Grumpy Bumpy surfed on the net
Grumpy Bumpy found his MOSFET
All the forumistas and perhaps a wise member
Should set Grumpy Bumpy on a glowing ember
This was a nice old thread that has not seen the light of day for a while, so I thought I would give it a BUMP. Over the next couple of days, I will be toying around with several IRF3708 MOSFETs. There are actually three of them in my current project, but for the moment, I will only be working with two of them. One will be controlling the heating element for a 3D printer extruder and the other will be controlling the fan for the extruder. I am not really sure where this is heading, but most likely the MOSFET used for the heating element will be controlled with PWM, whereas the fan will most likely be a simple on and off procedure. However, I may also use PWM for the fan.
At the moment, I am currently thinking about the grand scheme of things, pertaining to heating, cooling, and all around temperature control.
EDIT: I should also note that I will be using Phil's circuit that was shown in Post #15. Providing I made my board correctly
EDIT: I must have used another drawing as a reference, because now I see he has an additional resistor in there.
Grumpy Bumpy surfed on the net
Grumpy Bumpy found his MOSFET
All the forumistas and perhaps a wise member
Should set Grumpy Bumpy on a glowing ember
This was a nice old thread that has not seen the light of day for a while, so I thought I would give it a BUMP. Over the next couple of days, I will be toying around with several IRF3708 MOSFETs. There are actually three of them in my current project, but for the moment, I will only be working with two of them. One will be controlling the heating element for a 3D printer extruder and the other will be controlling the fan for the extruder. I am not really sure where this is heading, but most likely the MOSFET used for the heating element will be controlled with PWM, whereas the fan will most likely be a simple on and off procedure. However, I may also use PWM for the fan.
At the moment, I am currently thinking about the grand scheme of things, pertaining to heating, cooling, and all around temperature control.
EDIT: I should also note that I will be using Phil's circuit that was shown in Post #15. Providing I made my board correctly
Just a glowing ember on that circuit before you get too grumpy and maybe a little bumpy, in that where I join the power ground and signal ground I elect to use a small value resistor just in case. If the power ground got disconnected somehow then the power would want to go through the signal ground and possibly the poor little Propeller which would end up all lumpy and bumpy. Here's some typical values and the idea with R1 is that it should provide a pull-down which might only be needed during reset when the pins float but that is the worst time as that float level may not turn on the MOSFET fully and even a chunky package will start to smoke. R2 is desirable in that it provides "some" protection for faults as well as current limiting (charge/discharge gate capacitance) the Prop I/O while R3 will hopefully fuse long before everything else does if the power ground is disconnected.
JP1 Small Fan
X1-4 Bed Heating Element
X1-6 Extruder Heating Element
12 volt supply
Will it work or am I dreaming?
Oh, I think you had plenty of cooks in the kitchen at the time, it would have been too crowded. Unless you are doing some high frequency PWM you shouldn't need to worry about drivers as long as the MOSFET is a proper logic level device. By proper I mean that they fully switch on at 3.3V, not just start to conduct microamps as datasheets like to spec that as a threshold and then translate that as "logic level" *&%$!
For heating you don't need PWM in the normal sense, it's best handled in "bang-bang" style but you could do a very low frequency PWM if that suits. I don't know about the fan as most of them are BLDC.
As for MOSFETs I try to use smaller SMD packages with very low Rdson and strong conduction levels at 3.3V gate drive. The TO220 packages are over the top even for 10A.
Since you did not say that it won't work, I will assume that it will
For my next design, I will use the previous drawing you provided, but it is to late for this board, since it is already made.
Just in case any one is interested, here is a pic of the board layout.
EDIT: By the way, that is a 2oz. copper clad board.
Yes, 2oz always and I like to leave the power traces unmasked, just tinned which means it's easy to tin up a little more too. But the placement of your mosfets has meant that you have foregone all the advantages of low Rdson and current capability because you now run relatively narrow and long traces (fuses) back to the connector. Best to place the mosfets as close as possible to the connector and make the connector pads suitable for 0.2" spacing screw terminals. I could show you some pcb layouts I've done but basically try to minimise the length and maximise the width as best as you can, even making it double-sided as well. Of course the ground should be made to match and use redundant vias too.
