How to use ADL5910 peak/envelope eval board to get peak voltages with peak period 1.67 seconds

hi there,

i have bought an ADL5910 peak/envelope eval board to get peak voltages with peak period 1.67 seconds.

basically, the element i am trying to measure has a peak voltage of around 0.4 to 40 volts which occurs every 1.67 seconds.  All I want to do is get the peak voltage from each pulse and output those values to a microprocessor.  

I have been trying to use the data sheet which describes the 5910 but it doesn't seem to indicate where to measure the output - I tried Q but it seems this is just the flip-flop as it outputs 0 or 1 only.  If i'm running the signal into Rfin_SW, do I measure output peak values at Rfout_SW?  Also, it seems that the board has to be reset manually from outside after each peak is detected - is that right?

Any help much appreciated as, clearly, I am a newbie on this stuff.

Cheers, Nick

  • +1
    •  Analog Employees 
    on Jun 11, 2019 9:04 PM

    Hello boydy33,

    The output voltage at pin 3 (VCAL) is the envelope detector voltage.  It is also corresponding to the RF input power from RFIN pin. 

    The "Evaluation Board" section on data sheet page 19 talks more about the RFIN_SW and RFOUT_SW connectors.   Table 6 starting on page 20  lists out each function of the ADL5910 evaluation board.  

    The Eval. Board has "S1 Switch" for manually reset the flip-flop.

    Hope these can help.

    John_C

  • Hi John, 
    Thanks for getting back to me sport.
    Page 20 says:
    First, measure the output voltage on the TCAL yellow clip lead with no RF signal applied to RFIN (this voltage is approximately 750 mV). 
            - Done - was only a bit above at around 765mV
    Second, apply the RF input power to RFIN, which causes the circuit to trip, and again measure the voltage on the TCAL yellow clip lead. 
    - this is the bit i'm finding a bit confusing.
    - for this, I just connected my input signal which is basically background noise without the voltage spike/pulse every 1.6 seconds (image attached shows signal with the pulse) which i want to detect.  Hence, it's basically just a varying noise signal which goes up to around 250 mV say.  So does this mean that instead of the above, I should connect a setup dc voltage of the threshold i want (say 250mV) and measure the subsequent voltage at Tcal or is this instruction referring to connecting the actual background voltage signal to RfFn?
    - by the way, when i measured Tcal with the RFin connected to my background noise input, I barely registered anything more than the 765mV of step 1 which didn't seem right.
    The difference between these two voltages is equal to the voltage that must be applied to VIN− during operation. This voltage can be applied either to the VNEXT SMA connector or to the VIN− yellow clip lead. This voltage can also be set by the R8 mechanical potentiometer on the board.
    - for this part, i just checked the voltage on Vin and adjusted the pot until the voltage on Vin equaled the difference between the above readings.  Is that right or do you need to supply an external voltage to Vin?
    Regarding the reset of the board, do i need to send a signal to the S1 switchevery 1.6 seconds in order to reset the board to read the next peak.....i certainly can't do this manually but it's hard to believe that this board would not be setup to enable a reset in an automatic or easier way?  If I do need to reset electronically, do you know what signal input it would require and where exactly?  I would be doing this via a nodemcu esp8266 microcontroller.
    Thanks a lot for your help on this man,
    Nick
  • Hi John,

    One other question is, with the peaks detecting properly, I assume the output would be sensed at RFout_SW and if so, is it a digital or analog output?

    Cheers,

    Nick

  • 0
    •  Analog Employees 
    on Jun 14, 2019 11:56 PM in reply to boydy33

    Hello boydy33,

    • Is there a RF carrier frequency for your “…spike/pulse of every 1.67 second in signal...” ? You need an input RF signal from the tenth of kHz to 6GHz.  The ADL5910’s TCAL (or VCAL ) pin has a typical output voltage at about 750mV with no RF Input    You measured 765mV indicating that the input signal is too low or your input signal could not get to the ADL5910 device.
    • You do not need external voltage supplying to the Vin pin.  You can adjust R8 potentiometer for it.
    • When the ADL5910 is latched, the Q output will stay on “High” .  At the same time, the RST pin is pulling low (by R5 resistor).  The ADL5910 will stay at latched condition until the S1 switch is pressed or a voltage that  greater than 2V that supplying to RST pin.
    • The signal at  RFOUT_SW is from the RFIN_SW input signal.  When the ADL5910 is latched, the Q output is  “High”.  The "Q_bar" is "Low" and it will open the switch (HMC550A) which prevent the signal flowing to the RFOut_SW.

    Hope these can help.

    John_C

  • Hi John,

    Thanks a lot for your help on this man - much appreciated - as you can see, I' m a mechanical engineer, not electrical ;-)

    • Is there a RF carrier frequency for your “…spike/pulse of every 1.67 second in signal...” ? You need an input RF signal from the tenth of kHz to 6GHz.  The ADL5910’s TCAL (or VCAL ) pin has a typical output voltage at about 750mV with no RF Input    You measured 765mV indicating that the input signal is too low or your input signal could not get to the ADL5910 device.
      • The input I am measuring is an electric fence, which is energised by a normal electric fence energiser.  This charges an internal capacitor and discharges 2 to 12 kV down the fence every 1.67 seconds in the case of my fence.  Hence, I don’t believe there is any carrier signal between peaks. 
      • My assumption in all this is that this is the point in the calibration procedure where you need to connect your background noise signal to RFin but is it the case that you are actually applying your full signal with voltage pulses at this point  in the procedure?
      • Does your above comment mean that if there is zero voltage applied between voltage pulses, then this board cannot detect the pulses?  Isn’t is possible to trigger reading of the pulse peaks in this case just by setting the Vin to the effective zero level of the ADL5910 board– ie. 750 to 767mV?
      • If I can’t do it with this ADL5910, do you think this type of circuit would be more appropriate? https://web.archive.org/web/20160624...ul/peak_holder
    • You do not need external voltage supplying to the Vin pin.  You can adjust R8 potentiometer for it. 
      • I suspected that was the case…thanks.
    • When the ADL5910 is latched, the Q output will stay on “High” .  At the same time, the RST pin is pulling low (by R5 resistor).  The ADL5910 will stay at latched condition until the S1 switch is pressed or a voltage that  greater than 2V that supplying to RST pin.
      • Right, so I’d basically have to trigger the reset to S1 by applying a voltage of > 2V to the RST pin.  That shouldn’t be difficult to program and supply from my microcontroller.  Thanks.
    • The signal at  RFOUT_SW is from the RFIN_SW input signal.  When the ADL5910 is latched, the Q output is  “High”.  The "Q_bar" is "Low" and it will open the switch (HMC550A) which prevent the signal flowing to the RFOut_SW.
      • Sorry John but I’m a bit confused by this.  I still don’t understand where exactly you would read the output, peak, signal and what exactly it would be – ie. Digital or analog signal? So say, the max peak voltage found during a given pulse is 30 volts, where would you measure that and in what form will it be….I assume it would be an analog signal right?

    cheers,

    nick