1
\$\begingroup\$

I am looking to build a circuit to control five different FET drivers individually. The input signal would be a single shot from a function generator, which should produce five output signals, with the same frequency and phase but the duty cycles need to be individually controllable. Basically, I am trying to replace five function generators (synchronized to the same clock) with one single function generator and a circuit board due to limited resources. The outputs need to have only two levels i.e pulse wave

EDIT: Olin's suggestion (using a MCU) or a FPGA seem to be the best options but any analog alternatives are much appreciated. Thanks :)

\$\endgroup\$
  • \$\begingroup\$ How much skew is allowable between the output signals' rising edges (for a rising edge on the input, for example)? \$\endgroup\$ – Brendan Simpson Mar 31 '16 at 20:12
  • \$\begingroup\$ Ideally, since I'm looking to basically replicate a func. gen I need it to be as matched as possible. Also the target frequency is 5-10MHz so anything more than 10ns would have an impact \$\endgroup\$ – Manu Mar 31 '16 at 20:20
  • \$\begingroup\$ How do you want to control the duty cycle of the five outputs? \$\endgroup\$ – Dave Tweed Mar 31 '16 at 20:20
  • \$\begingroup\$ That could be done with a micro, as Olin Lanthrop suggested, but perhaps an FPGA or CPLD would be better. \$\endgroup\$ – Brendan Simpson Mar 31 '16 at 20:21
  • \$\begingroup\$ @Dave: Possibly with passive components on the board \$\endgroup\$ – Manu Mar 31 '16 at 20:21
2
\$\begingroup\$

Since you only need pulses of up to about 200 ns, and control can be through passive components, I would just use three 74x123 chips, each of which contains two monostable multivibrators that can be controlled by one capacitor and one resistor each.

\$\endgroup\$
  • \$\begingroup\$ For what it's worth, the 74HC4538 dual monostables are much better than the 74123s. \$\endgroup\$ – WhatRoughBeast Mar 31 '16 at 23:53
1
\$\begingroup\$

This sounds like a job for a microcontroller with one capture input and five PWM outputs. There are many micros with these features.

Use the capture input know when the function generator fires and to measure its period after at least two cycles. The firmware would then adjust the PWM generators accordingly. Many of the PWM modules intended for motor control can be configured for a certain delay, or phase offset, from a master clock. I've used exactly this feature in multi-phase switching power supplies and other similar applications.

\$\endgroup\$
  • \$\begingroup\$ Thank you, that seems like what I might need. Would you have a suggestion for a low-cost MCU that can do this? I'm looking at signal capabilities levels of 15V. Also, any analog alternatives too would be great - this goes a test board that might get manufactured in the 50-100 qty range so having the devices programmed individually would be an option only if we don't have other good alternatives. \$\endgroup\$ – Manu Mar 31 '16 at 20:30
  • \$\begingroup\$ @Manu: Take a look at the Microchip dsPIC 33 and PIC 24, especially those that claim to be targeted for motor control or power conversion. Those then to have the fancy PWM modules that can do phase offset and various other things. \$\endgroup\$ – Olin Lathrop Apr 1 '16 at 11:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.