Launch controller circuit animation - Eyes-on and opinions welcomed

[OzHybrid]

Wasn't about the circuits shown in the thread. It was about your comment concerning a trend in using FET devices which Grog6 followed up by saying it isn't a good idea. The reply from kc9qzf and I agree with is a launch controller using FETs that is properly designed for using FETs is practical.

An unconnected gate on a FET is ON. An unconnected gate on a Transistor is OFF. They get activated with a whiff of any stray voltages or radio. Just from the stray voltages from powerlines, a FET would be on if the gate was unconnected. Clearly, an unconnected anything is not a good circuit and no one would create that deliberately, but the failure mode of both is Dangerous for FET and Safe for transistors. (I'm overgeneralising here, but what I've stated is true in principle) As a plus, implementing current limiting with a transistor circuit is easier. However, using FET SHOULD result in lower power dissipation. Like most things in life, there's a tradeoff has to be made.
And to say it one more time, there is not a properly designed circuit shown anywhere in this thread.
Looking forward to when @kc9qzf gets back to civilisation and posts his version.
My 2c. YMMV.

 

[sghioto]

An unconnected gate on a FET is ON

This is true but a properly designed circuit using FETs in a launcher would never have the gate floating. It's always at ground potential when not activated.
Current limiting through a FET is easily done using a current regulator IC.
Personally I wouldn't use FETs or more precisely mosfets to activate igniters, I use 20 amp auto relays but do use a SSR as a control relay.

 

[OzHybrid]

This is true but a properly designed circuit using FETs in a launcher would never have the gate floating. It's always at ground potential when not activated.
Current limiting through a FET is easily done using a current regulator IC.
Personally I wouldn't use FETs or more precisely mosfets to activate igniters, I use 20 amp auto relays but do use a SSR as a control relay.

Current limiting through a transistor is easily done with a resistor. No IC or further support componentry required. But there is an argument for both. You just need to be aware of the full limitations, advantages and disadvantages of any component you use in a firing circuit.
Unfortunately, it's easy to get the tunnel vision of only the advantages of a particular methodology.
In a firing circuit, risk is what needs to be mitigated. Risk= the chance of something happening x the magnitude of the consequence.

With your particular choice, 20A automotive relays, you've made a reasonable choice. Good overload characteristics. However a relay internally has a set of contacts and if you knock the relay hard enough in the action direction of the contacts, they will make contact. Dropping a firing box would usually be enough. 2 relays in series with the contacts facing a 90 deg angle to each other would be better. I doubt you've had that problem, but I've personally seen it happen in testing. Now you know, you cannot unknow. You now have to make the choice to mitigate the risk or accept it. If you accept it and there's an incident, you cannot say you didn't know. Is the chance of that happening high? No. But the consequence magnitude is high.

The point is, there is no simple answer. There are consequences for whichever method you use. Just make sure you know what they are and mitigate them.

 

[sghioto]

However a relay internally has a set of contacts and if you knock the relay hard enough in the action direction of the contacts, they will make contact.

That's why I use a solid state relay as a control device. Pads are selected using relay contacts but no voltage is present until the SSR is activated and no contact arching. Below is a 5 pad design I built awhile back.

 

[OzHybrid]

The battery is at the head end. There is no specified at the head end isolation for it such as a switch or keyswitch to apply power to the circuit. The control for everything is through a single IC which if it fails is an issue. If the cable is connected at the head end first, the circuit could pick up enough stray signal for operation unless you're using shielded cable.
Your whole firing position control has no ACTUAL power isolation for the circuit and depends on the uln2003 at the head end. There is no current limiting resistor for the test circuit. You are dependent on the buzzer/beeper self current limiting for safe test current limiting.

 

[OverTheTop]

ULN2003 devices are darlington relay drives, and each output has an inbuilt catch diode.

 

[OzHybrid]

ULN2003 devices are darlington relay drives, and each output has an inbuilt catch diode.

OK list edited. Still a lot of issues.

