MAX3232E: Charge Pump Capacitor Rated Voltage

More than about 6.6 V should not be needed.

Are you using electrolytic or ceramic capacitors?

Please show an oscilloscope trace of the RS-232 output of a failing board.

Hi Rohit,

Generally we suggest about 16V rated capacitors - as that should give a bit over 2x the expected voltage across capacitor - this helps limit derating of ceramic capacitors so they are closer to nominal value that we do suggest. 

Why do you think that the capacitor is the issue? Are the capacitors breaking (shorting out)  on the charge pump path? 

Do you have a scope shot of the communication bus - primarily the RS-232 driver outputs - it tends to be a bit easier to diagnose these issues at the driver outputs and the capacitor with the 16V ratings shouldn't be problematic. 

Also how is the driver being loaded - i.e.:

1) how long is the cabling between driver and receiver? (if you have a cabling part number that would be helpful)

2)  what data rate are you using?

3) What is on the other end of the RS-232 driver - is it just an RS-232 receiver or are there other components on the bus? 

Please let me know!

Best,

Parker Dodson

Hi Clemens and Parker,

1. We are using ceramic capacitors.
2. We were suspecting capacitors, since after changing capacitors in one of the boards, we did not observe any junk issue. We did not notice any breaking of capacitors or any dry-soldering of capacitors so far.
3. The cable length between the board and PC is ~1.5meters.
4. We are using a Baud rate of 115200.
5. There is an RS232 transceiver to USB serial port adapter which is being used to connect to PC. In DUT, there are no other components in path between processor and MAX3232E.

Below are the waveforms captured in the board where junk is seen on the serial port.

PS: We tried changing the cable and PC ports still the same issue is seen. With a Golden card and same setup, no issues are observed.

In these waveforms, the bus is idle (and looks OK). Can you show actual data transmission?

Hi Rohit,

Thanks for the additional information. 

Is it possible to see the bus communicating ?  - sending an alternating bit stream or just some random data so I can see the transitions on the bus. If there are junk issues due to the electrical performance of the IC it will most likely show at communication - but at the idle state shown currently it looks okay. 

Best,

Parker Dodson

Hi Clemens and Parker,
Please find the attached waveform during the actual transmission. Also have attached serial port logs FYR.

