Reverse impulses on a DC output

Discussion in 'The Projects Forum' started by Gestalt, Mar 9, 2012.

  1. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    I want to take a steady DC Voltage of about 15V and have it periodically drop to 0V. Below is the waveform I am looking for.

    [​IMG]

    I also want to be able to adjust the frequency of how often the downward spikes occur. Ultimately I want it to be able to do this with higher DC voltage of about 72V and at 1.5 amps.

    Things I thought of that may work:
    -Simple switch or Relay controlled by a timer or a waveform generator
    -An Oscillator of some kind?
    -A 555 timer
    -High Frequency Switch

    My base electronics knowledge is very rusty, any idea on how I might be able to achieve this? I am looking for precise instructions on how to do this. Right now I just have an adjustable DC power supply.
     
  2. PaulEE

    Member

    Dec 23, 2011
    423
    32
    Do those spikes have to be spikes?

    What is your min/max frequency?

    Are you familiar with microcontrollers? :)
     
  3. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    Yes, ideally I would like them to be spikes. Although I would consider compromising..
    Frequency somewhere between 1Hz - 7kHz.
    I am somewhat familiar with micro-controllers...as long as I don't have to do code programming I should be fine.
     
  4. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    I don't think it makes sense to build something that works for 15V and then transform it to something that works with 72V.

    1. Do you have the 72V/1.5A power supply?
    2. What is the slope of the spikes, i.e. what's the pulse width?
     
  5. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    "1. Do you have the 72V/1.5A power supply?" Yes

    And I do want to be able to work in a large voltage range. Hence a simple 555 will not work. I need something that can handle higher voltages. The application is not digital.

    "2. What is the slope of the spikes, i.e. what's the pulse width?"
    Pulse width is to be as small as possible, mimicking a transient spike.
     
  6. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    It'll work for the timing, but you'll need a transistor to take the output up to whatever you need. Old TVs are full of fast-switching, high-voltage transistors.

    You might consider using a timer to produce a square wave into a transformer. The spikes will be sharp because the transformer only reacts to a CHANGE in current, giving a big spike on the leading and trailing edges of a square wave. If your pulses are far enough apart, you can use a capacitor to slowly drain off the voltage after a trailing edge has passed, so there are only spikes in one direction.

    Here's a pulse generator I used recently, showing the basic idea.

    [​IMG]
     
    Last edited: Apr 2, 2012
  7. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Is this a resistive load you want to drive with the 72V?

    IMO you could simply put a p-channel FET in series with the power supply and switch it off for the desired pulse width.

    You will need to switch this MOSFET with a voltage about 15V lower than the 72V.

    I have nothing to draw here right now.
     
  8. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    Would a frequency generator hooked up to a relay work?
     
  9. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    It will give you an interruption on the 72V, but it will not be a highly precise pulse duration nor will it be a clean pulse like the one in your diagram. relay contacts bounce.
     
  10. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    A spike "as small as possible" does not tell us what is acceptable. You need to give a value so we know what type of circuit you need. Is it milliseconds, microseconds, nanoseconds....? What is your budget? :confused:
     
  11. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    +1
    I still contend that the circuit I suggested earlier - a transformer reacting to a rising or falling clock edge - is a good strategy for a brief, "shaped" and timed pulse. You have lots of controls - for instance the RC time constant and the transformer turns ratio.
     
  12. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    It may very well be an adequate circuit for the task. But unless the OP tells us what is acceptable, we can't know for certain. :rolleyes:
     
  13. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    Pulse off time is to be a constant while in operation and adjustable down the road. I am looking to have the pulse off time between 50 microseconds and 2ms.

    I have a fairly large budget. I just want it to be relatively simple to operate. In order to adjust parameters I would like to ideally use pots or some sort of computer program.

    Also wanted to say load is inductive.
     
  14. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    OK, now I better understand your requirements.

    I had a requirement at work to apply several different types of transients to the input of a military power supply while it was operating at load, so I needed a power amp that would deliver significant DC power plus simultaneously apply the desired transient. After much searching I found this amp from AE Techron which I purchased, and it worked very well for that purpose.

    I drove the amp with an arbitrary waveform function generator which I could program to provide both the required DC and AC transient combined waveform. Any waveform generator that can provide the desired analog waveform will work, for example a computer controlling a D/A converter.

    The amp is not cheap, of course, but it is a well-made industrial type amplifier and provides an output quite faithful to the input, even into loads with inductance and capacitance, such as the EMI filters at the power supply input I was testing. I did a lot of searching and experimenting to try to generate the required waveforms and the Techron amp was the only one I found that would adequately do the job.
     
  15. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    This argues for a LabJack. (I have their U3HV model - love it.) It's a USB data acquisition device with a range of output capabilities (clocks, PWM, analog, digital). You can control it with just about any programming language you might be familiar with. I use mine under control of visual basic and Excel, so that I can control it and log data into the same application. Once you're all set up to use and program your LabJack, creating the custom waveform you want would be fairly trivial.

    You would additionally need a MOSFET driven by the LabJack (or an amp as suggested) in order to scale up the power, since the LabJack is limited to USB power levels.
     
    Gestalt likes this.
  16. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    The U3 HV labJack looks awesome and is exactly the kind of thing I am looking for. However programming is the bane of my existence. Is there a set of code I can copy/paste where I can just change parameters?

    Or perhaps is there another device similar to the LabJack that requires no programming and just has a friendly UI interface?
     
  17. wayneh

    Expert

    Sep 9, 2010
    12,103
    3,038
    I don't know about friendly, but you could look at DATAQ, and a lot of folks here yammer on about the Arduino. The learning curve for anything is going to be steep, but you can get thru it. Your best bet with any of them is probably to find someone to point you to something close to what you want, and maybe willing to tweak it for/with you. The forums over at LabJack were helpful. Not always real speedy, but users contribute a lot of good ideas and the company seems pretty good about contributing as well.

    I should add that one reason I chose a LabJack when I went thru this process, was because of its flexibility in programming. You don't need hardly ANY special software beyond a USB plugin or something like that, and then you can access the device with almost any language, including over networks. For me, that meant it would work well with my Macs, and it does. I wasn't much interested in the devices requiring dedicated software running only on a peecee. I've seen too many such applications fall out of support when the programmer moves on, leaving users stranded after a few years.
     
    Last edited: Apr 7, 2012
    Gestalt likes this.
  18. Gestalt

    Thread Starter New Member

    Mar 9, 2012
    14
    0
    A friend referred me to this Advanced Waveform Generator by syscomp, which has a super friendly UI.

    I e-mailed the company and using it's arbitrary waveform function I should be able to get my desired output. Pretty stoked, and price point is reasonable, imo.
     
  19. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    That sounds good to generate the waveform. Have you decided what type of circuit to actually drive the power line? :confused:
     
Loading...