliers, and have close contacts with their reps/FAEs.
As for the overall project, whether such a solution is more economical overall -- in more general terms that is, such as project timeline/completion, or operating costs -- compared to the identified solutions (MOSFET/IGBT, give or take heatsinking), that's something you'll have to figure out.
I will strongly encourage mechanical contacts: do not ignore fault conditions. Inevitably something's going to short out the output, and all your supplies' capacitance discharges through the poor switches. In a series stack, some will be reversed (depending on their relative ESR and capacitance, or inductance even), and then you have to handle ungodly currents in reverse as well; fusing or (supply) control doesn't do you anything because we're talking ~ms time scales here. MOSFETs are useless for such (100s, 1000s A) currents; they can be turned off fast enough to protect themselves (10s us), but keep in mind the necessary overhead to handle the corresponding transient voltages plus clamping into a TVS (MOSFETs are rarely avalanche rated to high currents, and rarer still, repetitive avalanche), which also needs to be high enough to handle whatever expected or external surge/overvoltage might occur (whether because of switching or external surge). Mechanical contacts are extremely robust: the on/off ratio is literally unbeatable, so it should be no surprise that the surge capacity (current when on, voltage when off) is similarly excellent.
Also not to mention, keep in mind the possibility that such transients (~kV sparking) knock out the CANbus or what have you. At least momentarily (wasting a retransmit cycle), if not persistently (outright damaged transceivers). (The latter should at least be easily avoided with appropriate protection and isolation, but that does need to be verified to some extent e.g. ESD and surge testing.)IMHO this setup is better build using an adjustable power supply (or several in series) suitable to work up to 2kV. One of the major problems I see is sourcing relays that can deal with breaking 2000V DC at several amps.
Breaking isn't required.
Tim -
In general, one should wire voltage sources in series and current sources in parallel.
As discussed above, a reverse diode should be used to protect each individual voltage source.
If you must connect voltage sources in parallel, you must see carefully how to adjust the current settings of the individual units.
Most "power supplies" are voltage sources: an important exception are constant-current supplies for LEDs, etc. -
Better is to get proper relays. They will be quite specialty. Notice you aren't shopping for DC relays, necessarily: those involve special ratings to open and close at up to the ratings. The voltage and current are not so much intrinsic limits of the contacts themselves, but a type rating based on underlying assumptions involving a range of load characteristics (inductive, resistive, inrush, etc.) and the switching thereof.
That is only half of it. The main problem with switching DC is that the arc doesn't stop by itself like it does with AC. So if you open a relay while it is carrying several tens of volts DC (ballpark over 30V) and a few amps, the relay will just arc over and melt. I have destroyed 70A automotive relays this way during testing while developing circuits to use relays way above their rated voltages. Fuses for DC can be similary problematic if the load is somewhat inductive. The voltage spike can cause the fuse to arc-over and creates a nice ongoing arc which then sets the entire system on fire. Still writing about <60 Volt system voltage...Quote
The OP isn't clear about that. How about operator error and/or doing an emergency shutdown? Using relays above their specification requires a carefull system design with safeguards to make sure the relay can never switch (including due to impulses, noise, faults, wrong commands, etc) when it is outside its switching limits.IMHO this setup is better build using an adjustable power supply (or several in series) suitable to work up to 2kV. One of the major problems I see is sourcing relays that can deal with breaking 2000V DC at several amps.
Breaking isn't required.
However, relays rated for several kV DC and currents up to several tens of amps (DC!) do exist. At the cost of around US$400 / US$500 each though. But they could be worth the money.
Actually, having a good old thyristor as a crowbar in case things go wrong isn't a bad idea (and a properly rated ceramic wirewound resistor in series to dump some energy in). Simple, cheap and effective.
-
Thanks for all the input.
It gives me a lot to think about. The safety aspects are extremely important.
The overall system has a contactor on the AC side. That is where the emergency shutdown goes to. Now that wont prevent the capacitors from discharging and potentially causing some serious damage but it will kill the system. We do have diodes in parallel to the output of each power supply and as I mentioned we are aware of the sensitive current control issues with paralleling voltage source power supplies. As one of you pointed out a fault may influence or fully disable the control system. The consequences of that so far have not been thoroughly analyzed.
