Hello, I am trying to build a drive control system, which should be able to achieve the switching frequency of the inverter at about 100kHz. It means that my microcontroller should sample with the frequency about 200kHz (because of the PWM scheme I use). Do you possibly know any microcontroller or DSP which will be able to do the A/D conversions (at least 4 analog inputs), drive control, machine model calculations and PWM (6 signals) with such a frequency? Thanks, Pawel Szczupak PS: Greetings to everybody, for I am for the first time here :)
Microcontroller or DSP with 250kHz sampling rate
Started by ●October 26, 2004
Reply by ●October 26, 20042004-10-26
I suppose you need the ADCs to be onchip. The following are my suggestions in the microcontroller world (all from Analog Devices): - ADuC841/2/3: 8 channel, 400ksps ADC, 12 bits, 62kB Flash, 2kB SRAM, 16 - 20 MIPS. 8051 Architecture. - ADuC702x: 5 - 16 channel, 1 MPSPS ADC, 12 bits, 62kB Flash, 8kB SRAM, 45 MIPS, ARM7 TDMI Architecture. Regards, JaaC pawel <pawl6@OHNESPAM.op.pl> wrote in message news:<clkrit$gpl$1@news.onet.pl>...> Hello, > I am trying to build a drive control system, which should be able to > achieve the switching frequency of the inverter at about 100kHz. It > means that my microcontroller should sample with the frequency about > 200kHz (because of the PWM scheme I use). > Do you possibly know any microcontroller or DSP which will be able to > do the A/D conversions (at least 4 analog inputs), drive control, > machine model calculations and PWM (6 signals) with such a frequency? > > Thanks, > Pawel Szczupak > > PS: Greetings to everybody, for I am for the first time here :)
Reply by ●October 26, 20042004-10-26
Check out TI's TMS320F2812: http://focus.ti.com/docs/prod/folders/print/tms320f2812.html Just to give you a quick overview it has a 12-bit 16-channel 12.5 MSPS ADC, 256 kB flash, 36 kB SRAM, 150 MIPS, SPI, 2 SCI/UART, McBSP, CAN, 16 PWM, watchdog, external memory interface. If that is overkill for you there are some cheaper devices available in the 240xA family. If you want to go cheaper yet some of the MSP430 microcontrollers might work too. Brad "Jaime Andres Aranguren Cardona" <jaime.aranguren@ieee.org> wrote in message news:14a86f87.0410260345.3120f7cc@posting.google.com...>I suppose you need the ADCs to be onchip. > > The following are my suggestions in the microcontroller world (all > from Analog Devices): > > - ADuC841/2/3: 8 channel, 400ksps ADC, 12 bits, 62kB Flash, 2kB SRAM, > 16 - 20 MIPS. 8051 Architecture. > - ADuC702x: 5 - 16 channel, 1 MPSPS ADC, 12 bits, 62kB Flash, 8kB > SRAM, 45 MIPS, ARM7 TDMI Architecture. > > Regards, > > JaaC > > pawel <pawl6@OHNESPAM.op.pl> wrote in message > news:<clkrit$gpl$1@news.onet.pl>... >> Hello, >> I am trying to build a drive control system, which should be able to >> achieve the switching frequency of the inverter at about 100kHz. It >> means that my microcontroller should sample with the frequency about >> 200kHz (because of the PWM scheme I use). >> Do you possibly know any microcontroller or DSP which will be able to >> do the A/D conversions (at least 4 analog inputs), drive control, >> machine model calculations and PWM (6 signals) with such a frequency? >> >> Thanks, >> Pawel Szczupak >> >> PS: Greetings to everybody, for I am for the first time here :)
Reply by ●October 26, 20042004-10-26
Check the TMS320F2810/12 DSPs. http://focus.ti.com/docs/prod/folders/print/tms320f2812.html 12-Bit ADC, 16 Channels 2 x 8 Channel Input Multiplexer Two Sample-and-Hold Single/Simultaneous Conversions Fast Conversion Rate: 80 ns/12.5 MSPS http://www.spectrumdigital.com/cgi/catalog.cgi?show_product=761128 http://www.spectrumdigital.com/cgi/catalog.cgi?show_product=761129 Dziekuje, Yuri
Reply by ●October 26, 20042004-10-26
Hello, Dnia 26.10.04 13:45, U�ytkownik Jaime Andres Aranguren Cardona napisa�:> - ADuC841/2/3: 8 channel, 400ksps ADC, 12 bits, 62kB Flash, 2kB SRAM, > 16 - 20 MIPS. 8051 Architecture.This one seems not to have 3-Phase PWM, which is a must in my system.> - ADuC702x: 5 - 16 channel, 1 MPSPS ADC, 12 bits, 62kB Flash, 8kB > SRAM, 45 MIPS, ARM7 TDMI Architecture.This one seems to be ok, I will have to check the details. Thank you very much for the information. Greetings, Pawel Szczupak> Regards, > > JaaC > > pawel <pawl6@OHNESPAM.op.pl> wrote in message news:<clkrit$gpl$1@news.onet.pl>... > >>Hello, >> I am trying to build a drive control system, which should be able to >>achieve the switching frequency of the inverter at about 100kHz. It >>means that my microcontroller should sample with the frequency about >>200kHz (because of the PWM scheme I use). >> Do you possibly know any microcontroller or DSP which will be able to >>do the A/D conversions (at least 4 analog inputs), drive control, >>machine model calculations and PWM (6 signals) with such a frequency? >> >>Thanks, >>Pawel Szczupak >> >>PS: Greetings to everybody, for I am for the first time here :)