Here is a sample of a H-bridge a did a while back using fairly small dual mosfets and the circuit is mirrored on the other side too to make it a full-bridge.
and just for interest the solder mask exclusion area
Here is a sample of a H-bridge a did a while back using fairly small dual mosfets and the circuit is mirrored on the other side too to make it a full-bridge.
Very cool Peter. I wish I could design circuitry like that, especially with my interest in stepper motors. Please tell me about it. What kind of current can it handle? Is it a chopper drive?
EDIT: Been a while since I have talked drives, I almost forgot the lingo. Now you said full bridge, but I don't suppose that is dual full bridge..... Otherwise you would have said dual
EDIT: And are you still playing with the stepper ICs?
Very cool Peter. I wish I could design circuitry like that, especially with my interest in stepper motors. Please tell me about it. What kind of current can it handle? Is it a chopper drive?
EDIT: Been a while since I have talked drives, I almost forgot the lingo. Now you said full bridge, but I don't suppose that is dual full bridge..... Otherwise you would have said dual
EDIT: And are you still playing with the stepper ICs?
That particular board was just a handy example really, both to show the size of the mosfets with their placement and also the "ruggedness" of the copper connections. This board had a specific use as a fast pin follower so the output would float if the input floated etc. The devices are rated at 20A which is achievable for this design also because it uses full gate drive but I only needed 5 to 10A and yes, this is actually a quad independent half-bridge or dual full-bridge.
I don't generally use a lot of TO220 mosfets but when I do I have the IRLB3036PbF which is a 1.9mohm device rated at 195A @60V and drain to source voltage is around 100mv @30A with 3.3V drive which makes the IRF3708 look a bit pale by comparison.
Haven't done a lot with steppers lately but you never know.
In reference to Post #66, I had forgotten that I changed that design to a single sided board. The pic below is the actual circuitry.
Oh my, I hope you don't expect those long thin traces of the top two mosfets to carry anything more than an amp or so. The bottom mosfet has the most chance of carrying a fair load without vaporizing the copper. This is where you could have placed the mosfets next to each load pin and also used double-sided both for routing and to beef up the current handling.
Here is a suggested layout you could use in the style that you have used (minus your jumper) and even though it is single sided only it has the capacity for loads of many amps.
The space on that board was very cramped. The only mounting holes I had were the two VGA connector holes on the protoboard. If I had any more room they would have been larger.
Your design looks nice but..... you forgot the inputs for the power supply Or are we powering them from the Propeller VDD?
Sorry, just had to be a smart.....
Actually the load for the top MOSFET is 0.24A
and the load for the middle MOSFET is about 1.8A
I am more worried about the lower MOSFET because that will be a huge load of unknown amperage at this point.
Here is a good dual n-mosfet, that is fully-on with 3.3v on the gate.
Shared Source (gnd), Drain pins both facing same way.
600mA pulse, 150mA continues
The space on that board was very cramped. The only mounting holes I had were the two VGA connector holes on the protoboard. If I had any more room they would have been larger.
Your design looks nice but..... you forgot the inputs for the power supply Or are we powering them from the Propeller VDD?
Sorry, just had to be a smart.....
Actually the load for the top MOSFET is 0.24Ao
and the load for the middle MOSFET is about 1.8A
I am more worried about the lower MOSFET because that will be a huge load of unknown amperage at this point.
What power supply? The mosfet board only requires input signals, mosfet outputs, and grounds. Besides this just had your pcb to model on and the top right hand "jumper" didn't look right but the main point is to illustrate that placement is very important. If you get all your components and connectors in the right place then routing the tracks should be easier and cleaner. But considering your load requirements I wonder why you bother with such large mosfets. A dual mosfet in an so-8 package should handle most of your needs. Even just single mosfets in DIP8 will work too.
btw, any "pulsed" current figures are not really applicable as those pulses are of such short duration and the duty cycle is extremely low. Only the continuous current figure matters in most applications.
