Circuit for detecting open/closed relay/contactor in 400-800Vdc applications

• Circuit for detecting open/closed relay/contactor in 400-800Vdc applications
  Posted by elima on 23 Sep, 2017 00:19
• As per the title, I am looking into designing a circuit that will be able to detect the status of a contactor and give an isolated digital output
  for example. I know that a contactor with auxiliary circuit would do this job easily but unfortunately I can't use such.
  The current can flow in both directions if that matters.
  The contactor is used in 400-800Vdc applications and the currents can peak at 170A.
  
  Any ideas?
• #1 Reply
  Posted by forrestc on 23 Sep, 2017 00:30
• Quote from: elima on 23 Sep, 2017 00:19

  As per the title, I am looking into designing a circuit that will be able to detect the status of a contactor and give an isolated digital output
  for example. I know that a contactor with auxiliary circuit would do this job easily but unfortunately I can't use such.
  The current can flow in both directions if that matters.
  The contactor is used in 400-800Vdc applications and the currents can peak at 170A.
  
  Any ideas?
  

  
  Could you be more clear about exactly what it is that you are looking for:
  
  If you are looking into detecting if the contactor's coil is energized, then you have several options, including an optoisolator powered from the same source as the coil.   Or a magnetic sensor picking up the coil's flux (depending on the design of the contactor, etc).
  
  If you are actually looking for the contact closure itself, you probably will need to do something to measure the voltage on both sides of the contactor and make sure they are the same within some reasonable limits.   This is assuming that there is always a voltage potential across the contactor.
  
  Another thought would be to sense the current through the contactor using a hall effect current sensor.  This only works if there is always a current flowing.
  
  I have many more similar thoughts, but without knowing the exact goal of the sensing, additional ideas would also be similarly 'a shot in the dark'.
• #2 Reply
  Posted by Gregg on 23 Sep, 2017 00:42
• A simple magnetic reed switch sandwiched between two output wires with some heat shrink might work but would depend on current flowing.  Veris Industries makes some split core Hall effect current transducers for DC that might be worth looking into.  Veris is a small company and you can actually call and talk to an actual engineer ( not a "sales" engineer).
• #3 Reply
  Posted by Gregg on 23 Sep, 2017 00:53
• Another option may be a neon opto-isolator.
• #4 Reply
  Posted by Teledog on 23 Sep, 2017 02:22
• current sense transformer/loop
  isolated & ...
• #5 Reply
  Posted by DrGeoff on 23 Sep, 2017 03:29
• Hall-effect sensor?
• #6 Reply
  Posted by elima on 23 Sep, 2017 07:30
• Thank you all for the replies.
  Let me make this more clear. I have a DC source (700Vdc battery for example) and there are 2 separate contactors connected to each pole, one on the positive terminal and one on the negative. I need to be able to detect if any of these contactors stays closed for some reason when it shouldn't (welded contactor). I need to detect the contact closure itself and not the coil.
  Hall effect sensor/current sensing methods are simple and easy but they are not really suitable here due to the fact that there isn't current flowing always(or it is very very small).
  
  Haven't found much about neon opto-isolators but I would guess they have a polarity so perhaps it won't be suitable as well. I should check further though, so I would appreciate if you can point to a good reading source.
  
  Voltage drop across the contactor terminals might be perhaps a good idea but it seems to be quite complex due to reference voltages and so on.
  
  Any more ideas?
  
