spikey Posted February 12, 2017 Share Posted February 12, 2017 I need to power two simple 12v relay modules, each of which lights one of two LEDs to indicate which way two points on my layout have been set, and I need to do this from my Gaugemaster 100 transformer/controller. With no load, the "12v DC" output from the Gaugemaster measures 15v, and I know from experience that the relays won't tolerate much more than 12v. I therefore obtained from the maker of the modules a 12v regulator board, which consisted of an L7812CV (without heatsink) and a couple of tantalum capacitors. This worked well for over a fortnight until it decided to run amok, overheat bigtime and emit thick acrid smoke. So much for my belief that these things had inbuilt protection against thermal runaway. Having overcome his incredulity, the maker sent me a replacement, which I took the precaution of mounting temporarily on the outside of my control panel. This worked well for six hours yesterday, with the 7812 never more than 15-20°C above ambient. This morning it worked well for 2-3 minutes before I smelt it getting hot. I killed the mains PDQ but even so, by the time I did, the 7812 was far too hot to touch. I can find no fault with my wiring, which has not been altered since I installed the first 12 module and subsequently had it working perfectly for a fortnight or so. I have little faith in the firm I've been dealing with coming up with something which is both reliable and above all safe. A quick search on Ebay got me nowhere for a suitable regulated DC voltage converter. I know I can solve the problem with a cheapo regulated 12v mains power supply, but that means yet another plug and trailing wire, plus a new socket on the side of the control panel. I'm up for knocking together my own simple circuit on a bit of Veroboard, but if two of these 7812s have run away on me now, what's to stop another doing the same? Other than into the kitchen to get myself a large glass of wine, where do i go from here, please chaps? Link to post Share on other sites More sharing options...
cliff park Posted February 12, 2017 Share Posted February 12, 2017 7812s do indeed have inbuilt thermal and short circuit protection. They also have a theoretical MINIMUM input of 19V, and a maximum of 35V . The Gaugemaster 100.O has a nominal 14V output at 2.5 amp, the 7812 is rated at 1.5 amp, but should be ample for a couple of 12v relays. I suspect a fault elsewhere in the circuit. Were the relays (either or both) operated at the time ? You say that the 7812 was running quite warm all the time, that seems a tad strange. Can you measure the resistance of the relay coils ? Is the wiring such that one relay is always operated ? Sorry for all the questions but it is hard to imagine quite what is causing this. Link to post Share on other sites More sharing options...
cpman46 Posted February 12, 2017 Share Posted February 12, 2017 Two 12V relays say at 35 to 40mA ea plus two LEDs at say 20mA ea should put a load of no more than 120mA on a 7812 1.5A device with both relays operated and two LEDs lit. Can you give a bit more info re the relay modules in terms of a simple circuit diagram and how they are wired to the LEDs? What drives the relay input? I have no knowledge of the Gaugemaster 100 but I wonder how smooth the '12V DC' output is. If there is little or no smoothing i.e. the O/P is raw rectified ac then some smoothing will be required for the voltage regulator to work. Link to post Share on other sites More sharing options...
Brian Posted February 12, 2017 Share Posted February 12, 2017 Just for clarity can we determine which Gaugemaster controller you're using? There are two - Gaugemaster 100MO or the smaller Gaugemaster Model 100M. The latter offers uncontrolled DC at a nominal 12 volts at 1.0Amp while the former is 14volts DC. Though either type are fine to use. Assume you have used the DC outputs on either type? Also you're not using common return shared with any track feed common return. The voltage regulator should have an electrolytic capacitor connected to the positive input pin and its zero volts pin and a smaller ceramic capacitor across the output and zero volts pins. The electrolytic is often 100uF and offers smoothing to the DC to help remove any AC ripple on the DC. You mention tantalum caps so that possibly is correct. Note the basic 7812 should have a heatsink fitted. They also need a minimum of approx 2 volts more on their input to give the required output. Link to post Share on other sites More sharing options...
