PROP Firing Circuit - using HIGH SIDE SWITCHES - UPDATED

DavidM

HI,

My previous Firing Circuit did not seem to go down too well with people , So I have looked into using HIGH SIDE POWER SWITCHES instead.

I have attached a Schematic showing a basic set-up for 8 Outputs ( using the BTS452R device in this example) , I have attached the PDF's of the various types of Switches that I would like to use..

1) INFINION - BTS452R - "SMART HIGH SIDE POWER SWITCH" - ( DIGIKEY $1.78 USD per 100 pcs)
2) INTERNATIONAL RECTIFIER - IPS7091S - "INTELLIGENT POWER HIGH SIDE SWITCH" - ( DIGIKEY $2.20 USD per 100 pcs)

These devices have the following features which seem to be what I am after.

* Separate Digital ground to Power ground
* Open Load Detection ( for continuity Check, no need for extra 9-12V power)
* Current Limiting ( no need for extra fuses)
* CMOS Control ( 5V logic with I/O Expander chip)
* Short Circuit protect and detection.
* Only 4 resisters required (very easy to design with)
* and many other features..


The INFINEON - BTS452R is cheaper that the International rectifier IPS7091 version, so I will more than likely use the INFINEON Brand.
It also has High voltage and amps specs.


In my schematic, I am showing just a basic 8bit I/O chip, but this can be any similar device, it will have to be the PCA9505 40Bit I/O if I plan to drive 32 Outputs.

I have left off ALL of the LED's, (I realise that I can create a separate LED Driver circuit/matrix instead)
Its still powered via the PROP using I2C, Each Expander will have its own address ( not shown in schematic)

Note that I have not included any CAPS for the IC's or the address pin Connections.

With this schematic I should be able to show both CONTINUITY ( GREEN LED) and SHORT CIRCUIT - (RED LED) conditions during test .

My Questions are..
Q1) What do you think of using these devices for my Firing Circuit? Are they Safer? More Suitable?
Q2) Has anyone had any experience using these kinds of switches? In particular the Fault detection circuits?


Regards

Dave M

Post Edited (DavidM) : 5/18/2009 6:58:29 AM GMT

kwinn

Much better. Those chips were an excellent find and make for a much better and safer design. I was going to work on the promised schematic this weekend but it looks like that is no longer necessary.

DavidM

HI Kwinn,

At first, your previous comments were discouraging me, but, you also made me try a little harder , so I have spent the last few days googling for ideas, I have seen these types of devices before , but I never bothered to look at them in detail, now that I have I realise they have already built in them all the circuitry I am trying to figure out. so thanks for making me try a little harder!

Q1) Do you believe the BST452R will do? in regards to AMPS & VOLTS required? I need up to 48V for 500ms ON TIME and possible up to 2-3 amps.

Q2) Any comments regarding my Schematic?

Q3) Do you think CONTINUITY TEST can be easily achieved with these Devices? ( I am assuming that OPEN LOAD = FALSE would be the criteria)



thanks

Dave M

kwinn

I didn't mean to discourage you in any way, but I was concerned about safety. All the extra circuitry for fault detection and the attendant increase in possible failures could have had serious consequences. This circuit looks much better.

According to the data sheet it has a maximum input voltage of 48V and a short circuit current of 4.5A when the input voltage is above 40V which exceeds your requirements.

The only catch may be the length of time it can maintain that current before the chip shuts down. Hard to tell from the data sheet, ant to some extent it depends on the heat sink used on the chip and the ambient operating temperature. Since each chip only fires once I don't think that is a problem, but the only way to be sure is to test one.

It looks like the continuity is always being tested when the device is off ( in is low ). The diagnostic pin should be low if you have continuity. To read out the diagnostic pins you could use a shift register like the 74HC165.

I was going to put a separate readout panel for the status leds in the schematic I planned to send you, but I was planning to have the data from reading the status bits with the 74HC165's be shifted to 74HC595's at the same time using the same clock pin as well as inputting the data to the prop for serial transmission to a remote display if that was required. This would be real time as far as a human being is concerned since it would be updated several times a second and less than a fraction of a second behind the actual status condition.

DavidM

HI,

OK, I have updated the SCHEMATIC to include SHIFT REGISTERS instead of the I2C expanders.
I am sticking with the INFINEON High-Side Switch.

I already have code for both type of shift registers ( from previous projects) and I already have other designs that use multiple registers. I know that you can drive both types with the same set of serial lines, but this is not my concern at the moment.

