Maker Pro
Maker Pro

diode for collapsing magnetic field

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
You lost me. The coil diode may delay the opening of the contacts a tad but we're taking milliseconds here.

Chris

I think that the issue is that instead of the current in the relay coil falling to zero almost instantaneously and the relay snapping open, the current decays at a rate determined by the inductance of the relay coil and its resistance (and the voltage drop across the diode).

The use of a zener diode increases the losses so the decay is faster.

It strikes me that a resistor in series with the diode would do a similar thing. You would choose a resistor such that the voltage drop at the operating current produces a voltage that the switching device can withstand.

It would be interesting to compare the effects in real life...
 

CocaCola

Apr 7, 2012
3,635
Joined
Apr 7, 2012
Messages
3,635

The diode for the coil should be reworked as such...

attachment.php


Not

attachment.php
 

Attachments

  • cor.jpg
    cor.jpg
    8.7 KB · Views: 216
  • cor2.jpg
    cor2.jpg
    9.8 KB · Views: 202
Last edited:

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I think that the issue is that instead of the current in the relay coil falling to zero almost instantaneously and the relay snapping open, the current decays at a rate determined by the inductance of the relay coil and its resistance (and the voltage drop across the diode). The use of a zener diode increases the losses so the decay is faster.

It strikes me that a resistor in series with the diode would do a similar thing. You would choose a resistor such that the voltage drop at the operating current produces a voltage that the switching device can withstand. It would be interesting to compare the effects in real life...
Yes, a resistor would do a similar thing. The PDF I linked to (which is written by a relay manufacturer) recommends the diode and zener combination as the best option.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
To the OP. You're getting confused. Here is a quick summary.

Your circuit has two inductive components: (1) the load (the pump), and (2) the coil of the big relay.

Inductance produces a voltage spike called back EMF when the inductance was being fed with a current, then something tries to break that circuit. In the case of the pump, this occurs when the big relay contacts open.

This back EMF voltage spike is of the opposite polarity to the voltage that was applied across the inductance. It can cause arcing in the contacts that break the circuit.

In the case of the load, the back EMF can be suppressed by using a large diode, rated for at least as much current as the load normally draws, such as the 1N1190 suggested by CDRIVE, which is connected in reverse across the load. When the supply to the load is cut off and the load tries to generate a back EMF voltage spike, this is clamped by the diode, and this prevents arcing and damage in the contacts that are interrupting the current (i.e., the contacts of the big relay).

Steve and others have suggested adding this diode since way back in the thread. This is the big diode that is rated for 30A or more. If possible, connect it directly across the terminals of the pump.

In the case of the coil of the big relay, which is also inductive, it's common to use a reverse-connected diode across it in the same way, and a 1N4001 would be suitable here. However, as explained in the PDF I linked to, doing this will cause the big relay to open much more slowly than it would otherwise, so its contacts will move apart more slowly than they would without this diode. This could cause arcing between the contacts as they separate, because of the heavy current the contacts are switching, although suppressing the back EMF from the load should greatly reduce or eliminate this arcing. If you want the big relay contacts to open as quickly as possible, you can use a zener and a diode in series, as described in the PDF, across the coil of the big relay. The back EMF will be limited to some extent but not clamped completely to zero, and the magnetic field in the big relay will collapse relatively quickly so the relay contacts can open quickly and cleanly.

I suggest you try the big diode across the pump first, with the simple 1N4001 across the big relay coil as shown by CocaCola. If the big relays still fail, look into improving the coil circuit by adding a zener diode in series with the 1N4001.

In answer to your other questions.

If back EMF in an inductive component is not suppressed, it will produce a voltage spike across the contacts that are in series with the inductance, which open LAST, and therefore break the circuit. Nothing happens while the current is flowing steadily; it's only when a contact opens and interrupts the circuit that the inductance generates the back EMF. Since the contacts that open last do not open instantaneously, and the back EMF is quite a high voltage, it can cause arcing between the contacts as they open. This is why it may be important for the big relay's contacts to open quickly, which is why I suggest the zener diode in series with the 1N4001 across the coil of the big relay. The need for the contacts to open quickly is related to arcing at the instant of disconnection, not to any functional need for quick switching for the operation of the control circuit.
 
