ADS1293: Voltage on 1298 or 1293

Hello Abe,

Your ticket has been viewed. Please give us 48 hours for a response. 

Kind regards,
Nick Z

Hello Abe, 

RLD uses feedback from the body to drive the body toward common voltage. Lead-off detection however, does not. LOD will work without RLD, but the signal that's being input is only measured to see if it's inside or outside of the bounds to be considered "on the body"; this will not ensure that the body is at a stable and desirable common mode for the ADC to make accurate measurements. 

Kind regards,
Nick Z

Thank you Nick,

RLDREF is a dc generated by the device that puts the body at this dc value whereas the LOD circuit puts avdd/2 on the leads during normal operation. The body is now under RLDREF and the Avdd/2. Do you see a conflict here? The LOD is always on and so is the RLDREF on the body. 

Hi Abe, 

I understand your confusion. If both the LOD and RLD achieve or are near common mode, there isn't any conflict. Since the LOD will use the RLD for a return path, it'll be used in feedback in the RLD and help drive the body using the LOD's influence. 

Here's a couple resources that'll help: 
/cfs-file/__key/communityserver-discussions-components-files/73/Improving-Common_2D00_Mode-Rejection-Using-the-Right_2D00_Leg-Driver-Amplifier.pdf 
https://www.ti.com/lit/an/sbaa196a/sbaa196a.pdf 

Kind regards, 
Nick Z

Dear Nick,

There is no confusion and thank for shedding some light! the problem here is that using the LOD is an optional feature whilst RLD is not optional! So to rely on the LOD DC voltage to set the body's DC for common mode is questionable!  Referring to Figure 2 of application report sbaa196a, and normally when all is running well with all leads are attached to the body firmly, it would seem that the source AVDD will divide equally so the upper lead would have a voltage of (AvDD+AVSS)/2 while the lower lead closes the loop through the current source to AVSS. This is a logical Kirkoff voltage law. When the upper electrode comes off, becomes unattached, the upper lead  immediately hits the rail and becomes AVDD and hence the LOD! It is not understood so readily how would you rely on the LOD and its said voltage to set the body's common mode in lieu of RLDREF! There is no where in the document an indication that states that RLD reference voltage should be related to  the LOD rails etc. The question is, if all is normally working and both RLDREF and the leads are under normal conditions, how could you explain the potential  difference  of the DC voltage applied to the body via  RLDREF and the DC voltage from the LOD voltage. Also, the RLD return path was intended for the AC mains be it 60 or 50Hz and not for the LOD DC current. Please clarify and you guys are always great!

You kindly indicated the following

"If a patient-drive electrode is used by the system, the ADS129x includes an on-chip right leg drive (RLD) amplifier that connects to the patient drive electrode. The RLD amplifier function is to bias the patient to maintain the other electrode common-mode voltages within the valid range. When powered on, the amplifier uses either the analog midsupply voltage, or the voltage present at the RLDREF pin, as a reference input to stabilize the output near that voltage"

According to figure 2 of the document sbaa196a; Is it wrong to assume that  the body,  from figure 2,  is under the influence of two DC voltages? One from the RLDREF and the other from the DC from LOD. Are you saying that it is OK for these voltages to be acting on the body and 

to be different since they are separated by z impedances of the electrodes? Remember, LOD circuit puts the lead and consequently the body through the electrode on its DC voltage.  Please refer to figure 2 of the document sbaa196a. IN other words, there is a voltage due to the LOD circuit and another from RLD reference both are applied to the body albeit through these Z impedances. Are you saying these voltages are independent and the Z impedances protect against any conflict. 

Also, how is that seeming dichotomy in putting the body  under two different DC voltages may affect the biasing of the CMOS. Remember that DC biasing currents must flow back to ground. And if  the biasing is internal, does that impedance to ground affect, i.e., reduce the differential impedance of the amplifier since essentially the impedance to ground from the DC biasing circuit would take away from the preferred high differential impedance. Thanks you!  

Do you have an evaluation board/kit EVM?

Figure 2 is from legacy that is to show a rough concept and may not be regarded/used as actual implementation; the actual implementation could be more complex than it.

Main purpose of the RLD is to drive a voltage to human body so that 1. stabilize body potential so body won't be completely floating. 2. possibly to cancel some noises.

Voltage induced by the small(nA) DC current source is only localized to the area between the two electrodes that source and sink the current on top of the voltage that is driven by the RLD if RLD is used.