I see what you did now. The hot completely bypasses the board. Interesting. Why have 3 seperate incoming grounds?
In my design, I figured 12V+ and 12- coming in and (3)-12V+ and (3)-12V- going out. I believe that will help with wire and cable managability.
EDIT: 1 ground must come in.... Then only 2 going out?
This was only meant as an example as I had no parameters to work with other than your layout. If you want a complete layout then that only comes after understanding the requirements of the circuit, the application, the installation etc. But you mean power ground not -12V don't you as -12 is always quoted with respects to the "ground" or common.
Don't confuse your pcb with cable manageability, especially when things are cramped as you say. Although you only need a 4 way connector for your mosfets you can see that the spacing afforded by the 6-way allows the high current drain connections to be more direct and shorter while there is still room to make these better again. You don't have to use all the grounds of course but they are there anyway, no need to fuss about the ones that aren't needed.
Because there is room to spare on the board you can even reduce its height if necessary by bringing the 4-pin input connector down centrally just above the resistors.
Best to run your load supply into separate terminals as these can get quite cumbersome and would put a lot of undue mechanical stress on a tiny pcb.
EDIT: Here is an optimized board layout based on the limited parameters I am working with.
Peter just so you know... I like your design. If I had bypassed the hot, my board would certainly have a lot more meat on the traces.
Those traces are the most important thing on a mosfet board, they can't be left to second or third place. If you really insist on having the "hot" on the board rather than off on a suitable screw terminal junction (just like those ones that float around in electrical junction boxes) then this is very easy too. I missed what that long skinny trace was doing on your pcb and didn't realize that you were trying to carry power on it. But here is a slightly amended circuit that gives you 2 hot terminals that you can link with a wire or zero ohm link on a single sided pcb.
Much neater though IMO to connect to a floating hot terminal as it can be a single large terminal and you can also avoid any potential disasters on the mosfet pcb with what would otherwise have no function on the pcb itself. This is what I tend to do especially since the hot supply wiring can become quite cumbersome.
I certainly learned a couple of things with this discussion, however I at least have to give my board a roll of the dice, especially since it is already made.
And at this point, at least until I make a design change, the mounting holes need to be more in the center of the board, since it is attached to a ProtoBoard that fits lengthwise in a junction box with a 4" depth. In the photo below, you can see the board outlined in red. The prop plug will be eliminated to allow the board to fit in a 4" deep box.
As an end result, providing the controller works well, I will most likely have a board made at a PCB house, including changes that you suggest.
I certainly learned a couple of things with this discussion, however I at least have to give my board a roll of the dice, especially since it is already made.
And at this point, at least until I make a design change, the mounting holes need to be more in the center of the board, since it is attached to a ProtoBoard that fits lengthwise in a junction box with a 4" depth. In the photo below, you can see the board outlined in red. The prop plug will be eliminated to allow the board to fit in a 4" deep box.
As an end result, providing the controller works well, I will most likely have a board made at a PCB house, including changes that you suggest.
Well seeing that then it is another simple matter to rearrange the mosfets and reduce the length of the board down to 1.1" while moving the mounting centrally. My screw terminal footprint is a little larger than the one you are using so I've allowed for a bit more squeezing in. I'm a bit worried about that big heatsink on your mosfet as that shouldn't be the case unless you are trying to do some high frequency PWM in which case you really need good mosfet drivers.
But if I were to do any kind of pcb it wouldn't be this little mosfet patch board, I would be far more inclined to do the whole Prop pcb and also put a decent switching regulator on it too and end up laying the mosfets flat to the board. There is nothing at all complex about the main pcb so why wouldn't you perhaps do it that way?
I already have various pcbs smaller than the one you are using which take those two of those little plug-in modules including the quad half-bridge you saw in post #67. They include 12-way screw terminals for power, RS485, and up to 8 module I/O as well as serial USB or Bluetooth etc. They would say you a lot of work and having to get pcbs done.
But if I were to do any kind of pcb it wouldn't be this little mosfet patch board, I would be far more inclined to do the whole Prop pcb and also put a decent switching regulator on it too and end up laying the mosfets flat to the board. There is nothing at all complex about the main pcb so why wouldn't you perhaps do it that way?