Last edited:

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
That's my response as well...

That's actually wrong as well, it should be grounded on the lower side (as pictured) not the positive switch leg... Diode should be tied parallel to the coil...

Oops, I didn't notice that error but I don't think that Blank Stare intended to draw it like that either. That's a schematic typo! :p

screwball, I saw what I expected to see and those ovals on his schematic just served to obscure or distract my eyes. Yes, wired like that creates a latch.
Chris
 

screwball

Jan 9, 2012
89
Joined
Jan 9, 2012
Messages
89
That's my response as well...

That's actually wrong as well, it should be grounded on the lower side (as pictured) not the positive switch leg... Diode should be tied parallel to the coil...

there we are, this is what I was getting at, I was almost certain it would continue to stay open as we use it on 230v systems but with a contactor

the anode was on a would have been on a pretty much permanently live track rather than just across the coil
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
there we are, this is what I was getting at, I was almost certain it would continue to stay open as we use it on 230v systems but with a contactor

the anode was on a would have been on a pretty much permanently live track rather than just across the coil

screwball, sorry about that. If I seemed dismissive it's because I probably was. I had peeked at Blank Stare's modified schematic earlier but I didn't catch that error. :eek:

Chris
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
Thats a hardcore diode haha not seen one like it

Anyhow, just to repeat my previous statement incase it was missed, the diode across the relay coil would keep the relay closed even when the pressure switch is released right?

The path through the diode and through the coil to ground is easier than the path through the motor then to ground so it would almost definitely keep the relay closed when the pressure switch is released?

The cathode of that diode would go to the 30Amp fuse potential or stay where it is but a PTBS or something to break the flow to allow the relay to reopen and stay open until pressure switch is pressed again

You lost me. The coil diode may delay the opening of the contacts a tad but we're taking milliseconds here.

Chris



I think that the issue is that instead of the current in the relay coil falling to zero almost instantaneously and the relay snapping open, the current decays at a rate determined by the inductance of the relay coil and its resistance (and the voltage drop across the diode).

The use of a zener diode increases the losses so the decay is faster.

It strikes me that a resistor in series with the diode would do a similar thing. You would choose a resistor such that the voltage drop at the operating current produces a voltage that the switching device can withstand.

It would be interesting to compare the effects in real life...

I had to piece together pertinent posts to give context to my quote. As you can see It was in reference to screwball's latching statement because I missed the error on the schematic. I have no dispute with the manufacturer's data on improving drop out time.

While on the subject I would think that this would also be true:

If a inductor has an external parallel resistance equal in value to the inductor's resistance (Rs), it would limit BEMF to no greater than 50% of the original applied voltage.

Chris
 

Blank Stare

Oct 10, 2012
48
Joined
Oct 10, 2012
Messages
48
CocaCola is correct, I simply made a transcription error.

Anyway, I went outside, with the intention of measuring the voltage when the relay activates/deactivates.

I checked the relay, just to make sure it was all plugged in correctly. Taking each lead off, and then putting it back on again, I determined that, as I had previously checked, all of the contacts were good and tight.

Then I activated the relay, so I could measure the voltage...

And the pump came on...

This is the same relay that was in the system, when last it quit working...

So now I am wondering what the heck is going on? These relays are made to be connected with crimped terminators on the wires. I think maybe I will get some dielectric grease, and slobber some on every crimped terminator. I can't believe that there was a loose one, but maybe there was, or maybe it corroded a bit, and taking it off a second time cleaned it enough for it to work again. Either way, they get some grease.

That still does not explain the couple of times that the relays failed in the closed position, causing the pump to run continuously. (Mechanical pressure relief valve takes the danger out of a tank that gets too full.) So, I expect this relay will fail, and it's just a matter of time.

Which seems to leave me several options...

  • Leave well enough alone, and wait for it to fail again
  • Or, add the diodes for the purposes previously mentioned
  • Or, add the diodes, and the zener, just for good measure.
I think, since I have only ever had a problem with the relay, that I will wait, and see what happens. If history is any guide, I will be replacing the relay in a few weeks, or months. If so, I will go the full boat, and add both diodes, and zener diode.

