Silicon Controlled Rectifier

[teh_pwnerer795]

Hey i am currently building a circuit and several websites have suggested to use an SLR. My question to you guys is, What is a Silicon Controlled Rectifier? What is it's sole purpose within a circuit?

Thnxs

[Binome]

An SCR is kind of similar to a transistor, yet significantly different. An SCR acts like an electronic switch, that once a current is applied to the gate, current can flow through anode and cathode until the current is removed from the entire device, compared to a transistor where the current has to only be removed from the gate.

[EricVanWyk]

What are you making that uses an SCR? Something high(ish) voltage?

Think of an SCR as a transistor that feeds itself. This allows a tiny control signal to activate it, but it means that you can't turn it off at the gate - something else needs to cut the current off.

This means that it has a lot of transistor like properties. Most importantly, its dissipation mode is a voltage drop - this makes it good for high voltage applications.

Please let us know what you are up to, it sounds interesting.

[teh_pwnerer795]

This year our robotics team could not afford to take part in FIRST Robotics. So we divide our team into two. Our group is building a EM Gun. Its going to be a 3 stage coil gun using #23awg copper wire. Our barrel is a aluminum hallow shaft measured at 3/8 of an inch. Im still trying to work out the numbers and schematics but everything takes time and research:) I am going to be using voltages around 350-400 and SLR's are apparently really good at controling and switching these high voltages and ampere's.

[Al Skierkiewicz]

An SCR is a silicon controled rectifier. This particular type of device can act as a switch and can be turned on with a very low control voltage but as it's name implies (and it's schematic symbol shows) it can only control current in one direction. It's other failure is once turned on, it cannot be turned off without removing the power that it turned on. This charachteristic while it sounds bad works very well in controlling pulsed DC power or in acting as a phase controled switch on AC power. It actually makes an interesting power supply when used in a bridge configuration. It's AC version is a triac and those are sued in virtually every light dimmer in use today.

[EricVanWyk]

I thought it might be something like that.

A word of warning - be careful when paralleling these devices. Add a small resistor to the gate of each SCR. This will prevent the gates from "ringing". Gate ringing destroyed my friend's coil gun that used SCRs. To over simplify, gate ringing causes the SCRs to turn on and off very very rapidly.

Also, I'm not sure if you need to add a ballast resistance to spread the current over multiple devices. Can someone verify this?

[lukevanoort]

While SCRs can work for you application, I'd only use them in a single coil coilgun design. MOSFETs or IGBTs are probably a better choice since you can turn them off. Also, don't forget to put a diode across each coil so that you don't fry your electronics from the current generated when the field collapses.

[teh_pwnerer795]

What type of resistor would i use>? 10 ohms? or like 10 million ohms?

[Kevin Sevcik]

This year our robotics team could not afford to take part in FIRST Robotics. So we divide our team into two. Our group is building a EM Gun. Its going to be a 3 stage coil gun using #23awg copper wire. Our barrel is a aluminum hallow shaft measured at 3/8 of an inch. Im still trying to work out the numbers and schematics but everything takes time and research:) I am going to be using voltages around 350-400 and SLR's are apparently really good at controling and switching these high voltages and ampere's. As far as your barrel material, you should remember that metals conduct and thus will generate eddy current opposing the magnetic field of your coil and robbing you of some power. There a few ways around this, but I'll leave that for you guys to discover.
While SCRs can work for you application, I'd only use them in a single coil coilgun design. MOSFETs or IGBTs are probably a better choice since you can turn them off. Also, don't forget to put a diode across each coil so that you don't fry your electronics from the current generated when the field collapses.You could use SCRs if you powered each coil separately, which I think you'd be doing anyways. I think starting with SCRs would be a safer bet than IGBTs since they'll stay on until they've dissipated all the energy in the caps. This would reduce your rate of fire since you'd have to recharge from zero every shot, but you wouldn't have to worry about a charged cap wanting to kill you after a firing.

Also, it's highly important to understand the purpose and placement of the extra resistor and diode you'll be adding. The coil of the gun is basically an inductor, and will be trying to keep a constant current going through it. Once your cap is spent, the constant current from the coil will act to charge it in the negative direction, which is baaad for the polarized caps you'll be using. So the idea is to re-route this current through a diode and resistor to dissapate the energy stored in the coil. The circuit would look something like:
SCR
+ --------|>-------
| |
Resistor Z 3
Cap | 3 Coil
Diode ^ 3
| |
- -----------------
The size of the resistor depends on how quickly you want to dissapate the energy, and how large a voltage differential your circuit can handle. The maximum current going through it is determined entirely by your coil's inductance and cap's size. After that, Ohm's law will tell you what the voltage drop across it will be and how many watts it will dissipate at this peak current.

[Scott L.]