Yea, I wasn't referring to a small pcb, I was talking about the whole controller, incorporating your design into the main board. Odds are that when I have the machine fully functional, I will layout out a general design draft that would include all the important components with wiring, and submit it for proofing to someone like you. For instance, whereas the photo above shows an upright ADC breakout board, when design a main controller board, I would actually add all the components of the breakout board to the actual layout and although not shown, the same would hold true for the Propeller Memory Card components.
Additionally, in your design, I see you added the component outlined in red. What is it? And what is it's purpose?
Driving a mosfet hard.
Though a mosfets gate don't hardly use any current while staying in one state.
For very short time it does use some power when you charge up its capacitance.
If you can give the gate a large burst of power, the faster you can switch it (in the mhz) and also less power loss.
So you need push-pull, the props pin is push-pull but at 40ma not much power behind it.
And with 3.1v it may not fully turn most Mosfets on, so if you use an external push-pull that have a 5v Vcc you're in power-mosfet-haven.
Pin 4 goes to the mosfet (a resistor in series with a value of 1ohm to 3ohm is good design )
Tony,
This is an interesting part that operates in saturation mode. With a 1k base resistor the VCE should be <80mV and control 100ma of Icc. See Table 7 specs.
I've used this part from Onsemi to drive relays - 5V and 12V: NUD3124DMT1G
It is a SOT23 with MOSFET, diodes / zeners, and resistors in one package. I'm controlling it with 3V LV logic, 74AHC594 8 bit shift registers tied together in series. Data is clocked in serially at 15MHz. It's tiny, sits next to each relay, has no external components and controlled by low voltage logic. Great part.
Additionally, in your design, I see you added the component outlined in red. What is it? And what is it's purpose?
You know I mentioned that isolation resistor between the signal ground and power? Well this is it and its purpose is to protect the signal ground path in the event of the power ground being disconnected. If this was the case then all the many amps from the load would try to flow through the signal ground which is more than likely not designed or capable of handling this current and what's more, it has the potential to take out the Prop and other devices in its path. So the resistor will both limit the current and if sustained will fuse open thereby protecting the signal circuits. I don't know if this is any kind of standard practice though, it's just what I do.
You know I mentioned that isolation resistor between the signal ground and power? Well this is it and its purpose is to protect the signal ground path in the event of the power ground being disconnected. If this was the case then all the many amps from the load would try to flow through the signal ground which is more than likely not designed or capable of handling this current and what's more, it has the potential to take out the Prop and other devices in its path. So the resistor will both limit the current and if sustained will fuse open thereby protecting the signal circuits. I don't know if this is any kind of standard practice though, it's just what I do.
Okay, I get it and I believe that is good practice, but I imagine a fuse would be much better. So if it was a fuse instead, what kind of ampacity would be required to trip a common MOSFET? And at this common ampacity, if current were to reverse flow through the fuse due to loss of ground, what is the likelihood of damaging the prop anyhow?
Okay, I get it and I believe that is good practice, but I imagine a fuse would be much better. So if it was a fuse instead, what kind of ampacity would be required to trip a common MOSFET? And at this common ampacity, if current were to reverse flow through the fuse due to loss of ground, what is the likelihood of damaging the prop anyhow?
Trouble with fuses, takes too long to blow
and in the meantime, too much current flows
Take a resistor, as cheap as a seed
Limits the current, and does what you need
Btw, even if your Prop didn't get any high current flowing through the chip itself there could be a voltage differential between the Prop's grounds pins which could end up damaging it or the ground potential itself could rise leading to other problems.
I've been using a Crydom CMX100D10 solid state relay to control a 28 VDC solenoid coil. A 2N3904, configured as a low side driver, controls the 5V logic for the CMX100.
Transient voltage protection is provided by a Littlelfuse 5KP33CA TVS Diode.
Some people have grumbled about the cost of these relays but, for my application, I consider them cheap.
