Maker Pro
Maker Pro

Unstable latching circuit`

Rob_K

Sep 20, 2013
59
Joined
Sep 20, 2013
Messages
59
Ah, here, this is the one I am using, which values should I be paying attention to?

My guess is the threshold voltage Vt0?
 

Rob_K

Sep 20, 2013
59
Joined
Sep 20, 2013
Messages
59
Ok, this has now stopped working, what am I doing wrong, I just tried to up the resistor from gate to ground, but no joy. It must be possible to do this surely, I have seen latching switches everywhere, how to they do it?
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
i did not see any data in #21

I show a sketch of how the SCR should be connected. I also show how a relay can be latched without electronics.
 

Attachments

  • Rob-K.png
    Rob-K.png
    3.8 KB · Views: 132

Rob_K

Sep 20, 2013
59
Joined
Sep 20, 2013
Messages
59
Ok, basically, every time the thyristor is soldered to a strip board, with no other components but a resistor, it triggers without the trigger. I have tried a 150Ω resistor, 200Ω resistor, Ω1000Ω resistor, all trip, yet take it off and put it on a bread board, works fine????????

Can anyone actually confirm that they have successfully used a thyristor as a switch, because this just does not work. I really must be missing something. Is it possible that the copper on the strip board is triggering the thyristor.

Here is the thyristor that I am using http://pdf.datasheetcatalog.com/datasheet/stmicroelectronics/6162.pdf

I hope the link shows up this time.

Ok, this is the last thing I am doing with this, as I am starting to think that for one shot switches i.e. switch on and stay on electronics is not the way forward and the only way is with a relay. But one last bit of understanding before I throw this idea away.

The resistor used from trigger to ground, I understand this is like a pull down resistor is that right? Also, it does appear that the smaller the resistance the less sensitive the trigger, as I have it now working with a 100Ω resistor at the moment, no other components on the board, trigger not even attached. What is confusing me is that logic, mental logic that is, says that if the lower the value of the resistor, the less sensitive, then surely no resistor at all would be the ideal. Yet in reality, there has to be some resistance. So in that case there must be a cutoff point, like a low pass or a high pass filter, is this right? and if so, this data should be on the data sheet right? so in the data sheet, the only graphs I can see that demonstrate that are figs 6, 7, or 10.

Am I on the right track anywhere here? don't worry, I'm not going to pursue this for much longer without a human in front of me, if I can find one.

Thanks all for your input and patience.

Rob
 
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
As you say, the lower the value of the resistor, the less sensitive it will be. If you connect the gate to the cathode it should not work at all.
You have chosen a thyristor with a very sensitive gate.
You say you have checked the stripboard but have you measured the resistance from one strip to the next. I go cross eyed when working with strip board !

The only other thing is that a rapid rise in voltage across the SCR can turn it on. A capacitor across it may stop this.

I have used thyristors in electronic ignitions and electric fencers without problems.
 

Rob_K

Sep 20, 2013
59
Joined
Sep 20, 2013
Messages
59
Thanks, I am starting to understand this a little more now and actually, yes I did measure the resistance from one strip to the next, but I dismissed it as I thought my multimeter was faulty. Just checked again, this is interesting, The empty part of the strip board, where I have not soldered, averages out around the 40MΩ mark, but where I have the components, it goes right down to the 20kΩ mark, which is quite low when you consider a 10kΩ resistor was used in my transistors attempt on the trigger pins!!!

So the big question, is my soldering that bad (this really is a distinct possibility, see attached photo, or is this the nature of strip boards to act, I guess like a capacitor really isn't it, conductive material and nonconductive material sandwiched together.

If this is the case, I should be able to get away with matrix board and wire couldn't I, it's just that PCB is a little out of my depth right now, I can design them, I just don't know how to get them made up for less than £50.
 

Attachments

  • Soldering.jpg
    Soldering.jpg
    53.7 KB · Views: 122

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Hi Rob_K

If you want to control your circuit's sensitivity you might be better to go back to your original two-transistor circuit. I would suggest several changes to the circuit you posted in post #1 on this thread. Most of these were suggested by duke37 in post #6.

1. Change R3 from 100k to 33k (or 10k if you don't have any 33k resistors).
2. Move C1 so it's connected across R4.
3. Add a 10k resistor from the base to the emitter of the BC547.

I recommend connecting a diode across the relay coil in all cases. You could also connect a diode (e.g. 1N400x) acoss the PNP transistor, with its cathode to the PNP's emitter and its anode to the PNP's collector. This all helps to reduce spikes which can be a problem in an automotive environment.