; said:


I spent an hour looking at the mouser.com guided catalog, and all I could find
were those monster zeners, like the one that you linked to earlier, Kris.

Is there a smaller, barrel shaped zener that will get the job done, or is that my only choice?

Once I have spec'd the proper zener, I will consider this thread complete, for my purposes.
If I run into more trouble, after installing the diodes and zener, I will open a new thread.




IN A WORLD WHERE PEOPLE OFTEN WON'T EVEN GIVE YOU THE TIME OF DAY,
YOUR UNDYING PATIENCE IS A VALUABLE COMMODITY.

THANK YOU EVERY ONE !

~ Blank Stare

(Note: The attached schematic is a corrected version of the one I posted most recently.)
 

Attachments

  • AAA finaldraft expanded colorized.gif
    AAA finaldraft expanded colorized.gif
    11 KB · Views: 106
Last edited:

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
You're confusing yourself again.

1. The diode ACROSS THE PUMP must be rated for 30A OR MORE. It is NOT A ZENER DIODE. The 1N1190 looks suitable. The 1N4001 is NOT suitable for this position; it's only rated at 1A.

2. The diode(s) across the COIL of the main relay can be a lot smaller. The 1N4001 you have in your schematic is correct and will work. However, connecting a single diode like this across the coil will slow the release of the contacts. To avoid this problem, you can ADD A ZENER DIODE IN SERIES with the 1N4001 that's across the COIL. It will be connected in series, pointing the opposite way from the 1N4001, like this:

. . . . . . . coil
. . . -----oooooooo------
. . . | . . . . . . . . . . . . |
0V------|>|------------|<|------------ positive feed from gating diodes
. . . 1N4001 . . zener

For the zener I would suggest something around 100 volts with a good power dissipation rating, e.g. 5W. The 1N5378 fits that specification.
 

screwball

Jan 9, 2012
89
Joined
Jan 9, 2012
Messages
89
screwball, sorry about that. If I seemed dismissive it's because I probably was. I had peeked at Blank Stare's modified schematic earlier but I didn't catch that error. :eek:

Chris

No problem mate, i did not see at first, then i spotted it and it confused me but now all is well :cool: Thanks
 

Blank Stare

Oct 10, 2012
48
Joined
Oct 10, 2012
Messages
48
Looks like its getting there :p
Do you understand the purpose of the diodes across the motor and relay?

I think so. When the magnetic field collapses, the energy that was in the electro-magnet (coil), looks for someplace to go. Also, it increases the voltage to a very high number, while it arcs across the contacts. The diode gives the voltage somewhere to go, instead of back at the contacts, where it potentially erodes the contacts, much like an Arc-Welding machine does with consumable "sticks". It kicks that positive charge back at the positive source that feeds the relay, forming a dead-end. I'm not quite sure why that works, but I can take it on faith, rather than study the chemistry of semi-conductors.

-----diode---><--zener-----What I was confused about, was which component to use it on. I now understand to use it on the 5 amp side of the relay.

I may have mis-stated it, but I understand to put the zener in series with the diode, pointing the (negative?) sides at each other. (The whole anode/cathode thing has me confused... cathode = negative?... or did I get that backwards?) Anyway, that's easy enough to look up, whenever I am ready.

1. The diode ACROSS THE PUMP must be rated for 30A OR MORE. It is NOT A ZENER DIODE. The 1N1190 looks suitable. The 1N4001 is NOT suitable for this position; it's only rated at 1A.

Got it. In my head I understood that, but no doubt my fingers betrayed me - again. :eek:

2. The diode(s) across the COIL of the main relay can be a lot smaller. The 1N4001 you have in your schematic is correct and will work. However, connecting a single diode like this across the coil will slow the release of the contacts. To avoid this problem, you can ADD A ZENER DIODE IN SERIES with the 1N4001 that's across the COIL. It will be connected in series, pointing the opposite way from the 1N4001, like this:

Understood. When this relay fails, as it most certainly will. I will take the time to add the zener to the diode, on the relay on the coil. (Did I get that right this time? :D )

For the zener I would suggest something around 100 volts with a good power dissipation rating, e.g. 5W. The 1N5378 fits that specification.