Comments
Grumpy Bumpy found his MOSFET
All the forumistas and perhaps a wise member
Should set Grumpy Bumpy on a glowing ember
This was a nice old thread that has not seen the light of day for a while, so I thought I would give it a BUMP. Over the next couple of days, I will be toying around with several IRF3708 MOSFETs. There are actually three of them in my current project, but for the moment, I will only be working with two of them. One will be controlling the heating element for a 3D printer extruder and the other will be controlling the fan for the extruder. I am not really sure where this is heading, but most likely the MOSFET used for the heating element will be controlled with PWM, whereas the fan will most likely be a simple on and off procedure. However, I may also use PWM for the fan.
At the moment, I am currently thinking about the grand scheme of things, pertaining to heating, cooling, and all around temperature control.
EDIT: I should also note that I will be using Phil's circuit that was shown in Post #15. Providing I made my board correctly
EDIT: I must have used another drawing as a reference, because now I see he has an additional resistor in there.
Just a glowing ember on that circuit before you get too grumpy and maybe a little bumpy, in that where I join the power ground and signal ground I elect to use a small value resistor just in case. If the power ground got disconnected somehow then the power would want to go through the signal ground and possibly the poor little Propeller which would end up all lumpy and bumpy. Here's some typical values and the idea with R1 is that it should provide a pull-down which might only be needed during reset when the pins float but that is the worst time as that float level may not turn on the MOSFET fully and even a chunky package will start to smoke. R2 is desirable in that it provides "some" protection for faults as well as current limiting (charge/discharge gate capacitance) the Prop I/O while R3 will hopefully fuse long before everything else does if the power ground is disconnected.
I sure could have used a little of your guidance when I was designing my board
Anyhow, this is what I ended up with:
R1 220 Ohms
R2 10K Ohms
R3 220 Ohms
R4 10K Ohms
R5 220 Ohms
R6 10K Ohms
JP1 Small Fan
X1-4 Bed Heating Element
X1-6 Extruder Heating Element
12 volt supply
Will it work or am I dreaming?
Oh, I think you had plenty of cooks in the kitchen at the time, it would have been too crowded. Unless you are doing some high frequency PWM you shouldn't need to worry about drivers as long as the MOSFET is a proper logic level device. By proper I mean that they fully switch on at 3.3V, not just start to conduct microamps as datasheets like to spec that as a threshold and then translate that as "logic level" *&%$!
For heating you don't need PWM in the normal sense, it's best handled in "bang-bang" style but you could do a very low frequency PWM if that suits. I don't know about the fan as most of them are BLDC.
As for MOSFETs I try to use smaller SMD packages with very low Rdson and strong conduction levels at 3.3V gate drive. The TO220 packages are over the top even for 10A.
Btw, 220R is fine
Since you did not say that it won't work, I will assume that it will
For my next design, I will use the previous drawing you provided, but it is to late for this board, since it is already made.
Just in case any one is interested, here is a pic of the board layout.
EDIT: By the way, that is a 2oz. copper clad board.
Yes, 2oz always and I like to leave the power traces unmasked, just tinned which means it's easy to tin up a little more too. But the placement of your mosfets has meant that you have foregone all the advantages of low Rdson and current capability because you now run relatively narrow and long traces (fuses) back to the connector. Best to place the mosfets as close as possible to the connector and make the connector pads suitable for 0.2" spacing screw terminals. I could show you some pcb layouts I've done but basically try to minimise the length and maximise the width as best as you can, even making it double-sided as well. Of course the ground should be made to match and use redundant vias too.
Here is a sample of a H-bridge a did a while back using fairly small dual mosfets and the circuit is mirrored on the other side too to make it a full-bridge.
and just for interest the solder mask exclusion area
Very cool Peter. I wish I could design circuitry like that, especially with my interest in stepper motors. Please tell me about it. What kind of current can it handle? Is it a chopper drive?
EDIT: Been a while since I have talked drives, I almost forgot the lingo. Now you said full bridge, but I don't suppose that is dual full bridge..... Otherwise you would have said dual
EDIT: And are you still playing with the stepper ICs?
