TXB0106: TXB0106PWR issue with w/f,dive capability

Intellectual 435 points

Part Number: TXB0106
Other Parts Discussed in Thread: TXU0104, TXU0102,

Hi,

I am using TXB0106PWR for converting 3.3v to 5.V logic levels, facing challange in getting clean output

all signals are uni directional o/p of 3.3V microcontroller system with CLK signal is less than 6MHz

please find the attached schematic for the same

level shifter.jpg

let me know possible rootcause & solution for the same , also let me know if any other dropin replacement available for the same

1 month ago

0 Clemens Ladisch 1 month ago

Guru 325980 points

For unidirectional signals, use unidirectional translators like the TXU0104/TXU0102.

Anyway, what exactly is the problem? Please show an oscilloscope trace of the wrong output signal.

0 Tyler Townsend 1 month ago

TI__Guru 63428 points

Hi AG,

Our team will get a response to you by eod tomorrow 7/7.

Regards,

Tyler

0 Jack Guan 1 month ago

TI__Guru 60151 points

Hi AG,

I do not see any immediate concerns with the current implementation, please provide the additional clarification similar to Clemens' ask.

Regards,

Jack

0 AG 1 month ago in reply to Clemens Ladisch

Intellectual 435 points

input signal

output with peripheral load connected with load

lwithout peripheral load

0 Clemens Ladisch 1 month ago in reply to AG

Guru 325980 points

Apparently, the capacitive load on the outputs is too high (see section 7.3.4 of the datasheet). How long are the traces? Are there other components like resistors/capacitors connected anywhere? Are they any connectors or cables?

0 AG 1 month ago in reply to Clemens Ladisch

Intellectual 435 points

these are direct traces to connector with 8mil thickness & ckearence, no capacitor or resistor in between, traces length is 10 to 15cm long,

0 Jack Guan 1 month ago in reply to AG

TI__Guru 60151 points

Hi AG,

Are the signals probed via tip and barrel method or active probes? Is this behavior observed on all channels? Typically we see this behavior if the output is heavily loaded with capacitance, as shown in this app note: www.ti.com/.../scea147.pdf

Regards,

Jack

0 AG 30 days ago in reply to Jack Guan

Intellectual 435 points

Hi jack,

still debugging it, all readings are with passive probes & not with tip barrel method,but primary observations are

1. similar tracks connected to same peripheral such as B1, B2, B3, B4 having almost same length, do not have issue, but the frequency is less about 122Hz and pull up on input side

2. facing issue with CLK & DI1 range 2.6MHz to 6MHz

0 AG 30 days ago in reply to AG

Intellectual 435 points

Does each o/p signals require 22E series resitor?

0 Jack Guan 29 days ago in reply to AG

TI__Guru 60151 points

Hi AG,

Yes, the dampening resistor should be placed as close as possible to the output of the device to maximize the impedance matching technique. How does the SI look at the slower 122Hz compared to 6MHz?

Regards,

Jacl

0 AG 27 days ago in reply to Jack Guan

Intellectual 435 points

Hi jack,

with loawer feq., & i/p pull up , waveform for B2

o/p with load

output with peripheral load connected

o/p without load

output without peripheral load

0 Jack Guan 26 days ago in reply to AG

TI__Guru 60151 points

Hi AG,

Similar to Clemens' feedback, since your signals are fixed direction per channel, a fixed direction translator would be used to benefit the system, such as the TXU0104 and TXU0102. It is also not suggested to drive slow transition edges into the TXB0106 (see section 3 of app note https://www.ti.com/lit/an/scea135/scea135.pdf?ts=1784583579809&ref_url=https%253A%252F%252Fwww.google.com%252F for additional details).

Regards,

Jack

0 AG 26 days ago in reply to Jack Guan

Intellectual 435 points

Hi Jack,

Thanks for reply, actually i followed the info provided in datasheet,

anyway is their any limitations on placing pullup/pull down on i/p or o/p side of channels of TXU0104?

also let me know, if TI have any drip in replacement for TXB0106PWR ? or similar 8 channel chip?

Regards,

0 Jack Guan 25 days ago in reply to AG

TI__Guru 60151 points

Hi AG,

AG said:
anyway is their any limitations on placing pullup/pull down on i/p or o/p side of channels of TXU0104?

The limitations are not as stringent as the TXS/TXB type translators, due to its pure CMOS IO structures. However, as stated in section 7.3.1.1, it is suggested to not use any value larger than 1MΩ.

AG said:
also let me know, if TI have any drip in replacement for TXB0106PWR ? or similar 8 channel chip?

Unfortunately no direct replacement, as the similar 6/8 channel devices involve a DIR pin to control the direction of the signal direction (i.e LXC8T245)

Regards,

Jack

0 AG 18 days ago in reply to Jack Guan

Intellectual 435 points

Hi Jack,

thanks for reply

Does the current drive/drive strengh will remain applicable for both A to B as well as B to A data transfer,?

Regards,

0 Jack Guan 17 days ago in reply to AG

TI__Guru 60151 points

Hi AG,

Yes that is correct, please refer to the IOH/ IOL values in the datasheet for the respective drive strength based on VCC condition. The DIR pin should be pulled high/ low depending on the direction of the signaling required to set A or B side port as output side.

Regards,

Jack

Logic

Logic forum

TXB0106: TXB0106PWR issue with w/f,dive capability