Thank you, having the number to look for when I am ready to order is more help than you can imagine.

Is it just my imagimation, or is that zener large enough to plug my :ahem: porcelain throne?

Anyway, to sum up everything - I hope

  • 1n4001 diode, and 1N5378 zener reversed at each other across the coil of the relay.
  • 1N1190AR 40A, 600PIV Rectifier Diode, in parallel across the pump.
http://www.mpja.com/1N1190AR-40A-600PIV-Rectifier-Diode/productinfo/18393+DI/

Sorry Electrician, the warrant was denied on the grounds that the manpower to exhume it from from the city dump is unavailable. :eek:

Thank all...
 
Last edited:

Blank Stare

Oct 10, 2012
48
Joined
Oct 10, 2012
Messages
48
I didn't get an answer about "how to test for a bad diode?"

If power goes both ways?

If power can't go either way?

Do I need a meter to measure how much power gets through?

Am I totally off-base? :eek:

Not necessarily relevant to the current discussion, but I figured it couldn't hurt to ask, since someone mentioned a "blown diode".

Thanks
 

screwball

Jan 9, 2012
89
Joined
Jan 9, 2012
Messages
89
I didn't get an answer about "how to test for a bad diode?"

If power goes both ways?

If power can't go either way?

Do I need a meter to measure how much power gets through?

Am I totally off-base? :eek:

Not necessarily relevant to the current discussion, but I figured it couldn't hurt to ask, since someone mentioned a "blown diode".

Thanks

Diodes you can sometimes tell just by looking at it, but you can use a multimeter to test a diode, most of them have a setting purposely for testing diodes, search it

power can only flow one way through a diode, hence its purpose although at some point (cant remember its name) much higher than its forward rating it will obviously pass through, for example putting 230v 20amp the wrong way on a 1N4001 diode (1 Amp)
However it is part of the zener diodes function for a certain amount of current (or voltage cant remember) to flow both ways, search it

Your looking too much into it again, if your motor uses 30amps then you want a diode to take the flow or 30amps or more when power to the motor is dropped

Your not far off, however you need to look more at basic functions and principles of components again/more IMO
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
............power can only flow one way through a diode, hence its purpose although at some point (cant remember its name) much higher than its forward rating it will obviously pass through, for example putting 230v 20amp the wrong way on a 1N4001 diode (1 Amp).

the diodes PIV .... Peak Inverse Voltage
for example the 1N4001 has a PIV of 50V, a 1N4007 has a PIV of 1000V

Dave
 

Blank Stare

Oct 10, 2012
48
Joined
Oct 10, 2012
Messages
48
Diodes you can sometimes tell just by looking at it, but you can use a multimeter to test a diode, most of them have a setting purposely for testing diodes, search it

power can only flow one way through a diode, hence its purpose although at some point (cant remember its name) much higher than its forward rating it will obviously pass through, for example putting 230v 20amp the wrong way on a 1N4001 diode (1 Amp)
However it is part of the zener diodes function for a certain amount of current (or voltage cant remember) to flow both ways, search it

Your looking too much into it again, if your motor uses 30amps then you want a diode to take the flow or 30amps or more when power to the motor is dropped

Your not far off, however you need to look more at basic functions and principles of components again/more IMO

Thanks, that was the answer I needed. It was more a matter of curiosity, than need. However, I will do some google searches, just to learn a bit more.

I went to mouser, and searched for the parts discussed here. They all seem to come with letters and/or numbers on the end, and sometimes on the front, but contained in the middle were the "names" - so to speak.

I am assuming that the extra information is there for the purpose of cataloging, and warehousing the parts, rather than being a different spec?

Looking at the data sheet, leaves me thinking so.

Here's an order I put together, then screen captured, and cropped. Whole things under 13 bucks. (I bought a few extra 1n4001 diodes, and the zeners, in case I fry some, when I do the soldering.