• #7 Reply
  Posted by oldway on 23 Sep, 2017 11:59
• 700V battery ? contactors ?  You are kidding ? There are serious safety concerns here....a commun contactor is not able to interrupt safely 700V dc....
  And using an electronic circuit to detect the opening of the contactor is not safe at all....
  If you would be qualified to work with such dangerous voltages, you should not need to ask help on this forum....
  Ask for help to a qualified engineer....On this forum, you will have a lot of answers of hobbyist who does not know nothing about high power, high dc voltages nor about safety.
• #8 Reply
  Posted by SeanB on 23 Sep, 2017 12:41
• 700VDC then you better be using a contactor rated for that kind of use, and preferably one that has already built in multiple contact poles and arc separation that will break the current of a shorted battery pack. Then back it up with appropriate DC bus fuses right by the battery ( preferably a few between the individual cell units as well, so that you stand a chance of them actually breaking the fault irrespective of how the battery pack is failed) and then use an auxillary NO contact block on the contactor to show that the contacts have closed, and a NC block to show again that they are open. The NO and NC blocks typically do have a bit of travel on them so they do not overlap while the relay is operating.
• #9 Reply
  Posted by elima on 23 Sep, 2017 16:31
• First of all, please don't do any assumptions regarding violating any safety rules. I am qualified to work with high voltage systems and the systems are following all safety rules.
  As I mentioned earlier, there is one contactor on each pole of the battery. The contactors are rated to work with the voltage/current specifications of the system which are taken with very high safety margins of the capabilities of the battery pack. Moreover, there is a fuse that is appropriately rated protecting the whole pack. Precharge and dishcharge circuits exist as well. So everything is done safely with no risks. The idea about using a contactor without auxiliary contacts as I used to do thus far, is mainly for weight reduction and mainly because there doesn't exist a light contactor for very small currents with auxiliary contacts. I challenge you to find me a light relay/contactor(light weight wise) that is rated for 600V and has an auxiliary contacts that won't be an overkill for a precharge current which in my case is max. 2.5A. The contactors/relays that u will find will be rated over 100A. (overkill, waste of money, extra weight)
  The chances that both contactors will weld at the same time and the fuse won't burn are close to ZERO. The detection of welded contactor is just another safety measure taken to make the system more fault proof.
  
  Now that I hope things are more clear, does anyone have a good idea on implementing such circuit?
  
  Thanks.
• #10 Reply
  Posted by Gregg on 23 Sep, 2017 17:10
• elima, I fully understand what you are trying to accomplish as I have worked on DC machines like huge lathes and milling machines.  If one of your contactors sticks closed you have a potential and real safety hazard. 
  You could possibly make your own neon opto-isolator but keep in mind that the small neon bulbs only light on one of the two internal electrodes with DC depending on the polarity and the neon bulb would have to be ballasted with a proper resistor or better yet several resistors in series and potted to make it less likely for an arc to jump.  I have only seen neon opto-isolators on one large 480 volt AC transfer switch and thought it was innovative.   
  A picture of your contactors and setup may help.  Perhaps a limit switch could be rigged with insulating materials to provide an isolated aux contact.  Industrial limit switches are readily available, some have plastic housings; you could fabricate an insulated actuating arm out of fiberglass or polycarbonate. 
  
• #11 Reply
  Posted by technogeeky on 23 Sep, 2017 17:43
• Quote from: oldway on 23 Sep, 2017 11:59

  ...
  If you would be qualified to work with such dangerous voltages, you should not need to ask help on this forum....
  ...
  

  
  Bullshit and gatekeeping. There could be a whole host of legitimate reasons that he can't / doesn't want to ask a fellow (local to him) engineer about this. Time constraints. Weekend. He's a solo operator at a small company. Just moved. In any case, asking a fellow engineer is precisely what he's doing here right now!
  
  The way health and safety advocates overwhelm discussion on this forum sometimes is very annoying. We all know there are risks. We know that the risks increase with voltage and current.
  
  If you have a safety concern and an answer to the question, then voice your concern, answer the question, and move on.
  
  If you have a safety concern and no answer to the question, then don't bother.
  
  
  As for the original question, my first attempt would be a pair of hall effect sensors. Or, because presumably the current through the contacts is so much higher than the current through the solenoid that it would saturate the hall sensor on that side, use a current shunt instead on that side. So when there is no current through the shunt (through the solenoid), but the hall effect sensor is still high, then you have welded contact situation.
  
  
• #12 Reply
  Posted by David Hess on 23 Sep, 2017 18:11
• Quote from: elima on 23 Sep, 2017 07:30

  There are 2 separate contactors connected to each pole, one on the positive terminal and one on the negative.
  
  Hall effect sensor/current sensing methods are simple and easy but they are not really suitable here due to the fact that there isn't current flowing always(or it is very very small).

  
  So we cannot rely on a current or voltage measurement at one contactor by itself.
  
  Do we care if we detect a failure in one of the contactors without knowing which one?  That might make things easier.
  