RMweb Premium kevinlms Posted February 13, 2017 RMweb Premium Share Posted February 13, 2017 To me it does sound like a Common Return issue. If you connect 2 different outputs fed from the same input, via a common return, you will always experience such problems. I suspect the problem only occurs when you reverse the train and create a partial short. Forwards is OK, but not backwards. Two solutions are available, use a different power source, or separate the common return from the relay return. Take a look at Brian's site & look at Simplified Common Return. Note the separate power supplies, they are there for a reason. http://www.brian-lambert.co.uk/Electrical.html#Wiring Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 Thank you gentlemen. I've checked the 12v regulator board against the basic 7812 circuit diagram and assuming that tantalum caps marked 334 and 104 are the correct values, it's good. I'd prefer not to take out one of the relay modules in order to draw a circuit diagram of it, but I will later if I have to. Hopefully all (or at least enough) will be revealed by this snap ... The red and black wires running to and then under the right-hand module are the DC feed from the 12v regulator module, which is wired straight to the "12vDC" output of my Gaugemaster 100M. Nothing else connects to that output. The 16vAC output is used only for the CDU. Track 1 and Track 2 from the controller are wired to the layout, which is common return. In the above picture, the DPDT switches are just section isolators. Yellow and orange are point motor (SEEP) solenoid feeds, white is solenoids commons back to the CDU except for that one white from the bottom left-hand SPDT switch centre tag to the switches common bus which should be a different colour ... If the 12v regulator circuit which started getting hot yesterday is still working when I wire it back in, I'll try to establish if the problem only occurs with only one loco on the layout when I reverse direction of travel. Meanwhile, if you guys need any more details, I'll be on the case PDQ. Link to post Share on other sites More sharing options...
Suzie Posted February 13, 2017 Share Posted February 13, 2017 I would check the input voltage to the regulator. You should be seeing around 20V, but the '334' capacitor sounds a bit small - there should be a bigger one somewhere as well. It will need a heatsink, but a lack of one should not be a cause of failure. It does sound like there is some connection somewhere between the output of the regulator and another voltage source. Link to post Share on other sites More sharing options...
RMweb Gold Tony_S Posted February 13, 2017 RMweb Gold Share Posted February 13, 2017 Is it this module? http://modellingelectronics.co.uk/products/Single-points-position-indicator.php Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 Is it this module? http://modellingelectronics.co.uk/products/Single-points-position-indicator.php Yep. Link to post Share on other sites More sharing options...
Tricky Dicky Posted February 13, 2017 Share Posted February 13, 2017 I think you you have over thought the situation. It is quite common for unregulated PSUs to present a higher voltage open circuit than their stated nominal voltage but as soon as you present a load that voltage drops down to the stated voltage and as the current draw increases that voltage can drop below the nominal voltage. As Brian above stated the L7812s need an overhead on input of at least 2V, but as soon as your circuit presents a load and the voltage drops towards the nominal voltage you lose that overhead so it does not surprise me they overheat you are trying to make them do something without providing the adequate input. I think you need to take the regulator circuits out, once your relays trigger they will present a load to the PSU, measure the voltage then it should have dropped to towards the nominal voltage. Even if the voltage is slightly higher I would be surprised if 12V rated relays could not stand a slightly higher voltage momentarily or even for a short period. If you are still concerned about the higher open circuit voltage put anything a 12v bulb for instance across the output of the PSU to present a load and thus bring the voltage towards the nominal. Richard Link to post Share on other sites More sharing options...