That particular board was just a handy example really, both to show the size of the mosfets with their placement and also the "ruggedness" of the copper connections. This board had a specific use as a fast pin follower so the output would float if the input floated etc. The devices are rated at 20A which is achievable for this design also because it uses full gate drive but I only needed 5 to 10A and yes, this is actually a quad independent half-bridge or dual full-bridge.
I don't generally use a lot of TO220 mosfets but when I do I have the IRLB3036PbF which is a 1.9mohm device rated at 195A @60V and drain to source voltage is around 100mv @30A with 3.3V drive which makes the IRF3708 look a bit pale by comparison.
Haven't done a lot with steppers lately but you never know.
Oh my, I hope you don't expect those long thin traces of the top two mosfets to carry anything more than an amp or so. The bottom mosfet has the most chance of carrying a fair load without vaporizing the copper. This is where you could have placed the mosfets next to each load pin and also used double-sided both for routing and to beef up the current handling.
Here is a suggested layout you could use in the style that you have used (minus your jumper) and even though it is single sided only it has the capacity for loads of many amps.
The space on that board was very cramped. The only mounting holes I had were the two VGA connector holes on the protoboard. If I had any more room they would have been larger.
Your design looks nice
Sorry, just had to be a smart.....
Actually the load for the top MOSFET is 0.24A
and the load for the middle MOSFET is about 1.8A
I am more worried about the lower MOSFET because that will be a huge load of unknown amperage at this point.
Shared Source (gnd), Drain pins both facing same way.
600mA pulse, 150mA continues
$0.157 each @100, SC74A SOT25-5 (0.95mm pin spacing = average person can smt it)
http://www.mouser.com/ProductDetail/ON-Semiconductor/CPH5617-TL-E/?qs=%2fha2pyFadui98sBmglo16LeJOF377SG8C3AmnN%252bSaW9nVdaHZ2EDyg%3d%3d
What power supply? The mosfet board only requires input signals, mosfet outputs, and grounds. Besides this just had your pcb to model on and the top right hand "jumper" didn't look right but the main point is to illustrate that placement is very important. If you get all your components and connectors in the right place then routing the tracks should be easier and cleaner. But considering your load requirements I wonder why you bother with such large mosfets. A dual mosfet in an so-8 package should handle most of your needs. Even just single mosfets in DIP8 will work too.
btw, any "pulsed" current figures are not really applicable as those pulses are of such short duration and the duty cycle is extremely low. Only the continuous current figure matters in most applications.
I see what you did now. The hot completely bypasses the board. Interesting. Why have 3 seperate incoming grounds?
In my design, I figured 12V+ and 12- coming in and (3)-12V+ and (3)-12V- going out. I believe that will help with wire and cable managability.
EDIT: 1 ground must come in.... Then only 2 going out?
This was only meant as an example as I had no parameters to work with other than your layout. If you want a complete layout then that only comes after understanding the requirements of the circuit, the application, the installation etc. But you mean power ground not -12V don't you as -12 is always quoted with respects to the "ground" or common.
Don't confuse your pcb with cable manageability, especially when things are cramped as you say. Although you only need a 4 way connector for your mosfets you can see that the spacing afforded by the 6-way allows the high current drain connections to be more direct and shorter while there is still room to make these better again. You don't have to use all the grounds of course but they are there anyway, no need to fuss about the ones that aren't needed.
Because there is room to spare on the board you can even reduce its height if necessary by bringing the 4-pin input connector down centrally just above the resistors.
Best to run your load supply into separate terminals as these can get quite cumbersome and would put a lot of undue mechanical stress on a tiny pcb.
EDIT: Here is an optimized board layout based on the limited parameters I am working with.
Those traces are the most important thing on a mosfet board, they can't be left to second or third place. If you really insist on having the "hot" on the board rather than off on a suitable screw terminal junction (just like those ones that float around in electrical junction boxes) then this is very easy too. I missed what that long skinny trace was doing on your pcb and didn't realize that you were trying to carry power on it. But here is a slightly amended circuit that gives you 2 hot terminals that you can link with a wire or zero ohm link on a single sided pcb.