This site may help with your project.
http://www.powerlabs.org/index.html

[teh_pwnerer795]

Awesome.. im gonna wrap my mind around this for a couple of days. Im not that logical so it will take me a while. I will be posting my circuits within a fews days. It would be wonderful if any of you could lend suggestions.

thnxs:)

keep posting if you theres an easier way to explain either the SCR or bleeder resistor. :)

[teh_pwnerer795]

Hey! actaully... i think i somewhat understand this stuff guys :D :D below are two photos that i scanned just recently.. let me noe if this makes any sense before i blow myself up :yikes:

NOTE* You will have to save the pictures and rotate/resize them. Srry any inconvenience.

Charging Circuit:
http://server6.theimagehosting.com/image.php?img=Charging%20Circuit2.JPG Circuit2.JPG

Fire Mech:
http://server6.theimagehosting.com/image.php?img=Firing%20Mech.JPG&public_view=1 Mech.JPG

C1 and C2 are basically linkers between the two separate circuits.

My power supply, Transformers, and capacitors are a big unknown question mark...?????? so again ... as mentioned before im still working out the numbers :)



NOTE* ifthe images are really unreadable... then let me know i'll rewrite and scan them with a better dpi

[Kevin Sevcik]

1. The cap goes between C1 and C2. When you're firing, you want high current provided by a high voltage through a low resistance. Your HV power supply couldn't supply the current on its own, so you charge up a cap and use it as a source of high current (for a very short time). Ideally, you disconnect the HV power supply from the cap just before you fire so the power supply doesn't melt trying to supply that kind of current.

2. You can trigger the SCR directly from the anode. That is, if you had a switch (with a resistor in series) connecting the anode to the gate, and you closed the switch, your coil would fire. Of course if you're serious about a multi-stage coil gun, you'd be triggering the SCR with yet other solid state devices like some transistors, etc.

[EricVanWyk]

Can you please post the part numbers you are proposing? I couldn't read them very well in the jpg.

I'm not sure what all the diodes next to the coil are for. You just need a single diode in anti parallel to the coil.

If you are using a single SCR, you can ignore what I said about ballast resistance. Just the gate resistance is necessary.

[Al Skierkiewicz]

Mitch,
The diodes in series with the coil are redundant and everything looks to be in the wrong polarity to allow current flow. The capacitor that is wired in series with the coil will form an LC circuit which will modify the resultant current in the coil. Also be advised, that fire switch and battery will be elevated to the supply voltage if and when the SCR fires.

[teh_pwnerer795]

2. You can trigger the SCR directly from the anode. That is, if you had a switch (with a resistor in series) connecting the anode to the gate, and you closed the switch, your coil would fire.

Im planning to have a spike set off the gate. The multistage will be set off by the program.

[Kevin Sevcik]

Im planning to have a spike set off the gate. The multistage will be set off by the program.
Note that they have a definite response time for the mechanical switching, plus I don't know that it would be easy to direct hook it up to the gate. Honestly, you'd probably have a lot better results if you used a solid state device like a few high voltage transistors connected directly to digital outputs on the RC. The response time would be loads better and, hey, you'd get to learn about transistors too!

Also, if you care at all about precise timing, you get to learn about working in the ProcessLocalIO() loop and working with interrupts and timers.

[teh_pwnerer795]

Cool. So i've made a small list of what i have to change.

1. reverse diodes
2. move capacitor
3. remove the bleeder diode and diode just after the anode of the coil.

We are using these pneumatics sensors that detect a metal object. I will be placing these over the aluminum to detect the projectile passing through. The program will then sorta calculate when to shot it off or if the projectile even moved. -.-'

The digital outputs and transistor idea sounds really good and i'll be looking into it. As well with the ProcessLocalIO() loop and interrupts :)



The only part that i have a model # for is the SCR.. The model # is (M25/16-1k). This part can be found on the following website. *at the very bottom. http://www.surplussales.com/Semiconductors/Diodes-1.html

[Scott L.]

Not sure what the timing is on the robot controller, but you will need sensors and actuators with millisecond and microsecond response time and program proccessing. all solid state components(no mechanical relays). Also I'd recommend using optical isolators between any controls and components that may inadvertantly have HV. Another thing to consider is that HV will arc through are if both potentials are near each other, so use wires rated for HV/HC, if current is high enough you also may risk crushing the barrel of the coil gun, and keep computer equipment at a safe range as a EM Pulse may be generated at the time of firing. also when working with high voltage, have someone else around that can get medical help if you inadvertantly get a severe shock.

[teh_pwnerer795]

ahaha ye so if my radio stops working i'll know whY:O lmao ye im sort of worrying about that now. The EMP thats gonna be created. I'll just try and move the components are far away as possible.