Reply by ●October 26, 20042004-10-26
Hello, thank you very much for your answers. I will look at the one from TMS320F2812 and this from ADuC7024. I have heard about dSMC, do you know it? How fast it is? Is it able to do what I would like it to do? I can not any find information about it. Greetings, Pawel Szczupak PS: Prosze bardzo Yuri :)
Reply by ●October 26, 20042004-10-26
pawel wrote:> Hello, > I am trying to build a drive control system, which should be able to > achieve the switching frequency of the inverter at about 100kHz. It > means that my microcontroller should sample with the frequency about > 200kHz (because of the PWM scheme I use). > Do you possibly know any microcontroller or DSP which will be able > to do the A/D conversions (at least 4 analog inputs), drive control, > machine model calculations and PWM (6 signals) with such a frequency? > > Thanks, > Pawel Szczupak > > PS: Greetings to everybody, for I am for the first time here :)That's asking a whole lot from your microprocessor. Sampling a TMS320F2812 at 80kHz (with a uP clock of 150MHz) is doable but pushing it. Why do you feel you have to sample at faster than your PWM rate? Most of the DSPs that are designed for motor control have very good PWM generators that are augmented with overcurrent shutdowns to allow you to run the processor core much slower than the PWM. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com
Reply by ●October 26, 20042004-10-26
> That's asking a whole lot from your microprocessor. Sampling a > TMS320F2812 at 80kHz (with a uP clock of 150MHz) is doable but pushing it.That is what I am affraid of. I have to be able to sample at least with 100kHz (see lower part). I am affraid this frequency is a must for me. Do you know if there are other uC which are capable of doing such a work in such a time?> Why do you feel you have to sample at faster than your PWM rate? Most > of the DSPs that are designed for motor control have very good PWM > generators that are augmented with overcurrent shutdowns to allow you to > run the processor core much slower than the PWM.In my PWM scheme I have sampled at every zero vector. It was sometimes useful to be able to change the PWM duty ratios during the second part of the PWM cycle (let's say 0int-1-2-7int-2changed duty-1changed duty-0int). But I think it is possible in my case to sample only during 0 vector. I'll have to think about it. Best Regards, Pawel Szczupak
Reply by ●October 26, 20042004-10-26
pawel wrote:>> That's asking a whole lot from your microprocessor. Sampling a >> TMS320F2812 at 80kHz (with a uP clock of 150MHz) is doable but pushing >> it. > > > That is what I am affraid of. I have to be able to sample at least with > 100kHz (see lower part). I am affraid this frequency is a must for me. > Do you know if there are other uC which are capable of doing such a work > in such a time? > >> Why do you feel you have to sample at faster than your PWM rate? Most >> of the DSPs that are designed for motor control have very good PWM >> generators that are augmented with overcurrent shutdowns to allow you >> to run the processor core much slower than the PWM. > > > In my PWM scheme I have sampled at every zero vector. It was sometimes > useful to be able to change the PWM duty ratios during the second part > of the PWM cycle (let's say 0int-1-2-7int-2changed duty-1changed > duty-0int). But I think it is possible in my case to sample only during > 0 vector. I'll have to think about it. > > Best Regards, > Pawel Szczupak > >You may want to look at floating point DSP chips, these may be fast enough (but wouldn't include the PWM generators). Another choice is to use an FPGA; this will let your run your algorithm on highly parallel data paths and pretty much go as fast as you want, perhaps with a processor of some sort to provide supervision. The downside is cost, but if you are careful with your algorithm and digital design then you may be able to fit it into a $20 FPGA. Xilinx and Altera both have cost-conscious FPGA lines, I don't know about the other vendors. Atmel came out with an FPGA with an embedded 8-bit processor, but I don't know if they're still selling it or how much it costs. You'll have a big tradeoff between FPGA size and processor performance, but that's your job... -- Tim Wescott Wescott Design Services http://www.wescottdesign.com
Reply by ●October 26, 20042004-10-26