Z impedance shown in the figure denote the impedance for cable and electrodes.

The concept of GND is a reference which you want to use to compare the potential. In the figure, it uses the reference signal/voltage, i.e. relative to the circuit GND, if RLD is used, to attach to human body to make the body closer to the reference voltage, which is not necessary have to be GND, so RLD somehow helps the sensitivity and specificity of the electrode off detection with less false detection.

If one of the RA or LA electrode is off, the electrode off detection circuit/comparator senses voltage very close to AVDD or AVSS and compare with a pre-select threshold to send a electrode off trigger signal.

This article may provide you some info -

www.planetanalog.com/.../

It　ｃａｎ　ｂｅ　ｕｎｄｅｒｓｔｏｏｄ　ｗｈｅｎ　 the RLDREF is different from the voltage generated by LOD then this conflict is mediated by the Z impedances. Does the DC voltage on the body interfere with the amplifier DC　biasing voltage?　

Ｉ　ａｍ　ｓｏｒｒｙ　ｍｙ　ｋｅｙｂｏａｒｄ　ｉｓ　ａｃｔｉｎｇ　ｕｐ．　Ｔｈｅ　ＫＶＬ　ｓｔａｔｅｓ　ｔｈａｔ　ｔｈｅｒｅ　ｉｓ　ａ　ｖｏｌｔａｇｅ　ｄｒｏｐ　　ｆｒｏｍ　ＡＶＤＤ　ｏｎ　ｕｐｐｅｒ　ｌｅａｄ　ａｎｄ　ａｎｏｔｈｅｒ　ｏｎ　ｔｈｅ　ｌｏｗｅｒ　ｌｅａｄ　ｔｏ　ＡＶＳＳ．　Ｉ　ｂｅｌｉｅｖｅ　ｉｆ　ａｌｌ　ｉｓ　ｇｏｉｎｇ　ｗｅｌｌ　ｉｎ　ｎｏｒｍａｌ　ｏｐｅｒａｔｉｏｎ　ｔｈｅ　ｕｐｐｅｒ　ｌｅａｄ　ｉｓ　ａｔ　ＡＶＤＤ／２　ｔｈｅｎ　ｏｎ　ｔｏ　ｔｈｅ　ｎｅｘｔ　ｅｌｅｃｔｒｏｄｅ　ｔｈｅｎ　ｔｏ　ｔｈｅ　ｌｏｗｅｒ　ｌｅａｄ　　ｗｈｉｃｈ　ｉｓ　ｓｔｉｌｌ　ａｔ　ｓａｍｅ　ＡＶＤＤ／２　ｔｈｅｎ　ａｃｒｏｓｓ　ｔｈｅ　ｌｏｗｅｒ　ｃｕｒｒｅｎｔ　ｓｏｕｒｃｅ　ｗｅ　ｌｏｓｅ　ａｎｏｔｈｅｒ　ＡＶＤＤ／２　ｏｎ　ｔｏ　ＡＶＳＳ．　Ｔｈａｔ　ｉｓ　ｌｏｇｉｃａｌ　ＫＶＬ．　　That　is　KVL　whｉｃｈ　ａｎｄ　ｌｏｇｉｃａｌｌｙ　ｐｕｔｓ　ｔｈｅ　ｂｏｄｙ　ａｔ　ＡＶＤＤ／２．　Ｔｈａｔ　ｂｅｇｓ　ｔｈｅ　ｆｏｌｌｏｗｉｎｇ　ｑｕｅｓｔｉｏｎ；　Ｉｆ　ｔｈｅ　ＬＯＤ　ｐｕｔｓ　ｔｈｅ　ｂｏｄｙ　ａｔ　ＡＶＤＤ／２　ａｓ　ｅｘｐｌａｉｎｅｄ　ａｂｏｖｅ　ｂｙ　ｔｈｅ　ｖｉｒｔｕｅ　ｏｆ　ＫＶＬ，ａｎｄ　ｉｆ　ｔｈｅ　ＲＬＤＲＥＦ　ｉｓ　ａｌｓｏ　ａｐｐｌｙｉｎｇ　ａ　ｖｏｌｔａｇｅ　ｏｆ　ＲＬＤＲＥＦ　ｔｏ　ｒｉｇｈｇｔ　ｌｅｇ，　ｗｅ　ｗｏｕｕｌｄ　ｈｏｐｅ　ｔｈａｔ　ｔｈｅｓｅ　ｔｗｏ　ｖｏｌｔａｇｅｓ（　ｍｅａｎｉｎｇ　ＡＶＤＤ／２　ａｎｄ　ＲＬＤＲＥＦ　ｄｏ　ｎｏｔ　ｒｅｐｒｅｓｅｎｔ　ａ　ｃｏｎｆｌｉｃｔ　　ａｎｄ　ｉｎｄｅｅｄ　ｔｈｅｙ　ｓｈｏｕｌｄ　ｎｏｔ　ｓｉｎｃｅ　ｔｈｅ　ｉｍｐｅｄａｎｃｅｓ　ｏｆ　ｔｈｅ　ｅｌｅｃｔｒｏｄｅｓ　ｉｎ　ｔｈｅ　ｐａｔｈ　ｃａｎ　ｂｒｅａｋ　ｔｈｉｓ　ｐｏｓｓｉｂｌｅ　ｃｏｎｆｉｌｃｔ．Ｔｈｅ　ｃｏｎｆｌｉｃｔ　ｂｅｉｎｇ，　ｉｔ　ｉｓ　ｉｍｐｏｓｓｉｂｌｅ　ｔｏ　ｈａｖｅ　ｔｏ　ｄｉｆｆｅｒｅｎｔ　ｖｏｌｔａｇｅ　ｓｏｕｒｃｅｓ　ｔｏ　ｔｈｅ　ｓａｍｅ　ｃｏｎｄｕｃｔｏｒ（　ｉｆ　ｗｅ　ｃｏｎｓｉｄｅｒ　ｔｈｅ　ｂｏｄｙ　ａｓ　ａ　ｃｏｎｄｕｃｔｏｒ　ａｌｂｅｉｔ　ｎｏｔ　ｔｈｅ　ｂｅｓｔ　ｃｏｎｄｕｃｔｏｒ！）