[EricVanWyk]

I've attached a picture of what I think the schematic ought to look like. I didn't include a transformer because I don't have a symbol for it handy. ;)

The SCR will trigger when the gate to cathode voltage (the two wires sticking up) exceeds the threshold.

For the love of cheeseburgers with mustard, please don't plug anything into this that you hold dear to your heart. At least not on your first attempt.

Check the capacitor you are using to see how much reverse voltage it can handle. It might be a good idea to add a second snubber diode hooked right across the cap's leads. This is not necessary in a perfect world, but parasitics can really get you in this high di/dt environment.

I might have read your link wrong, but I think that SCR is only rated for 16A.

Tape your experiments!

[teh_pwnerer795]

Ye i just noticed that the SCR is only 16 amps :( .. i was looking at the package surge rating. which is 400 amps.

[Al Skierkiewicz]

Mitch,
I am just getting around to rereading some of this post and your other question about bleeder reisistors. Bleeders are most often used to "bleed" off the charge on a high voltage cap when the power is removed. High quality and computer grade caps can store lethal voltages for quite a while. The bleeder then slowly discharges the cap to prevent these voltages from causing other problems in your design or hurting someone. The value is chosen so that the cap is discharged in a reasonable length of time but does not provide too great a load on the power supply. My old ham power supply, HP23B, uses 2 100k 2 watt resistors across a 125 microfarad cap in the 850 volt supply. In that configuration the resistors also help stabilize the power supply by providing a permanent load. Since the supply is used in the transmitter, load current doesn't reeally appear until you transmit without the resistors. (BTW the 850V supply is a voltage doubler)
In one of your schematics, there is a resistor in series with the cap. That resistor limits the "inrush" current to the cap. A capacitor that is discharged appears as if it is shorted when voltage is first applied. By using a series resistor you limit the current to a point that will not blow the primary fuse but still charge the cap. The digital TV transmitter I work on uses a bank of such resistors on a three phase high voltage power supply. There is a timed delay relay that shorts out the resistors after the caps in supply are partially charged. This transmitter uses a 35,000 volt power supply at 2 amps and the two main filter caps (in parallel) are 15" x 6" x 30" each and live inside an explosion proof enclosure with the transformer and rectifiers.

[EricVanWyk]

Ye i just noticed that the SCR is only 16 amps :( .. i was looking at the package surge rating. which is 400 amps.

Oh, you should be fine then. My bad...

If you are up to it, it would be a good exercise to guesstimate and then calculate what the surge current will be.

[teh_pwnerer795]

so if i add another SCR that is parallel with the capacitor and have a high resistor in series with the SCR. When the gate is activated, i can discharge the Cap saftey and slowly without sparks flying? this is similar to the idea of a bleeder resistor

[Al Skierkiewicz]

Mitch,
I don't know if who you are asking the question of here. An SCR across the cap is not needed if just the bleeder is used. If you plan on using an SCR to connect the bleeder, then you never really know if the SCR has fired and the charge has been depleted. Don't forget that a short across a charged capacitor will draw immense currents for a short period of time. This could have devastating effects on both the SCR and the cap when done repeatedly. The bleeder draws much lower current over a longer period of time. Of course you don't know if the bleeder resistor has gone open circuit without your knowledge and that is why you always want to make sure your caps have discharged before working on the equipment. Check twice, then check again before going to work inside the covers.

[teh_pwnerer795]

K i see where your coming from about the SCR being fired or not. I just wanted a safe discharge if the robot was about to be turned off. I was thinking of using a bleeding resistor to do that ... or just a HV Load, maybe a halogen bulb light bulb.

[Al Skierkiewicz]

maybe a halogen bulb light bulb.

That would produce a significant load on the power supply.

[teh_pwnerer795]

That would produce a significant load on the power supply.

im sort of confused when you mean power supply. If the power supply is off. ie. not charging. and the capacitor is somewhat charged... how would putting a load across the cap effect the power supply?

.... or.. i must of forgot to say that i would put a SCR or transistor in series to set off the HV load.

[Al Skierkiewicz]

Mitch,
My fault, as often when someone talks about putting a lamp across a source, it is implied to be permantly wired without a switch. BTW discharging a high voltage cap into a lamp shocks the filament due to the pulse like current during discharge. A lamp is a very low impedance when cold.

[teh_pwnerer795]

k thanxs for your help:) i really appreciate it.

another quick question. Me and another student are having a heated argument over how long the barrel of the tube should be. Saying we have a four foot tube, my suggestion was to cut it in half. This way the three coils are kept close together and has a stronger punch at the beginning.

On the other hand he suggests we use the entire four feet of tubing to gain a greater momentum.

my question is. For this type of experiment, would the projectile travel faster in a short tube with the coils close together, or longer tube spread out?


by saying this, keep in mind that our robot is only 2 feet wide, 1 (1/2) feet long.