Much neater though IMO to connect to a floating hot terminal as it can be a single large terminal and you can also avoid any potential disasters on the mosfet pcb with what would otherwise have no function on the pcb itself. This is what I tend to do especially since the hot supply wiring can become quite cumbersome.
I certainly learned a couple of things with this discussion, however I at least have to give my board a roll of the dice, especially since it is already made.
And at this point, at least until I make a design change, the mounting holes need to be more in the center of the board, since it is attached to a ProtoBoard that fits lengthwise in a junction box with a 4" depth. In the photo below, you can see the board outlined in red. The prop plug will be eliminated to allow the board to fit in a 4" deep box.
As an end result, providing the controller works well, I will most likely have a board made at a PCB house, including changes that you suggest.
Well seeing that then it is another simple matter to rearrange the mosfets and reduce the length of the board down to 1.1" while moving the mounting centrally. My screw terminal footprint is a little larger than the one you are using so I've allowed for a bit more squeezing in. I'm a bit worried about that big heatsink on your mosfet as that shouldn't be the case unless you are trying to do some high frequency PWM in which case you really need good mosfet drivers.
But if I were to do any kind of pcb it wouldn't be this little mosfet patch board, I would be far more inclined to do the whole Prop pcb and also put a decent switching regulator on it too and end up laying the mosfets flat to the board. There is nothing at all complex about the main pcb so why wouldn't you perhaps do it that way?
I already have various pcbs smaller than the one you are using which take those two of those little plug-in modules including the quad half-bridge you saw in post #67. They include 12-way screw terminals for power, RS485, and up to 8 module I/O as well as serial USB or Bluetooth etc. They would say you a lot of work and having to get pcbs done.
Yea, I wasn't referring to a small pcb, I was talking about the whole controller, incorporating your design into the main board. Odds are that when I have the machine fully functional, I will layout out a general design draft that would include all the important components with wiring, and submit it for proofing to someone like you. For instance, whereas the photo above shows an upright ADC breakout board, when design a main controller board, I would actually add all the components of the breakout board to the actual layout and although not shown, the same would hold true for the Propeller Memory Card components.
Additionally, in your design, I see you added the component outlined in red. What is it? And what is it's purpose?
This is an interesting part that operates in saturation mode. With a 1k base resistor the VCE should be <80mV and control 100ma of Icc. See Table 7 specs.
It is a SOT23 with MOSFET, diodes / zeners, and resistors in one package. I'm controlling it with 3V LV logic, 74AHC594 8 bit shift registers tied together in series. Data is clocked in serially at 15MHz. It's tiny, sits next to each relay, has no external components and controlled by low voltage logic. Great part.
Link: http://www.digikey.com/product-search/en?x=0&y=0&lang=en&site=us&keywords=NUD3124DMT1GOSTR-ND
You know I mentioned that isolation resistor between the signal ground and power? Well this is it and its purpose is to protect the signal ground path in the event of the power ground being disconnected. If this was the case then all the many amps from the load would try to flow through the signal ground which is more than likely not designed or capable of handling this current and what's more, it has the potential to take out the Prop and other devices in its path. So the resistor will both limit the current and if sustained will fuse open thereby protecting the signal circuits. I don't know if this is any kind of standard practice though, it's just what I do.
Okay, I get it and I believe that is good practice, but I imagine a fuse would be much better. So if it was a fuse instead, what kind of ampacity would be required to trip a common MOSFET? And at this common ampacity, if current were to reverse flow through the fuse due to loss of ground, what is the likelihood of damaging the prop anyhow?
Trouble with fuses, takes too long to blow
and in the meantime, too much current flows
Take a resistor, as cheap as a seed
Limits the current, and does what you need
Btw, even if your Prop didn't get any high current flowing through the chip itself there could be a voltage differential between the Prop's grounds pins which could end up damaging it or the ground potential itself could rise leading to other problems.
Thanks for all your input. I learned quite a few things during our discussion.
Bruce
Transient voltage protection is provided by a Littlelfuse 5KP33CA TVS Diode.
Some people have grumbled about the cost of these relays but, for my application, I consider them cheap.
Sandy