Which DC voltage on the body are your referring to? Are you referring to the floating voltage on the body before any other voltage apply/drive to the body?

Can you describe or provide more detail what do you mean by interfere?

Do you have an Evaluation board/kit(EVM)?

Can your describe or provide more detail what do you mean by "if all is going well"?

In the conditions that both electrodes ON,  the upper electrode may have a voltage slightly lower than the AVDD; and the lower electrode may have a voltage slightly higher than the AVSS; the exact is unknown as it depends on may other factors and sources, e.g. subject's skin conditions, body impedance, hydration, electrodes, patches, cables, and many more, etc.

LOD does not necessary to put body at AVDD/2.

Body is not a conductor; body has very high impedance and body impedance depending on many factors and sources and the way to measure it, which belong to another field of study/discussion named bioimpedance.

*if all is normally working well*  means that no lead has come off and therefore  the LOD circuit on one lead is putting the body at say close to avdd/2 while the the RLDREF is also connected putting the body at its DC, so with that said, don't you see a possible conflict between two DC  sources namely LOD and RLDREF acting upon the very body?

Also. don't you see the resulting effective DC voltage, from the two sources above,  on the body could possibly mess up the DC biasing of the high impedance CMOS sensing front end? Thank you guys and gals....you are awesome!

May I ask what are the conflicts?

And, What do you mean by mess up the DC biasing? There are many unknow high impedance between all the voltage nodes and across the different parts of the body.

So you agree that RLDREF will act on the body and so will LOD voltage but no conflict should be anticipated since they are separated by bioimpedance? It would seem that the ecg front end more likely will see the dc voltage from the LOD....and the dc from LOD will not disrupt the dc biasing circuit of cmos.

What do you mean by conflict?

You have two DC sources applied to the body namely RLDREF and the LOD!  That is what I mean by conflict.

Unless you are relying on the bio-impedance and/or the Z from the electrodes, as in figure 2, to break that otherwise conflict, Thanks! 

I see. I probably won't call it conflict.

Again, the figure 2 is from legacy and a very over simplified model that omits a lot of real and complex info as dig into/reveal the theory&concepts of bioimpedance might be beyond the scope of the document, so it might cause some misleading.

Human body's bioimpedance is very dynamic and varies from places to places, locations to location and depends on the frequency and many other factors.

Don't think/model human body as a simple pure resistor, think human body as a very complex RC or RRC network that contains many different RC or RRC cells/elements.

In this case, you can apply different voltage at different locations/places/nodes on human body as long as the DC current is limited to very tiny or AC also with very limited AC current.

Suggest read bioimpedance, bioimpedance measurement.