I have also attached an example of a board layout ( for my testing purposes only)

Q1) Can I run the shift registers at 3.3V? instead of 5V ( as I have shown ) , that's what all the 1K resistors are for.


thanks

Dave M

DavidM

Would anyone like to comment?

thanks

Dave M

kwinn

Sorry David, lost track of this posting for a few days.

Yes, those shift registers will work fine at 3.3V. All of the 74HC series chips will run from 2 to 6V, although slightly faster at the higher voltages. In this circuit
speeds is not a problem.

I am using the '595 for a similar circuit now.

DavidM

Hi kwinn,

Thanks for that,

Does your circuit that "similar" use these kind of "SMART" mosfets that i am using?

Davem

kwinn

No, I am not using the high side mosfet switches. I am using 3 '595's to drive 7 status leds and 15 optically isolated relays, and 2 '165 shift registers to read back 13 switches and limit sensors. The prop, these 5 chips, and optoisolated relays replace several boards of very old electronics and relays.

I have saved the data sheets for both of them for future use, and I thank you for bringing them to my attention.

DavidM

Hi kwinn,

I sent an email to INFINEON in AUSTRALIA, they said they provide technical advice and would like to help me with my project, ( that's nice of them!)

So I received a reply from them for some other questions I had..

---

Hi David,

RE: BTS452R..

Q1) I need to produce about 2-3 amps ( at 48V) per output for less than 500ms and only once. ( I will be "Firing" small PYRO IGNITORS, "SQUIBS" )
Is there any inrush current spike associated firing the SQUIBS? If the inrush current is large, this can cause the BTS452R to shutdown within milliseconds of being turned on.
What is the maximum voltage used to power the circuit? Is it 48V or is there some tolerance in this which will allow it to rise above 58V?

Q2) Can the mosfets (High-Side Switches) be controlled from 3.3V CMOS logic? or only 5V
I have checked the ST data sheet for the 74HC595 and with Vcc of either 5V or 3.3V this device will be suitable to drive the IN pin of the BTS452R. The Status line is will also drive a CMOS input correctly.

Q3) Have I chosen the ideal HIGH SIDE SWITCH for this design? ( i.e. the INFINEON BTS452R) , based on AMPS & VOLTAGE Required and minimum cost?
The BTS452R is a reasonable choice for this design however I would like to know more about the inrush current of the SQUIBS before confirming it as the correct device.

Q4) What kind of HEAT-SINK ( if any is required ), I would like to use just PCB COPPER for this.
Assuming that the maximum PCB temperature is 75°C, just copper will be ok for this design as the switches are not on for long and only used once. The more copper the better, 300mm2 each device if you have the room otherwise as much as you can afford.

Q5) What are the ideal resistor values for each switch? Does this depend on the OUTPUT VOLTAGE?
The value of Rext depends on the load resistance and the quiescent current you requirements you have. If the load is approximately 2A static, then the 100k Rext value you have looks to be ok. Maybe also consider a transistor to switch 48V onto these resistors (one transistor for 8 resistors) to prevent excess quiescent current or voltage on the lines when setting up equipment and also to allow differentiation between open load and short to battery in off state.

Regarding the digital series resistors, the use of 15k resistors is ok however if overvoltage is a possibility, CMOS logic at 5V will be better than 3.3V due to the internal protection Zener diodes on these pins.

Q6) How do you check for CONTINUITY? i.e. OPEN LOAD = FALSE while off
Open load in off is detectable thanks to the 100k resistor you have between 48V and OUT as long as the SQUIB will pull the output to ground with low current applied.

Q7) How do you check for SHORT CIRCUIT while OFF and or while ON
Short circuit to ground is detectable only when on. Short to 48V is detectable in off-state only however this is indistinguishable from an open load if you have the open load resistor fitted. A transistor to switch the 48V onto the resistors to the OUT pin can be used to overcome this limitation.

Q8) Are any other PULL-UP or PULL-DOWN resistors required (apart from what is shown) for interfacing to the SHIFT REGISTERS ( or any I/O logic )
Is there any chance that you will have an overvoltage event when the 48V line will go above 58V? If so, you should consider a ground resistor to limit the current through the internal clamping diode.

Best regards,

Chris A
Infineon Technologies Australia

---

I will follow these suggestions. I do need to get some specs on the ignitors though, I have tried to look and so far haven't found anything. They actually suggest to use 5V instead of 3.3.

Dave M

kwinn

You could run the '595's and 165's with 5V if you want. All you need are 1K or slightly higher ( 1.2K, 1.5K ) value resistors on the lines between the prop and those chips.

DavidM

thanks kwinn

The infineon guy said it would be best to run at 5V as well,


Dave M