▒C▒+▒▒|ogZ[▒S▒C▒Dmgq▒|u l맹▒0▒T▒+▒▒|ooin▒▒umXk▒,\s▒h▒>▒l▒a▒_m▒뵹▒in▒[镭▒5e▒ ▒!y▒corvYmWX▒uٍu▒t▒looinz0YJ▒EV]_▒A\H?▒pWD▒e▒▒▒K▒▒▒ynY▒ .
C#t▒v▒▒|ooZK▒▒▒▒▒▒񅱽▒▒▒▒B▒oin▒▒ 9▒▒▒6]▒WX▒wtop_swB▒▒Ӛ▒▒▒ILqմ▒{7^▒_Ë+/▒▒ő▒ͥ▒N▒▒no}▒d▒▒յ͵▒<▒aw▒빕▒▒bus0n][▒re▒▒k▒f▒{u mnmt0nYC▒y6<j▒:▒ 8
*LYw▒▒|▒oZ[▒▒\▒^▒▒▒▒l맹▒ ▒▒▒▒▒񅱽▒▒▒▒B▒/Lena}mt▒l▒oin▒▒▒]^▒▒Ձl맹▒▒K▒▒ѧ▒▒▒m▒ juw▒▒fK▒i6▒B/)▒▒1?.▒▒jn▒ڞ▒[B▒▒Wz+k▒k▒w.qw~tZ▒L▒~qt▒ {%ls▒▒
*D▒▒|▒oZ[▒▒n▒▒▒U▒+▒▒|▒oin▒▒nZ▒▒k▒w.qSte▒▒▒▒▒m▒▒|▒{{ (.▒▒zn▒ڞ▒m 0▒.qw~uynb▒م3▒eS MZ+k▒kr▒▒q+3▒e▒romu@6▒n▒B+)▒▒1?.▒▒jn▒▒▒m (▒enge(1063
c#t▒+▒▒|ooinz0▒?[ڞ▒m 0▒▒E▒45▒▒▒%)k▒{▒0(uytY▒j▒sl▒▒▒1;S▒▒ӵ▒▒▒▒▒t▒39\▒u 5\▒`+▒`looik▒l▒bl▒kk{{ (3.▒1(▒B/\▒UM   ▒▒▒▒▒*K*▒h▒{e 0>▒▒▒p: ▒
Lena}mt▒looinz0▒▒▒▒▒񅱽▒▒▒▒▒▒▒ЧH▒/.▒▒▒K▒▒wte▒▒▒▒510▒s▒a5\▒▒▒▒▒>l▒ynZ▒▒▒▒͕▒▒`g▒_▒j*B}an(qor|▒l▒ynZ▒ali{et▒ g▒▒▒j*Fman▒por|▒l▒ m▒5▒▒▒˥▒▒5
*Lena}m|!looin▒▒▒|ooinz0▒oЂu▒t>▒j▒hx▒▒\▒al▒n▒ oez▒e▒`fsom F▒T` (▒ L▒t▒ Wte▒▒0"0x▒0e▒Y▒▒5!▒ (▒` 02▒▒�▒>▒06q0020g▒*▒▒02▒▒▒▒▒�▒▒▒▒▒▒00▒
C#t▒v▒▒|ooinz0▒▒▒񅱽▒▒▒▒B▒YoagYC!▒0▒`G▒+▒▒▒▒▒▒▒▒▒ɗ▒͵▒5▒ wi~e 2u>▒R▒t▒s▒ e▒ٮ+ۥ Z▒齵▒▒M[\VV▒0▒`▒ (▒`O*▒▒"gzmp0opve{smdA5▒`[▒wn▒|N▒aw▒▒▒s
c▒\▒^▒▒Ձlooin▒▒ M▒W▒k▒▒▒▒.▒){-q Root G▒B▒▒7^ªѡR▒t▒{▒=▒10..j▒5* ▒` (▒rghmte,W▒▒;▒AIrg▒▒5Z▒▒▒▒OiB▒▒▒| 6<▒"▒b▒▒▒`  ▒r}in:N▒▒▒ͱ▒`z▒m▒y e
c▒\▒^▒▒Ձl맹▒▒c#t▒+▒▒|ooZK▒▒J▒s:!▒▒▒▒aling vam▒ysk(▒z▒m▒NIT(I}▒w▒ e▒▒04▒�▒.שH▒ue▒u▒▒ϡ/▒pe6<▒ gon▒^V▒w񥽹5▒]▒▒▒▒▒5▒1.1-ئ▒▒ݥ~g▒pZCa▒▒▒c l
c#T▒V▒▒|ooin▒▒▒▒▒▒▒▒|▒oZ[▒▒▒▒rwio▒▒▒▒.t.y▒I(▒n▒|▒~▒`I▒]AC9▒.▒▒L▒t▒c|▒t"[▒▒▒▒▒\▒`wZW▒▒▒▒▒sm▒ٓ▒▒▒2▒▒▒`e▒X▒▒▒▒`6u OZ▒GMA▒reg▒▒▒FM▒Po▒tu8
c▒\▒V▒|u l맹▒▒X▒Ufa}▒t▒looinz0C▒\▒^▒▒▒ logZKC▒▒`0< <0=▒`tc▒_lk.W▒▒▒ѕ▒dz_he▒h xe▒h`t▒嵕stc{}▒i|s=46!}ey_orte▒=6▒`bu▒ke▒_o▒W0▒]Y▒▒]▒ 7
c#t▒k▒▒|ogZk▒1U▒+▒񅱽▒▒▒▒0inz0▒▒▒▒▒▒|▒oinz0gkˆ▒jtn▒ im▒we▒a{(▒~m▒rao▒.▒ɩh▒▒▒▒in▒`in▒tv▒ }e{▒▒: 16▒7:K-MA:`NO▒E_D▒Ta`[oe}p{▒i:!▒▒Kin▒ ▒
▒U▒+▒񅱽▒▒▒▒▒inz0c▒\▒V▒|u logZ[▒▒▒▒▒▒▒|▒oZ[▒▒U▒+▒▒|ooin▒▒>▒rmgiw▒ur▒t n▒w m▒ternacu0lr▒ve▒ŵ▒_wW▒5Sl`TpEA U|hY.▒t`tv▒ver▒sl_dpi!soav▒ d
▒]V▒▒Ձl▒v▒[B▒cC!֖W▒▒|ooinz0[▒acu lri~▒r`slc_n▒} ]                                                u.▒▒O▒▒ re▒iw▒uv▒t n▒w i▒te▒nc▒bɥ▒ 0
c#t▒v▒▒|ooinz0z0▒񅱽▒▒▒▒▒ ▒00▒▒▒▒▒񅱽▒▒▒▒▒▒▒▒▒񅱽▒▒▒▒B▒▒ethe▒~e▒▒ɥve▒G▒usbk▒rm:Ю▒^▒▒er▒t n▒w m▒tervc▒b٥ve▒▒▒▒55▒1gt▒4 n▒ilgd▒th e▒ro     ▒
*LY+▒▒|▒oZk▒z.׶ձ`2.▒ /G▒cul'▒ho▒t G▒~|roll▒r`(EI▒i-▒Dz▒veruxg▒=pg▒:`EHCRA%▒▒▒ӹ▒rm▒dz▒ve▒e▒ci-▒|auvos}>▒UXCI`o▒~erig▒p|ctvo▒_▒e▒bɥ▒ r
▒Y^▒▒Ձlooin▒▒c▒\▒^▒▒Ձlooinz0▒▒▒▒▒▒|ooin▒▒smsc95▒u▒bso▒]▒▒▒▒▒▒ѕɕ▒▒▒uw▒Z▒uv▒acgЗWϕ́▒▒▒▒080w▒co▒e:▒rmgiste▒]Kr▒݁in▒uv▒akg`tr▒v!rmgiw▒we M
LenX_{▒|ooin▒▒c#T▒^▒▒▒ looZ[▒~▒re▒dz▒ver`SoruC▒Z▒5▒w.▒▒O▒▒Ю▒Vk▒er▒t n▒w m▒te▒ve▒u dv▒ver`smss7uxx                                  0
*LenX_{▒n▒oin▒▒▒▒▒񅱽▒▒▒▒B▒S▒1U▒+▒▒|▒oinz076d"▒ci ~▒rwn▒0x5▒p`q}▒ro▒`0x000000000:11085▒                                              ▒
▒]^▒▒▒ l▒oZk▒▒1U▒+▒▒|ooin▒▒▒ugaw|u l▒v▒[B▒▒+▒񅱽▒▒▒▒▒▒`ru▒`n▒}j▒r`1                          ▒▒=xc▒`xxg▒=xg▒>0.▒u|o:`yr▒`5=▒`io▒}e▒ 00
D▒▒|▒oZ[▒oinz0cG!T▒V▒|u l맹▒B▒c▒\▒V▒▒▒▒l▒v▒[B▒0essm▒~g▒▒̓~u{▒▒▒jxxg▒=xg▒`xxg▒=xc▒>0.▒u|o:`Host w▒ppo▒ts▒Uj(
*LYv▒񅱽▒▒▒▒▒▒              xu▒ 1-▒:5▒p:"l▒▒t tg▒]lW▒h▒xxg▒=xc▒▒፥▒xg▒>0.▒u|o:`xXGR!▒▒сG▒~|roll▒r

