Is your PWM object the type you recommend? I downloaded 3 of them and don't understand the code at all, so I'll have to work on learning how to program for a while before I can even use the object the way I want it to work in my program. I am planning on a RCTIME - PWM (maybe PID) loop for 7 Cogs and a data link for the 8th.
Thanks also on explaining the ground problem I was experiencing.
I don't know details like part numbers but it seems a fast opto-isolator function is available in fiber optic cable repeater amps. The signal must be converted from light to electrical then amplified and converted to light. So the high speed parts must exist.
kwinn,
As far as I can tell, I have a proportional cartridge hydraulic valve which is the same as a hydraulic servo valve but some vendors want more money for the "servo" brand.
your doing your "homework" well. (downloading PWM-objects and looking into the code)
If you post or attach some code with a question you will get answers and support quickly.
The PWM32_v4-object can create 32 different pwm-signals each with his own frequency and duty.
So it is not nescessary to occupy seven cogs for seven PWM-signals.
Please post more information about your project. If we (the forum-members) have an overview about what you want to achieve
good solutions can be found.
best regards
Stefan
Edit: wow Kirk! That's what I call unbiased free thinking. Of course glasfiber-optics need really fast optoelecronics.
With either kind, you will need a MOSFET driver between the isolator and the MOSFET to provide adequate gate drive. Examples are the Micrel MIC44xx series low-side drivers:
I have a new question the datasheet of the micrel MOS-Fet driver says 3A. Does this mean the driver-chip can deliver 3A constantly?
or just as a short impulse to load the input capacitance of a big POWER-MOS-FET?
best regards
Those chips are designed to drive capacitive loads. As such, they can handle high peak currents into any kind of load. But they're not packaged in such a way to accommodate the thermal interface required to dissipate heat from constant-high-current use. That said, I've used the MIC4469 (with paralleled outputs) to drive a speaker directly in a class D arrangement:
Improper grounding can cause other concerns, in addition to the previously mentioned one about creating an unintentional DC bias between points in the circuit if too much impedance is present in the grounding wires.
Although not necessarily the case here, there is normally a concern about digital noise (caused by the ultra-fast transitions associated with square-wave signals) being conducted through the ground path and affecting the operation of any super-sensitive analog circuitry, which is a primary reason for using bypass capacitors, and is discussed in detail here: Grounding in Mixed Signal Systems
Im thinking that the potential problem here is exactly the opposite. There is a likelihood that inductive kickback from the driven analog circuits (1A coils running at 12V and 20kHz) could be enough to thoroughly thrash the 3v3 logic circuits, and that this is the main reason why some folks here are suggesting that optoisolation makes sense (in order to completely separate the two power supplies from one another, with no common ground at all). Also, some EMI shielding of the 12V stuff may be necessary.
Comments
Is your PWM object the type you recommend? I downloaded 3 of them and don't understand the code at all, so I'll have to work on learning how to program for a while before I can even use the object the way I want it to work in my program. I am planning on a RCTIME - PWM (maybe PID) loop for 7 Cogs and a data link for the 8th.
Thanks also on explaining the ground problem I was experiencing.
I don't know details like part numbers but it seems a fast opto-isolator function is available in fiber optic cable repeater amps. The signal must be converted from light to electrical then amplified and converted to light. So the high speed parts must exist.
kwinn,
As far as I can tell, I have a proportional cartridge hydraulic valve which is the same as a hydraulic servo valve but some vendors want more money for the "servo" brand.
Everyone,
Thanks for your help.
your doing your "homework" well. (downloading PWM-objects and looking into the code)
If you post or attach some code with a question you will get answers and support quickly.
The PWM32_v4-object can create 32 different pwm-signals each with his own frequency and duty.
So it is not nescessary to occupy seven cogs for seven PWM-signals.
Please post more information about your project. If we (the forum-members) have an overview about what you want to achieve
good solutions can be found.
best regards
Stefan
Edit: wow Kirk! That's what I call unbiased free thinking. Of course glasfiber-optics need really fast optoelecronics.
googling with ultrafast optocoupler I came up with this http://www.nve.com/Downloads/il600.pdf
quality has its price but still affordable $5.67 for a 2-channel isolator http://www.nve.com/webstore/catalog/product_info.php?cPath=30_37&products_id=388
@Phil what do you think about his device?
With either kind, you will need a MOSFET driver between the isolator and the MOSFET to provide adequate gate drive. Examples are the Micrel MIC44xx series low-side drivers:
-Phil
I have a new question the datasheet of the micrel MOS-Fet driver says 3A. Does this mean the driver-chip can deliver 3A constantly?
or just as a short impulse to load the input capacitance of a big POWER-MOS-FET?
best regards
Stefan
Those chips are designed to drive capacitive loads. As such, they can handle high peak currents into any kind of load. But they're not packaged in such a way to accommodate the thermal interface required to dissipate heat from constant-high-current use. That said, I've used the MIC4469 (with paralleled outputs) to drive a speaker directly in a class D arrangement:
-Phil
Although not necessarily the case here, there is normally a concern about digital noise (caused by the ultra-fast transitions associated with square-wave signals) being conducted through the ground path and affecting the operation of any super-sensitive analog circuitry, which is a primary reason for using bypass capacitors, and is discussed in detail here: Grounding in Mixed Signal Systems
Im thinking that the potential problem here is exactly the opposite. There is a likelihood that inductive kickback from the driven analog circuits (1A coils running at 12V and 20kHz) could be enough to thoroughly thrash the 3v3 logic circuits, and that this is the main reason why some folks here are suggesting that optoisolation makes sense (in order to completely separate the two power supplies from one another, with no common ground at all). Also, some EMI shielding of the 12V stuff may be necessary.