  Is there any amount of low current which may be bypassed around the contactors?  For instance is the detection circuit allowed to leak 1 milliamp around each contactor?
  
• #13 Reply
  Posted by ratio on 23 Sep, 2017 18:19
• You're looking to detect in particular a mechanical failure where the contacts fail to open due to, say, welded contacts or a broken spring? A whisker switch set to trip when the armature is released is what I'd look at if it's an open frame.
• #14 Reply
  Posted by elima on 23 Sep, 2017 18:36
• Thanks for the replies. Let me point out that at this stage I do care about knowing which contactor is faulty. If it seems that the complexity of such circuit is really high, then I will consider
  settling down with a solution that will just detect the failure and not the location of the failure.
  Bypass current or what I was thinking of as a test current is allowed. I think we might have the same thinking direction at this point.
  My idea is to inject a test current through the contacts once in a while (at some frequency) but I can't figure out the whole idea of the circuit.
  I am quite positive that there should be a simple solution for this problem.
  
  Just to give an example of a contactor that I currently use for precharge(2.5A max): EV200HAANA http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=EV200_R_TBD_KILOVAC_EV200_Ser_Contactors&DocType=CS&DocLang=English
  
  and what I would like to use for example: https://www.cynergy3.com/sites/default/files/D%20series%202016_5.pdf
  
  I am using this contactor only because I have the aux. contacts that can help me to detect failures. It is a total overkill for my application and I am trying to cut expenses and weight by designing such circuit.
  
  To make things even more clear and to avoid solutions that will the kill the idea of this whole "upgrade":
  1. The system has to be cheaper. Just with using different relay I can save more than 75% (that's a lot!)
  2. The system has to be lighter. With the given examples the weight reduction is more than 90% (not taking into account the needed pcb though)
  3. The system has to be reliable.
  
  
• #15 Reply
  Posted by max_torque on 23 Sep, 2017 19:21
• If your safety case demands a contactor with auxiliary contact status indication, then that's what it needs.. "Cheap" or "light" doesn't come into it.  The reason such things aren't cheap is that they have been rated, tested and validated to do what they claim to do.  If you are going to engineer a DIY solution then to make it as robust as the specific parts designed to do that task then it's also going to be expensive and time consuming.
  
  
  There are numerous isometery systems that can detect welded contactors, most use some sort of small AC signal imposed on top of the DC voltage across the contactor in order to detect the real time impedance of the system.
• #16 Reply
  Posted by oldway on 23 Sep, 2017 19:56
• Do you realize that 700V is as high as the voltage of the third rail of a subway?
  I do not know what is the short circuit current of the batteries, but it is in all cases much higher than the 30mA which is considered as the lethal limit.
  
  For voltages as high as 700V, the only valid safety is a visible disconnection of both poles (disconnector) and a circuit / load earthing.
  All subway cars have such a device, it is mandatory.
• #17 Reply
  Posted by elima on 23 Sep, 2017 20:16
• I didn't come here to discuss about safety issues as my system has been designed by following all safety procedures. I don't need to go into details how the system works and about all safety matter but I can say that safety is not what I wanted people to be talking about in this thread. So thank you all for your safety concerns and worries, but they aren't the problem here. The system is normally de-energized, there is a discharge resistor and there is as well a visible one pole disconnection. There is insulation measurement device and so on..
  
  An engineer is a person who is supposed to think outside the "box" and try to be creative. Using ready made solutions is not really engineering.
  If everyone would just use what already exists, our world would be still stuck with technologies from few centuries back.
  
  Lets stick to the topic of this thread and leave safety issues and concerns aside. I could have perhaps made this problem look much simpler if I didnt mention that it is used in 400-800Vdc. People would then try to think about possible solutions and leave safety aside.
  
  So for those with safety concerns: You have a system that has a low voltage (0.5V) and high voltage (24V). On the high voltage side positive pole as well as on the negative pole you have a relay without auxiliary contacts. You need to detect whether there is continuity on the terminals of each relay (not the coil) and the detection circuitry should be totally isolated from the low voltage system.
  