Apparently you all like relays for this task despite the problems and cost associated with them. I will have another look at relays for this and have some conversations with manufacturers. While the relay should never open under normal operation I am worried about the arcing potential in a fault case. I like the SCR crowbar idea although it creates the problem of a possible misfire.
Thanks again! -
Start by making a 250V, 5A module with an output current clamp , set to 5A, so that it will simply reduce its Vout if it trys to supply more than 5A. So with a 90-265VAC input, that’s a PFC’d LLC converter.
You can parallel as many of them as you like.
You can also put them in series, with the described output diode (by others above)
Then put them all in a holder…bring all their outputs to your relay matrix. Have a “connection selection board”….you select what you want here, and then when you press “return”, it first shuts all the PSUs down. Then it discharges all of their input/output caps, then it switches the relays to connect them however required……then, once all the relays have switched, it will allow you to press “power up”.
-
Hi Faringdon
You gave a very nice description of the project here. Almost fully accurate to our project. Each module uses a PFC and a PSFB converter and is controlled by a microcontroller that is connected to a CAN bus. The output current and voltage of each module is continuously monitored. There is a current limit and current balancing between the modules. When the output configuration is changed all power supply modules first turn off, then when the voltages are low, the relays switch and the unit turns back on.
My question was really about the elements in the output switching matrix not about the whole project. It seems that everybody agrees that relays are the way to go here. I am uncomfortable to use relays way out of their specification and need to check a bit deeper what is available. If a relay opens due to a fault condition during operation it will likely arc and kill the relay. That is not nice but not the end of the world. More important is whether it creates a hazardous situation. That is not acceptable. -
I am just wondering, can you put some sort of TVS across the relay, or a relay snubber.
-
I am just wondering, can you put some sort of TVS across the relay, or a relay snubber.
That is certainly an option. -
at these power/voltages, safety and fault protection is paramount.
The devce to switch depends on the tye of load eg indictive/resistive/active.
I would use a contactor not a relay and avoid any solid state devices.
If the chengeover is done seldom, just use some heavy duty connectors, eg Superior lab plugs rated at 3kV 100A.
Jon -
just my 2cent.
I would avoid the whole matrix switching stuff, because complexity of HV switches.
A simpler solution is to have the dynamic of the supply such that it can give out constant power from 31V to 250V, then simply serialize them.
31V/40A up to 250V/5
Should be feasible. -
Dutch66:
A power supply (linear or SMPS) has a transformer with an insulation rating primary to secondary for working and maximum test voltage.
The insulation rating depends on the material, thickness and method of assembly eg split bobbins, mylar tape, wire insulation.
Your idea of several power supplies outputs in series is poor engineering practice and potentially a dangerous failure.
Depending on the low side return path, load and fault possibilities, your power supplies after the one closest to earth/return have insulation voltage far beyond the expected ratings.
Just look at the insualtion V P-S on each PSU for all in series, and short last HV to earth.
With 2kV and 5A lethal voltages, you need to consider the risk to life and property, effect of faults and failures, insurance liability etc.
Thus WORST CASE/FAIL SAFE design is obligatory even for a one off job.
Commercial HV supplies avoid these issues and have many fault protections.
I refer you to exisitng commercial multivoltage HV PS designs e.g. Del, Bertran, Stangenes, HV engineering.
Bon Chance,
Jon
-
see markup
-
Breaking 2KV 5A with a relay? Good luck with that!
-
Transmission and distrib breakers, vac, SF6, air, oil
Jon
-
Breaking 2KV 5A with a relay?
Op specifically said no hot switching.QuoteYour idea of several power supplies outputs in series is poor engineering practice and potentially a dangerous failure.
Parallel, series, both OK.
Series has the advantage that you can use much more standard components.
Of course, the insulation has to be in line with safety norms, etc.... considering safety factors for overvoltage faults..... Hint : not easy.