The RIN2 waveform looks correct.

The DOUT2 waveform never goes positive. And the slow rising edge looks as if the output never drives a positive signal.

Please check the bus voltages at pins 2 and 6, and that the flying capacitors at pins 1/3 and pins 4/5 are being switched. Some solder connection might be broken.

Hi Rohit,

Thank you for the additional scope shots - there definitely seems to be an issue on proper signal levels on the DOUT2 pin as shown. 

It does seem that the DOUT pin is not swinging up to its positive level and as Clemens mentioned the rise time is slow which generally indicates its not actively driving the positive output. 

Beyond the checks that Clemens suggested (which I also suggest you should check to ensure that there are connected properly and check if the flying capacitors are switching and the V+ and V- pins are outputting the correct voltage levels) 

I have a couple more checks on the system:

1. Could you please let me know the part marking on top of the MAX3232E? I want to trace it in our internal system and see if there has been able problems reported in the past + check its legitimacy status (sometimes counterfeit devices can show up and we can look into if the device is valid or not) 

2. Can you show the output voltage of DOUT2 when you disconnect the RS-232 serial cable - essentially an unloaded driver test - this will help me determine if the problem is showing up regardless of loading (which will help isolate the issue directly to the IC and not the IC + system setup). I think I found the correct serial converter but I can't find too much information on the device besides that it converts from 232 to USB - so I want to check if the loading from that device is causing any issues with this test. 

And just as a question - has any other component (diodes, resistors, etc...) been added to the RS-232 bus besides the cable and transceivers? 

Best,

Parker Dodson

Hi Clemens and Parker,
I shall capture the required data and share you my observations.

Thanks & Regards,
Rohit Dixit

Hi Rohit,

Thanks - just let me know when you have the scope shots and I will take a look!

Best,

Parker Dodson

Hi Parker,
I have summarized responses to all of your above queries in the attached document. Please have a look.
MAX3232E Debug.docx

After touch up, the DOUT voltage levels seem fine and I am not observing an junk on the serial console. However, I have some cards where I have performed A-B-A swap analysis and issue still persists. We have swapped the failing part to a good board and observed that the failure is still seen. The part replaced from the good board to the failing board works fine as expected.

Hi Rohit,

So the results look in line with what we more or less would expect. 

The part marking looks legitimate - so I am not too concerned with a counterfeit device. 

The diode used on the bus really isn't giving much protection - its clamping voltages are much higher than the driver and possibly the receiver as well - so if a transient occurred on the bus that was clamped above acceptable levels you could see damage - but the tests you show don't necessarily indicate that. 

The driver seems to output voltages okay with no loading - what exactly is connected to the serial bus when you see junk data - because that seems like it could be the source of the problem. 

Best,

Parker Dodson