 

[kc9qzf]

Ok, I am back. I am attaching a page of a schematic which uses a P-FET (Q3) as a high side driver. The control circuit below it uses a bjt (Q2) to control the gate. When the bjt is active, it pulls the current away from the gate which turns Q3 on.
As a side note, Q1 which is also a P-FET, is used as a reverse battery protection. Normally, the gate of Q1 is pulled to ground which means it will be turned on, which will allow a very low voltage drop as opposed to a normal diode. If however in a reverse battery condition, the gate will be forced to a potential as high as the source and the device will be shut off. In this case the body diode will aslo block any reverse current.
The bjt (Q2) is just an NPN device that turns on when current flows between the base and the emitter. This current is limited by R3. R5 is provided to assure no stray voltage condition from an open could turn on Q2. The dual diode is provided to make sure a reverse battery condition does not create a high reverse voltage against the base/emitter diode. The input [DSW] is a digital output from a microprocessor. As further protection against the dreaded reverse battery condition, a reverse biased diode at the output of the circuit will allow inductive devices a drain path and keep a reverse voltage to a minimum.
I just had this schematic handy and could explain further if requested.

 

[sghioto]

OK list edited. Still a lot of issues.

One thing I did forget to include on the schematic was a voltage divider on each input of the ULN2003 to decrease the sensitivity. I was redrawing the schematic from an older jpeg file, my apologies.
Remember a schematic shows how things are connected not how it's actually constructed. The Battery is not hardwired but connects using cables. The beeper is rated under 20ma at 12 volts. The control cable was standard Cat5. This system has been used for hundreds of launches with Cat5 cables up to 500ft with no problems.

 

[CameronMakesRockets]

/Wow, I wish I found this thread before I made my launch box!

Provide an AUDIBLE warning, either at the pad or on the controller. LEDs are ok, but a buzzer is better.

This is a good idea, helps everyone know if the controller is hot or cold.

Great resources here, I will keep this thread in mind when I build my next one!

 

[Curtis Enlow]

Here is a short video of the first test of the controller simulated at the IP of this thread.

This was tested with an e-match. (The Continuity LED doesn't go out because the e-match leads fuse together after igniting).
As I said before, this is only for Class 1 LPR/MPR launches at the local field, though the pad could probably do HPR, but for nowe the local L/MPR field is infested with soccer players so it will have to wait a bit for her initial launch.

Thanks to everyone for your invaluable comments & suggestions

 

[brockrwood]

Mosfets are NOT a good idea for the switch in a launch set. I'll tell you why. You need a mosfet, a voltmeter, and a dry table. the number on the side of the mosfet is it's type; you can search google, and see which pins are which. They are source, gate and drain. An N-channel mosfet needs the drain + , the source -, and the voltage on the gate turns on the channel between them. do, the Voltmeter on ohms, clip the black meter lead to the source terminal, and the red read to the drain. It should be open circuit.
But here's why it's a dangerous thing:Tale and touch the gate lead with a finger, it will usually turn on.touch the black lead with your other hand, then touch the gate, it will turn off, touching red will turn it on, thru the resistance of your dry body. The circuitry needs to be carefully designed; I've seen an open switch contact make a mosfet a random voltage detector, where it would switch when i reached for it, from the static on my hand.
something else on switch ratings. If you see 10a125v .1a dc or somesuch, it means the traveling contact is vulnerable to wear, and it won't hit it's million operation limit in the datasheet. That switch can handle 2x current for 20 secnds or so, look for overload ratings. DC is hard on switches, BUT not dc below 28 Volts. Only gold contacts wear at 12v. Coin silver contacts will last forever at the overload rating. But only below 28v.

How about a power Darlington as the switch to connect the battery to the igniter(s) in a launch controller? How about some other power BJT whose base I can feed enough current to that it saturates fully and sets off the igniter(s)? I am currently using a relay based home-built controller. I would love to replace the relay with something solid state. The relay contacts can handle the 12V lead acid battery connected to it, but the relay control input requires 12 volts to make it go. I would love to replace it with some sort of transistor that I can control with just a couple of volts.

 

[sghioto]

Many SSRs only require 3 volts to activate. There are several options for low voltage switching.
Use a npn transistor to activate the relay or a combination npn and P channel mosfet to make a high side SSR.

 

[OzHybrid]

Simple solenoid saver circuits. They do need to be matched to the solenoid you'll use. For most of us that will be a single solenoid that is unlikely to be changed and used for all hybrid launches. So it's a fixed value once you've worked out your values.

https://www.electronicdesign.com/te...0/whats-all-this-solenoid-driver-stuff-anyhow
Link above