[EricVanWyk]

k thanxs for your help:) i really appreciate it.

another quick question. Me and another student are having a heated argument over how long the barrel of the tube should be. Saying we have a four foot tube, my suggestion was to cut it in half. This way the three coils are kept close together and has a stronger punch at the beginning.

On the other hand he suggests we use the entire four feet of tubing to gain a greater momentum.

my question is. For this type of experiment, would the projectile travel faster in a short tube with the coils close together, or longer tube spread out?


by saying this, keep in mind that our robot is only 2 feet wide, 1 (1/2) feet long.

Bah, I just TAed a class that made coil guns. You'd think I'd know the answer off the top of my head... Make Greg Marra answer this. He was actually taking the class, I was just sorta teaching it.

My bet is that single coil would be a heck of a lot easier to do timing on.

DO BOTH!

Enjoy, and Good Luck!

[lukevanoort]

k thanxs for your help:) i really appreciate it.

another quick question. Me and another student are having a heated argument over how long the barrel of the tube should be. Saying we have a four foot tube, my suggestion was to cut it in half. This way the three coils are kept close together and has a stronger punch at the beginning.

On the other hand he suggests we use the entire four feet of tubing to gain a greater momentum.

my question is. For this type of experiment, would the projectile travel faster in a short tube with the coils close together, or longer tube spread out?


by saying this, keep in mind that our robot is only 2 feet wide, 1 (1/2) feet long.
Short tube, definitely. Your friend is probably thinking about explosively powered projectile weapons. In those cases, the additional barrel length allows the projectile to gain kinetic energy from the expanding gas for a longer period of time. In a coil gun, all that extra barrel length does is slow the projectile down due to friction from contact with the barrel. A longer barrel may help with accuracy somewhat, but you'd likely get similar improvement from rifling the barrel and using aerodynamic projectiles. One thing is for sure though, in a coil gun more barrel length will not increase the projectile's momentum. BTW, at school you may want to call this a 'Solenoid-based Electromagnetic Linear Accelerator' instead of a 'coilgun'... I don't know how uptight the administration is about weaponry in your school, but it might avoid some potential problems.

[teh_pwnerer795]

Ye i thought so as well with the short tube. shorter tube = less friciton.


Ye and our school is pretty strict. We were gonna call it MAC (Magnetic Accelerator Cannon).

Alrighty. We thanks alot everyone for your support. I'll be making a few more posts towards the end of november. I'll tape some of our experiments and if your interested, PM me for a couple of videos :)

Im starting up a ProcessLocal() thread in programming within a week once i learn about interrupts. Feel free to post :)

[teh_pwnerer795]

Hey guys im back with more questions :D

first question. is it possible to hook up two 6A diodes in parallel to get a total diode protection of 12A?

second question. if a bridge rectifer is rated for 25A, does that mean i have to put 25A through the rectifer for it actaully work?>?

Thanxs again..

if the question seems confusing, let me noe.

[Richard Wallace]

Hey guys im back with more questions :D

first question. is it possible to hook up two 6A diodes in parallel to get a total diode protection of 12A?

second question. if a bridge rectifer is rated for 25A, does that mean i have to put 25A through the rectifer for it actaully work?>?

Thanxs again..

if the question seems confusing, let me noe.1st: two 6A diodes in parallel will not share current perfectly, so their total capacity will be less than 12A. Also, the Ampere rating of a diode is generally tied to an assumed operating temperature; if actual temperature is higher then the diode will fail at less than its nominally rated Amperes.

2nd: the 25A rating on your rectifier is a maximum (again, at the given temperature, which is usually 25 Celsius) -- you can operate the bridge at lower current.

[teh_pwnerer795]

Hello everyone.

another question for my little project.

http://www.canadiantire.ca/browse/product_detail.jsp?PRODUCT%3C%3Eprd_id=84552444327 8837&FOLDER%3C%3Efolder_id=1408474396672500&bmUID=1196023856474&deptid=1408474396672395&ctgrid=1408474396672406&subctgrid=1408474396672500

this inverter supplys 120 VAC @ 300 watts (2.5A)

my question is, since there are two outputs, would it be possible to hook these in parallel to acheive a 120VAC @ 600 watts (5A)?

[Al Skierkiewicz]

this inverter supplys 120 VAC @ 300 watts (2.5A)

my question is, since there are two outputs, would it be possible to hook these in parallel to acheive a 120VAC @ 600 watts (5A)?

No, there are two outputs are in parallel for convenience. One must be careful when looking at AC inverters. Often they do not put out pure sine waves. They may be modified sine wave or they are sometimes quasi square waves. Use caution when feeding transformer inputs like power supplies as the unusual input waveform may cause core heating in the transformer.