Suzie Posted February 13, 2017 Share Posted February 13, 2017 Yep. Since you are using that module you do not need the regulator. The 12VDC input is just used for the LEDs so does not need regulating. The relays are powered by the output from the CDU so should be 24V dual coil latching relays. It is the point supply that is probably frying the relay, so measure the output of your CDU - the negative output should go to terminal 3. Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 I think you you have over thought the situation. Maybe, Richard. It wouldn't be the first time. OK ... back story coming up. I initially bought a couple of the point position indicator modules, which packed up as soon as I put power on. The maker was very helpful, but the outcome was that the failure was attributed to my having connected the modules to the 12vDC output of my 100M. Frankly I wouldn't have thought that was an unreasonable thing for a chap to do in the absence of any warning by the maker, but never mind. I forked out for a pair of replacements, and to his credit the maker went to great lengths to ensure that I wired them up correctly. He also kindly chucked in the 12v regulator circuit which he assured me would solve the problem, which it certainly seemed to have done until yesterday. Anyhow ... I can't lay my hands on a 12v bulb with which to load the DC output, but I do have a small assortment of 0.25W metal film resistors somewhere. Would one of those would do instead? Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 Since you are using that module you do not need the regulator. The 12VDC input is just used for the LEDs so does not need regulating. The relays are powered by the output from the CDU so should be 24V dual coil latching relays. It is the point supply that is probably frying the relay, so measure the output of your CDU - the negative output should go to terminal 3. Sorry, but how do I measure the CDU output when all I have is a cheap 'n' cheerful Maplin multimeter? Relays are http://www.futureelectronics.com/en/technologies/electromechanical/relays/signal-relays/Pages/2605020-HFD2-012-M(257).aspx?IM=0 Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 Gosh ... I just checked once again the no-load output from "12vDC" on the Gaugemaster 100M and it's steady at 15.6v. I then re-connected the 12v regulator circuit which overheated yesterday, checked my wiring, crossed my fingers and switched the mains on briefly. Nothing untoward seemed to happen, and the point indication LEDs were lit correctly on the control panel. I therefore put the mains back on, this time keeping a fingertip on the 7812. In no more than 2 seconds it was far too hot to keep my finger on it, so the power was knocked off PDQ. What's interesting (or maybe isn't, I have no idea any more!) is that this was with the 100M speed control at zero, direction switch centred, and nothing whatsoever happening on either layout or control panel. As I'm unable to think of anything more sensible to do for the time being, I'll now run the railway without any "12vDC" power connected to it and try my best to find something not working right ... Link to post Share on other sites More sharing options...
RMweb Gold Tony_S Posted February 13, 2017 RMweb Gold Share Posted February 13, 2017 Just opinions so please wait for others to criticise/review them. I am surprised that your first attempt at using the relay boards caused a problem. The 12 volt relay coils are rated to about twice that (from the data sheet on the suppliers website.) If under any kind of load the 15volt DC drops to about 12volts it won't enable the 7812 to work properly. It will sense an under voltage at the output, try to compensate and will probably be "on" with no regulation. Probably getting somewhat warm. My advice would be not to use the DC output from your Gaugemaster. Purchase a 12v regulated supply from someone like Rapidonline. They are quite cheap. This will(should) eliminate any problems with commons or whatever may arise from your Gaugemaster if that is the reason. Sorry if I have repeated anyone else's advice. I can't think and type. Best wishes Tony Link to post Share on other sites More sharing options...
cpman46 Posted February 13, 2017 Share Posted February 13, 2017 Since you are using that module you do not need the regulator. The 12VDC input is just used for the LEDs so does not need regulating. The relays are powered by the output from the CDU so should be 24V dual coil latching relays. It is the point supply that is probably frying the relay, so measure the output of your CDU - the negative output should go to terminal 3. The 12VDC supply is as Suzie says only required for the LEDs and is switched by the relay contacts completely separate from the relay coils.The relays are 12V dual Coil high sensitivity latching types. According to their spec the coils have a resistance of 960 ohms and can tolerate a maximum DC voltage of 27.7V. When you switch your point the CDU pulse is applied to the relay as well latching the relay and causing the output to change illuminating one or other of the LEDs. The coil is only momentarily energised the same as the point motor coil. Operating the point switch again pulses the other relay coil unlatching the relay and returning the LED outputs to their previous state. The relay contacts switch both the positive and negative rails of the 12VDC supply to the appropriate LEDs. If everything is wired correctly the only current consumed from the 12VDC supply will be that of the lit LEDs and the LEDs on the module. Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 My advice would be not to use the DC output from your Gaugemaster. Purchase a 12v regulated supply from someone like Rapidonline. They are quite cheap. That's actually the way I'm thinking right now, Tony, unless we (by which I actually mean the helpful folks on here) do end up finding the fault. Either that or I hurl these modules out the window after hitting them violently with a large hammer and buy a couple of those new-fangled point motors which have a switch for a pair of LEDs as well as one for frog polarity ... Link to post Share on other sites More sharing options...