• #18 Reply
  Posted by oldway on 23 Sep, 2017 21:21
• You posted a question in the beginner section of the forum about a circuit with a battery and a voltage of up to 800V and you don't want to discuss about safety issues ? You don't think there is something wrong ?
• #19 Reply
  Posted by elima on 23 Sep, 2017 21:54
• @oldway. You are not contributing to this thread at all. I haven't asked for evaluation of the safety of my system. The only wrong thing that I see here are your unrelated comments for solving the presented problem(apologies if this sounds harsh).
  I am very far from being a beginner but as I am new to this forum I thought it would be wiser to start here.
  Heard enough comments about safety concerns, but not many solutions to the actual problem. All safety concerns have been looked through and are taken into account in every step.
  Please, no more comments about safety concerns.
  Lets stick to the topic.
• #20 Reply
  Posted by oldway on 23 Sep, 2017 22:08
• I have already 1615 posts on this forum and I don't have nothing to proove anymore.
  
  Some electronic forums forbid posts concerning high voltages because the forum is read by hundreds of thousands of people including beginners.
  
  It is clear that a question about 800V voltages is not an innocuous question and that safety is a major concern for all who are reading this topic.
  You are posting on a public space, we are not talking in personal messages.
  
  So, your thoughts on the help or not that I bring in this topic are totally inappropriate.
• #21 Reply
  Posted by oldway on 23 Sep, 2017 23:39
• Here is a circuit that could work.
  Leakage currents remain below 30mA.
  I supposed that the load has a relatively low resistance.
   
  But I decline all responsibility, you will use this circuit at your own risk .... You say you are qualified to work with circuits of high voltages, you are therefore fully responsible for your acts.
  For the operating principle, the position of the contacts of the 2 reed relays are different if one contactor is faulty.
  For correct operation, the contacts of the 2 reed relays must be in the same position.

  ---

  
• #22 Reply
  Posted by David Hess on 24 Sep, 2017 00:42
• Quote from: elima on 23 Sep, 2017 18:36

  Bypass current or what I was thinking of as a test current is allowed. I think we might have the same thinking direction at this point.

  
  I was thinking of shunting each of the series connected contactors with a high voltage resistor (or 4 high voltage fusible resistors in a series and parallel arrangement for higher reliability) so they can drive an optocoupler with a current or 1 milliamp or even lower.  A quick search found some optocouplers specified down to 100 microamps or an even lower current could drive a relaxation oscillator to pulse an optocoupler (or transformer) at a higher current.
  
  I am a little leery of high voltage disconnect circuits which have such a leakage current however they may be suitable in some circumstances.
  
  

  Quote

  My idea is to inject a test current through the contacts once in a while (at some frequency) but I can't figure out the whole idea of the circuit.

  
  I could see doing that as well and it may be what max_torque referred to.  Use a small coupling transformer with the contactor side AC coupled across the contacts with a high voltage capacitor.  Now an actual resistance measurement may be made with AC current excitation.
  
  This is both safer with no contactor side leakage and perhaps more reliable without any temperamental optocouplers.
  
• #23 Reply
  Posted by floobydust on 24 Sep, 2017 00:46
• Is some quiescent current allowed?
  Or must this be zero-drain circuit when the contactors are off?
  Why does the battery bank "float" with respect to ground? This is unusual and not permitted with 125VDC battery banks I've worked with in buildings.
  
  A sensing circuit will probably add enough leakage current across the contactor's contacts to keep things unsafe. I don't know your application, but there must be a disconnect upstream of the contactor for people to work safely on the HV bus. It sounds like you know this.
  
  If some quiescent current is allowed, I can post some ideas.
  
  
• #24 Reply
  Posted by Neomys Sapiens on 24 Sep, 2017 00:49
• There is a circuit for continuity detection that is suitable for contact with voltages (certified for 600V in form of the Unitest V1x200). It is astonishingly simple, as it was build around three transistors. The´test connection forms part of an oscillator circuit of some sort. I just cannot find it, but I remember to have retraced the circuit one time.
  
  Anyway, it will come probably come out as something using a small AC signal across the contacts. So you would need well isolated supplies for both of them and then use capacitive or inductive coupling in such a way, that you can recover your signal and yet withstand the full voltage, even with a transient on top. So it might be helpful to raise the frequecy beyond the audible range. Any objections to the use of such signal from the other aspects of your setup?