(did some 60kV stuff in pretty small housings with all the safety normative checking...) -
Hi Johnpaul,
Dutch66:
A power supply (linear or SMPS) has a transformer with an insulation rating primary to secondary for working and maximum test voltage.
The insulation rating depends on the material, thickness and method of assembly eg split bobbins, mylar tape, wire insulation.
You are of course correct with your statement about the insulation rating of transformers. As I pointed out before we are not talking about the use of off the shelve power supplies. The transformers and power supplies are specially made for this and the maximum voltage is taken into account.With 2kV and 5A lethal voltages, you need to consider the risk to life and property, effect of faults and failures, insurance liability etc.
Thus WORST CASE/FAIL SAFE design is obligatory even for a one off job.
Commercial HV supplies avoid these issues and have many fault protections.
I refer you to exisitng commercial multivoltage HV PS designs e.g. Del, Bertran, Stangenes, HV engineering.
Yes absolutely. The safety implications are by far the most important aspects of the whole design. In the end the design needs to be safety norm compliant. That's one of the reasons why I don't want to use components out of their specified range. I started this thread because I try to understand the options for creating the switching matrix I sketched. One possible outcome would be that there are no options within reasonable commercial limits.
So far pretty much everyone was suggesting to use relays. I looked at even more relay data sheets and I see a class of relays that potentially could be used for this. These are relays with more than 3mm gap between the contacts. They have an insulation rating between contacts of more than 3000V and insulation rating between coil and contact of more than 5000V. They are also rated for 8A carry current. So far so good, but they are not rated for operation at 2000V. This is obvious as any attempt to open the relay under load would result in an arc, destruction of the relay and a potential fire as impressively demonstrated in the video liked by DavidAlfa.
Of course I do not intend to break the relay under load. But things happen and there needs to be a safety mechanism for that. Currently I am trying to address this aspect.
Thank you very much for your comments.
-
Hi DavidAlfa,
Impressive video. That is why we don't intend to open the switch of the relay under load. As pointed out by many contributors here the issue is not with the normal operation but with a potential fault condition. In normal operation the automatic sequence is:- Turn off the power.
- Wait until the voltages have been measured to be safe.
- Switch the relays.
- Ramp up the voltage again.
That is all fine. The problem is what happens when the relay coil looses power due to some fault. Now your scenario comes into play and I am currently looking for a safety mechanism to prevent that. So far I have not come up with anything I consider to be safe. -
Dutch 66 despite the restriction no hot switching, faults can and do happen. EG sys running, momentary mains dip causes relays to drop out and reclose.
But the SMPS holdup time is longer than the dropout and the substantial stored energy in the many SMPS bus caps is discharged via the openeing relay contacts.
The "relays" you want are for multi KW TX, with wide gaps, arc blowback etc.
NOT a relay but a contactor.
Worst case is all PSU in series, faults from earth to top PSU out.
Jon
-
Hi everybody,
I am working on a power supply based on 8 smaller power supplies that are either put in series or in parallel (or a mix of both). The total output power is 10 kVA and the maximum voltage is 2000 V. Each of the smaller supplies can deliver up to 250 V and up to 5 A. Before I get a lot of answers regarding the problems of putting supplies in parallel, the work contains a CAN bus based control circuit with master controller and slaves on every unit, and while that may be an interesting discussion topic it is not my concern right now. At the moment I try to find the best solution to switch these supplies either in series or in parallel. Here are the basics:- There is no need for hot switching. All switching can be done while the power is off.
- The power can be slowly raised. No need for a rapid turn on.
- The output power needs to stay on for several hours.
- The 8 supplies shall allow the following configurations:
- 2000 V, 5 A
- 1000 V, 10 A
- 500 V, 20 A
- 250 V, 40 A
If I were approaching this design, I would design the transformer to be capable of 2000V and 40A on the secondary side. Not terribly easy, but easier than handling all of the conditional switching. The primary would be designed for the max 10kVA at the 240V mains voltage. I would then use the SMPS control to adjust the output voltage to the desired setting. You could implement adaptive over current protection based on what the average primary voltage is (lower current as the output voltage goes up). This is just my first pass thought on the requirements, so take it with a grain of salt.