ADM3307E stange behavior of driver

Hi NiceJ,

Could you share a layout? 

This should help narrow down the cause of the behavior you're seeing. 

Regards,

Shane

Hi SFlanagan,

Thank for your reply.

I'm performing tests on COM4 and the correct schematic is (the same as previous, but different components name to match below layout) : 

Layout : 

U8

Layout 

Additionally, find here detailed scope screenshot : Green is COM4_TX_CMOS and Blue is COM4_TX_RS232

ZOom where CMOS bit aren't pushed to RS232 line by the driver : 

Thank you

Hi NiceJ,

Thanks for sharing the layout and the plots.

Have you been monitoring the V+ and V- voltages while observing this behaviour?

I'd be interested in seeing those plots if you have them available.

Regards,

Shane

Hi SFlanagan,

I performed more investigation on the Layout and I found that The V+ Cap was placed before VCC decoupling cap (see picture below), after this was corrected the RS232 link is better.

Please have a look at picture below, and tell me if it is what expected for the ADM3307E :

• 0,1µF charge pumps caps
• 230kbps
• YELOW : V+_CHIP1
• BLUE : V-_CHIP1
• RED : RX_CMOS(CHIP_2)
• GREEN : TX RS232(CHIP_1)

Zoomed

Replacing charge pumps caps with 1µF may help smoothing it, rigth ?

Is it possible to share internal block diagram of the chip ?

Thanks for your support

1µF pumps cap is lowering ripple at “light” load, but not so much at “high” load. But signal remain inside valid windows threshold.

Here are captures for V+, V- and TX_CMOS on the same chip, running at 230kbps.

I found something very strange by disconnecting the chip to GND, as shown in following picture : 

In this case, I have a very beautiful TX_RS232 signal !

Can you share the ADM3307E’s internal block driagram with us (espessially pin 23, 5 & 6 internal connection) ? Which could help understanding this behaviour.

Many thanks.

Hello SFlanagan,

any update on this topic ?

Thanks

Hi NiceJ,

Under high load conditions, the oscillator is on continuously if the charge pump output cannot reach 7.25 V, hence you get so many ripples. The ripple on the output voltage of a charge pump depends on three factors: the oscillator frequency (The high frequency minimizes the amount of capacitance required, as less charge needs to be stored and dumped in a shorter cycle), the value of the reservoir capacitor, and the load current. The pulsing nature of the higher voltage switched output is often smoothed by the use of an output capacitor. So by increasing the value of the capacitor there, you're smoothing out the ripples at light load. But seems like your load current is too much during heavy load conditions so increasing the value of the external capacitor isn't helping that much as the oscillator is continuously on and hence the ripples. Also, under heavy load conditions where the output never reaches 7.25 V, the Green Idle circuit is inoperative and the ripple voltage is determined by the load current, the same as in a normal charge pump. So by adding an external capacitor there (which you’ve removed by disconnecting the capacitor to the gnd), you’re increasing the load current only (It is stealing current away and recharging itself). Hope that helps answer your questions.

Naveen