If you continue to have problems with the stripboard version, post some photographs of your stripboard circuit.
 
Last edited:

Rob_K

Sep 20, 2013
59
Joined
Sep 20, 2013
Messages
59
Thank you, I will try this also.

For now, I am just going to leave my observations on this little project should anyone stumble upon this post with similar problems. For the record, I studied electronics as part of my degree which I finished just 2 months ago, but have found it a challenge in transferring it to the real world solutions, these are my findings based on a week of research and testing.

This project was to provide a way of temporarily securing the ignition of a (in this case very large and tatty) motorcycle that had been stolen and recovered with the ignition switch removed. Instead, an ordinary on off switch was to be used that would not work until a magnet was run across a reed switch hidden somewhere on the bike. This would activate a 30A automotive relay which would be the ignition switch. I always lock my bikes up, but this is to stop someone riding off with it when you go to pay at a petrol station, or nip into a shop for 5 minutes.

1. The system had to be safe, not catching fire, no loss of current while the bike was running throughout the various stages of vibration through the rev range.
2. The system requires that it fail in the off position. (Otherwise it defeats the purpose of having it)

Three circuits have been tried:
1 A relay, this is by far the most efficient, it just works for this kind of thing, trying to source out the right size is tricky as the only 12V relay with a large enough safety margin regarding Amps tend to be over 2cm in height (mine is 1 inch) that plus the board and a little space below when mounting it means a rather large project box for just 2 components (relay and diode). There are smaller ones about, but I have not found it easy to source as a noobie in the real world of electronics.
2. Two transistor circuit of which there is much excellent advice on this thread, though it is difficult to trust if like me, you are not familiar with physical circuits as opposed to theory.
3 A thyristor, this was by far the way I really wanted to go, it was small discrete, required minimal other components to work and in fact does just work... up to a point. Here the biggest problem by far is understanding the specification of these components as without experience datasheets can be a little jargonish to deal with.

At time of writing, I have a seemingly stable circuit that now works, thanks to the wonderfully helpful members of this forum, I used the top diagram that duke37 kindly provided, using an X0405 thyristor and after much trial and error, I reduced the 150Ω resistor which worked for a while and then was majorly inconsistent to just 50Ω which has now remained consistent for some 3 hours tested with the power switch "on" for an hour or so before triggering and then off for a while and coming back to it. There seems to be some sort of working threshold for this resistor at around the 150Ω 200Ω mark and above that the thyrister would just trigger without being triggered.

Now there are a couple of things I want to mention at this point and that is the fact that this resistor and the 560kΩ resistor on the trigger provide a voltage divider to the trigger itself. This is a part I have found a little confusing to work out. In an article I read here, it says that the maximum gate value must not be exceeded, and that appears to be from the datasheet 0.8V. I must have this wrong though, as once I finally got the thyristor not triggering immediately, it seemed to require about 2V to trigger. That said, it is currently triggering happily at 0.738V, but I am only using a 9V battery at present, I may have to change the 560Ω resistor as with a 12 V battery, it puts the gate value up to 9. something.

The other point is the confusion I have had with the triggering of electronic components. I had this with the transistors, hence the original post and I have had it with the thyristor. I am sure there is a way to do this and be confident about it, but the use of a strip board is not the way forward. After testing the resistance of the strips on the strip board, and I have come across this before, but I dismissed it because it turned out to be due to rf noise in the robotics lab at Uni and turned out that they don't like strip boards at university, because they can basically act like flat unpredictable capacitors throwing all sorts of messy values around. This will not help you debug an electronics circuit, it adds too many unknowns that are pretty much impossible for the beginner to know how to deal with (I was nearly completely defeated and put off electronics because of this). As it happens, this is still on a strip board, but as I am going to use a relay as and when I can find a small enough one to fit in the space I have to play with, this is just a work in progress. My next design for this will be using matrix board and wire or PCB if I can find a way to get them made at a reasonable price, throwing £50 at something you are not sure about is not how people become millionaires.

I hope this helps and I welcome any feedback re information that may be misleading or just wrong.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
A resistor to the gate will limit the current and hence the voltage. It will never get over a volt or so unless you are abusing it.

Some conductance between strips is not likely to be due to solder, that would either connect or not. The most likely cause is contamination with flux. Try cleaning with white spirit and then methylated spirits. Do not use the toothbrush again, it will taste horrible. Plumbers flux contains salts (zinc chloride?) which should never be brought anywhere near an electronic circuit.

If conductive flux is the problem, then going to a printed circuit will not help.
 
Top