• State Resolved
  • Replies 3 replies
  • Subscribers 47 subscribers
  • Views 50 views
  • Users 0 members are here
Support feedback
Options
Options
Related

THS3470REBEVM: Power Down Mode

Günter Hildebrandt
Prodigy 20 points
Part Number: THS3470REBEVM
Other Parts Discussed in Thread: THS3470

Tool/software:

Hi Support Team,

The documentation of THS3470 describes a power down mode controlled by P0 and P1 pin (No. 30 and 31). If both pins are connected to ground the power down mode is entered. The evaluation board can connect both pins to ground by switch s3. 

I monitor the output signal of the THS3470 with an oscilloscope. If I switch to power down mode, the signal form changes but the signal is still on the output. I would expect that the signal is turned off in power down mode. Will this behavior change in the next IC revision? I have Rev.B.

The customer of our planed product may overload the output of THS3470. E.g. by connect a long cable or a wrong load. Will the HTS3470 be protected against overload at +-30V supply voltage when OVTEMP_FLAG (1/2) is connected to P0/P1 power control? 

Thank You

Guenter 

  • 0
    TI__Genius 16245 points

    Hi Guenter,

    The voltage that can be seen at the output of the device when in power down can vary depending on the configuration. Are you seeing a DC offset at the output? Depending on the configuration as well as where the load is biased to, you could see a voltage due to the feedback network still being connected to the circuit and biasing the output node through there. Regarding your second question, the device itself does have some protections such as an internal current limit feature that helps limit any current conditions as well as a thermal protection feature. Could you elaborate on the kind of overload conditions you are concerned with? Is it a shorting condition that would cause the device to drive a large amount of current or is it a high voltage condition that exceeds what the device can handle? 

    Best Regards,

    Ignacio

  • 0 in reply to Ignacio Vazquez Lam1
    Prodigy 20 points

    Hi Ignacio,

    I use the evaluation kit with a modified feedback resistor of 750 Ohm. I apply a 10 MHz signal to the input and get an Output of 35 V peak-peak. The load is 120 cm long cable with 50 Ohm impedance connected to a high impedance oscilloscope input. The evaluation board uses about 117 mA at 30V for both power channels in this setup. When I switch to power down the current is reduced to about 106 mA on both power channels. The following pictures shows the output signal of the eval board. I'm surprised about the level of the output and the remaining current consumption in power down mode.

     

    I think the customer will not make a short circuit on the output. I have seen during my bandwidth measurements that a cable at the output causes a high power consumption at some frequencies. I think this is caused by a kind of resonance, when the wave is reflected on the open end of the cable. The open end is the use case for our application and this cannot be changed. I have already destroyed a THS3470 while the over-temerature output was connected to the power down input by sweeping the frequency at 50 V peak-peak output. The customer will send a digital generated signal to the amplifier and I have no control about the wave form. The load depends on the cable length and the wave form. I'm looking for the required protection mechanism to keep the IC at safe operating conditions. Do we need an external circuit to monitor the DIE temperature?

    Thank you,
    Guenter

  • +1 in reply to Günter Hildebrandt
    TI__Genius 16245 points

    Hi Guenter,

    The device when in power down should result in a high impedance output and should not amplify any signal or pull this much current. Is the circuit in an inverting configuration? Also, what is the gain and input voltage Vpp? If you remove the load all together, do you see this still? I would recommend increasing the Rf resistor to 2k and adjusting Rg to the gain you need. This will limit shoot-through current and should help with thermal performance.

    Regarding the protection circuitry, the device does have a natural delay which means for moments with large transients there is a chance the thermal protection will not kick in on time to save itself. I am curious if the failure you have seen is due to reflections or due to thermal issues causing the device to fail. However, the device's thermal protection mechanism is the temp flag to the power down pins as a way of protecting itself. Current limiting the device is another consideration for protecting the device. You could use an external circuit to monitor the die temperature, but this circuitry is relatively slow, so I believe one of your best options is the temp flag to your control pins.

    Best Regards,

    Ignacio