They cost next to nothing, but shipping to get more would be 25 times the cost of getting another one. Rather toss three in the trash, then have to order another...

Thanks
 

Attachments

  • components 20121020.jpg
    components 20121020.jpg
    25.4 KB · Views: 162

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I think so. When the magnetic field collapses, the energy that was in the electro-magnet (coil), looks for someplace to go. Also, it increases the voltage to a very high number, while it arcs across the contacts. The diode gives the voltage somewhere to go, instead of back at the contacts, where it potentially erodes the contacts, much like an Arc-Welding machine does with consumable "sticks". It kicks that positive charge back at the positive source that feeds the relay, forming a dead-end. I'm not quite sure why that works, but I can take it on faith, rather than study the chemistry of semi-conductors.
That's a pretty good description.

The characteristic of inductance is that current through the inductor cannot change quickly. While the inductance is energised by current flowing from the power supply (via some contacts), current flows from the power source, in the top end, out the bottom end, and to ground. (This is "conventional" current, which flows from positive to negative, rather than the actual electrons, which flow the opposite way.)

When the circuit is interrupted, the inductor "tries" to keep the current flowing at the same magnitude and the same direction. This means that the grounded end now becomes positive and the top end swings negative, the opposite polarity from the voltage that was applied to it. The inductor can generate a large negative voltage in its attempt to keep the current flowing. This is the voltage that can arc across the contacts.

If you connect a diode in reverse across the inductor, it provides a path for the current to take, so the inductor doesn't need to generate a high voltage in order to try to keep the current flowing through the inductance. This obviously protects the contacts that are in series with the inductance from being damaged by high voltage arcing.

The diode provides a direct path to keep the current flowing in the inductor, and the current (and the magnetic field) collapse relatively slowly because little energy is lost in the diode (it has a forward voltage of about 1V). If the voltage wasn't clamped so hard, they will collapse more quickly. This is the reason for the zener; it clamps the voltage, but only to 100V or so, so it dissipates the energy from the inductor fairly quickly. In the case of the relay coil, this allows the relay to open quickly and cleanly.

I may have mis-stated it, but I understand to put the zener in series with the diode, pointing the (negative?) sides at each other. (The whole anode/cathode thing has me confused... cathode = negative?... or did I get that backwards?) Anyway, that's easy enough to look up, whenever I am ready.
The series combination of a diode and a zener doesn't conduct at all when the top end of the inductor is positive (i.e. while the inductor is energised by the current flow from the power supply), and when the top end of the inductor goes negative due to back EMF, the series combination conducts with a 100V drop, i.e. it clamps the voltage in that direction to 100V. (Or whatever voltage you choose.)

Understood. When this relay fails, as it most certainly will. I will take the time to add the zener to the diode, on the relay on the coil. (Did I get that right this time? :D )
Sure. Personally I would add it anyway, but that's up to you.

Is it just my imagimation, or is that zener large enough to plug my :ahem: porcelain throne?
The zener isn't particularly big. The 40A diode that goes across the pump is pretty meaty though. I wouldn't sit on it though; it has sharp edges :)

1n4001 diode, and 1N5378 zener reversed at each other across the coil of the relay
1N1190AR 40A, 600PIV Rectifier Diode, in parallel across the pump.
http://www.mpja.com/1N1190AR-40A-600PIV-Rectifier-Diode/productinfo/18393+DI/
Right.

I went to mouser, and searched for the parts discussed here. They all seem to come with letters and/or numbers on the end, and sometimes on the front, but contained in the middle were the "names" - so to speak.

I am assuming that the extra information is there for the purpose of cataloging, and warehousing the parts, rather than being a different spec?
It varies. With semiconductors in general, the prefix usually identifies the manufacturer, although 1Nxxxx diodes don't have a manufacturer prefix. Suffixes for semiconductors in general can indicate many things, including the package, voltage rating, gain range (for transistors), improved accuracy (op-amps), temperature grade, lead-free manufacture, packaging (tape and reel) for surface-mounted components, and other characteristics. In this case, the presence of an R in the suffix is important because it indicates "reverse" polarity - see CDRIVE's comments in an earlier post.
 
Top