Post Go back to editing

LTC6602 input connection

Category: Hardware
Product Number: LTC6602

Hi,

I would like to use LTC6602 as filter for IQ demodulator. My modulator is MAX2021, which has differential output impedance of 53ohms. My signal has maximum frequency 640kHz and I'm would use ADC sampling rate of 2MHz - 4MHz. I would need DC block for LTC6602.

I'm planning to use circuit from this thread:  LTC6602 input and output

Instead of DC coupled single ended input, I need to modify input for AC and differential. I would use LTC6604-2.5 for output.

My question is what is optimal input circuit of LTC6602 as my source is low impedance? According to datasheet Figure 12b, it is recommended to have 1.87k resistors for bias. If I connect MAX2021 directly before 0.1uF in Figure 12b, I would get double to voltage. Would that be recommended way to use LTC6602 for low impedance AC coupled loads?

Parents Reply
  • Hi,

    I'm still waiting for response.

    Meanwhile, I'm waiting for response. I was building the evaluation board as DC1304A-A is not available. I have problem that there is fairly significant oscillation seen in output. It seems to have frequency around 1/6 of the clock (oscillation is 4MHz when my clock is 24MHz). My schematic is supposed to be same as evaluation board except I have 100 ohm resistors in output (R2,5,8,10) and 100pf capacitors (C4,7,...). 

Children
  • Hello,

    You can also use the circuit on Fig12a if you have a fixed low pass cutoff frequency. Fig12a sets the common mode voltage to Vsupply/2 which will also help with your input range. 
    In terms of impedance, I scanned thru MAX2021 datasheet and found that the 53ohms is the differential input impedance. If you check on page 14 of the DS, you can actually achieve a high impedance baseband amplifier by directly DC-coupling to baseband ports through an inductor to ground, or through a low-value resistor to ground as shown in Figure4.

    Best Regards,
    Kenneth

  • Hello,

    Thank you for the recommendations. I hope I could have variable lowpass cutoff frequency, so I guess I need follow figure 12b.

    About this ~4MHz oscillation seen in the output signal. It is actually common mode, so it is probably not the problem. Is this expected behavior? I'm using settings GAIN0 = 1, GAIN1=1, LPF0=1, LPF1=0, HPF0=1 and HPF1=0. It seems that changing settings will impact to this common mode oscillation.

  • Hi,

    Can you share with us your actual schematic with input/output connections?

    Thanks,
    Kenneth

  • My circuit should be identical with the evaluation board

  • What does your input look like? is it from MAX2021 already? or you used a different signal generator? What's the exact amplitude and frequency of the input signal?

  • Hi,

    I did the test connection according datasheet:

    It looks ok.

    As output from MAX2021 is going to be 2Vpp I adjusted gain to maximum, LPF and HPF settings as mentioned before. If I terminate both inputs with 50Ohm, I can see this 4MHz common mode noise in outputs. Amplitude is about 30mVpp.This is without MAX2021.

  • Hello,

    Have you tried using the default RC stated on the eval board user guide? I think the default is 49.9ohm and 10pF.
    Unfortunately, LTC6602 have no available model on the web so I can't simulate your circuit.

    Also, can you try increasing the cutoff frequency to 247.05kHz by setting LPF1 and HPF1 to 1 just like the one on the table below? If you are using 100kHz input signal, the LPF might be attenuating/blocking it if you are using a cutoff frequency of 82.35kHz, maybe that's why you are seeing the common mode oscillations.

    Best Regards,
    Kenneth

  • Hello,

    I have tried different things on output. Common mode noise is getting smaller if I increase the capacitance (cutoff is lower).

    Settings has some impact to oscillation seen in output. Default setting gives much better result. I doubt that input signal frequency has any impact, since I see this common mode oscillation without any input signal.

    My question is if it is expected to see some sub harmonics of clock in output? As there is no evaluation board available, I cannot do verification.