Suzie Posted February 13, 2017 Share Posted February 13, 2017 Just set your Maplin multimeter to the 200V DC range and connect the probes to the output of the CDU. If it is more than 27.7V you are likely to fry the relays if they are 12V ones discussed above regardless of the 12V DC supply to the LEDs. Looking at the figures you have provided I would expect the output of your CDU to be about 23V. These boards are very simple. What has gone wrong with the faulty boards? Presumably they are stuck one way or the other. The wiring shown on the website suggests the relays should be rated at least 18V to eliminate any risk of overloading the coils. Sensitive relays sometimes have helper magnets, and these require that the coil polarity is the right way round, and that you don't overload them because you can damage the magnet, or damage the very thin wire in the coils. Link to post Share on other sites More sharing options...
Tricky Dicky Posted February 13, 2017 Share Posted February 13, 2017 Spikey When you said the modules failed what exactly went wrong did the LEDs burn out or did the relay coils burn out. As Suzie has said the 12v supply to terminals 4&5 is merely to power the LEDs the internal switches of the relay are rated well above what is drawn by the LEDs. If it is the coils that are burning out then according to the circuit your CDU output is too high. According to the relay data sheet the coils can take a maximum of 27.5v and that should be momentary, the module manufacturer recommends 20V max. Richard Link to post Share on other sites More sharing options...
Tricky Dicky Posted February 13, 2017 Share Posted February 13, 2017 When the modules failed did the point motors still operate? Richard Link to post Share on other sites More sharing options...
John ks Posted February 13, 2017 Share Posted February 13, 2017 Have you tried measuring the current coming out of the regulator. if it is low (<100mAmp)then I would suspect the regulator if it is high(>1.5Amp) then something you are feeding may have a problem of you are feeding to many devices Have you tried running the regulator without any thing connected to it if it gets hot then I would suspect the regulator if it doesn't get hot then it is something that the regulator is feeding reconnect all the devices connected to the regulator, 1 at a time an ammeter connected to the output of the regulator will help When you connect a device to the regulator the current will increase (<100mA if the device is ok, >1A if the device is faulty assuming that there are only a couple of LEDs or relays per device) or it gets hot if there is a problem( Are the 2 relay modules the only things connected to the voltage regulator if not then what else is there John Link to post Share on other sites More sharing options...
spikey Posted February 13, 2017 Author Share Posted February 13, 2017 When you said the modules failed what exactly went wrong did the LEDs burn out or did the relay coils burn out. I can't remember with any certainty, but I still have them here and might be able to test them somehow? When the modules failed did the point motors still operate? Yes. ... Are the 2 relay modules the only things connected to the voltage regulator ... Yes Off now to apply the multimeter and report back ... ETA - voltage at CDU output measures 23.6v Link to post Share on other sites More sharing options...
Tricky Dicky Posted February 13, 2017 Share Posted February 13, 2017 Hi Spikey Your CDU output is slightly higher than the manufacturers recommendation but less than the 27.7V max. for the relay so there should be no problem there. On the failed indicator modules the state of the coils can be checked by a continuity or better a resistivity test across terminals 1 & 3 and 2 & 3. An open circuit or resistance less than 960R will indicate a fault. As said before if you were using the regulator circuit connected to terminals 4 & 5 then it is not needed if the LEDs are blowing increase the series resistors. Sorry I cannot be any further help at the moment. Richard Link to post Share on other sites More sharing options...
spikey Posted February 14, 2017 Author Share Posted February 14, 2017 I need to get a grip of this and crack on with other aspects of the layout, so I've ordered a 12v regulated power supply in the hope that removing the 100M from the equation solves the problem. If it does, I still won't understand what the problem was, but I can live with that. If it doesn't, these modules go in the bin and I fork out for a couple of point motors which have two sets of changeover contacts, one of which I use with the new 12v supply for my LED indication. Sincere thanks to everybody who's helped out on this one. I've learned a lot from you guys. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.