So what is the expected DC voltages seen by the device AFE? And would the resulting DC value from RLD and LOD disrupt the CMOS DC biasing circuit?

what is the expected DC voltages seen by the device AFE? 
Comment: at what conditions and situation?

would the resulting DC value from RLD and LOD disrupt the CMOS DC biasing circuit?

Comments: What do you mean by disrupt? and Why do you think there might be disrupt?

what is the expected DC voltages seen by the device AFE? 
Comment: at what conditions and situation?

under normal operating conditions where the LOD and the RLDREF are both applying their respective voltages! I understand that the body has

complex bio-impedance and you have two DC sources acting on the body namely, the RLDREF from the driven right leg electrode and another from the LOD.  Remember, you have LOD current sources acting on the leads and these current sources have large impedance with respect to the body bio impedance ( 100 times more) so therefore you would expect that you have pretty close to (AVDD+AVSS)/2 on one lead and approximately the same value on the lower lead (assuming two leads system with another RL lead totaling to three leads all together).

would the resulting DC value from RLD and LOD disrupt the CMOS DC biasing circuit?

Comments: What do you mean by disrupt? and Why do you think there might be disrupt?

Remember with BJT amplifiers, we typically isolate the DC from the source (body) by a capacitor before the transistor is DC biased by voltage divider to set the DC bias level. Now you have two DC coupled sources from the ( LOD and RLDREF) so are you saying that these combined will not interfere with the DC biasing of the transistors in the AFE of the ADC amplifiers. 

Thank you! You are the best! 

If ECG electrodes and RLD electrodes are attached to body(note: these won't all attach to the same spot/location/node on body and in fact they are far apart(thus, different nodes) and if DC LeadOff current method is selected and RLD is also enabled, hopefully and conceptually the body voltage will be between AVDD and AVSS with respect to the device's GND.

Don't quite understand your BJT example, if a picture can be provided, that will be great.

Body itself is huge complex impedance load(non-uniform, inhomogeneous, composed of many small complex impedance elements/cells), so when you apply any voltage or current source on any spot of the body and sink(as long as not exceed the IEC and current into body limit) in any other different spot of the body, there are huge complex impedance load(non-uniform, inhomogeneous, composed of many small complex impedance elements/cells) in between.

Hi,

This ticket/thread has been pending for more than 1 week.

For house cleaning, I am closing this ticket/thread.
If you may have other question, feel free to open another ticket/thread.

Thanks

Thank you for your patience and your depth of knowledge. We are almost done here so I am thankful in advance for the invaluable didactical journey. 

Imagine two leads system with a third RL reference RLDREF at say 1.5 volts. This puts the body at 1.5 volts DC. The two leads are also attached to current sources from rail to rail( from the LOD). This means, the upper lead is connected through a current source to AVDD whereas the second lead is connected  through a second current source to AVSS. Such arrangement puts the upper lead at (AVDD+AVSS)/2 through the upper current source and since the bio impedance from the two electrodes combined and  the body between the electrodes is a small fraction of the current sources impedances, it is viable to approximate that the resulting voltage division across the bio impedance remains a small portion of the (AVDD+AVSS)/2. We further assert that the lower lead is at just about (AVDD+AVSS)/2 as well. 

Now imagine the upper lead becomes detached from the body thereby is pulled up to the rail AVDD. The second lead remains connected to the body so we now have the body at RL electrode with 1.5 volts on the body and acting on the lower lead and its current source instead of approximate (AVDD+AVSS)/2 that was on that lead prior to the top lead becoming detached. Is the forgoing a logical inference?

lastly, I surmise that irrespective of the DC common mode on the body, and owing to the DC coupled transistors of the amplifiers, there is no worries that the common mode will affect the   DC biasing of the front end?

Thanks.   

Quote "This puts the body at 1.5 volts DC."

Comments: this puts the body, specifically the RLD electrode to skin contact node to have a path to 1.5V DC as there are impedance in between the path.

Quote " Such arrangement puts the upper lead at (AVDD+AVSS)/2 through the upper current source"

Comments: in practical/reality, it's not necessary to be exact at (AVDD+AVSS)/2, but we may say it's between AVDD and AVSS.

Quote " and since the bio impedance from the two electrodes combined and  the body between the electrodes is a small fraction of the current sources impedances, it is viable to approximate that the resulting voltage division across the bio impedance remains a small portion of the (AVDD+AVSS)/2. We further assert that the lower lead is at just about (AVDD+AVSS)/2 as well."

Comments: Not sure about this statement and what does it mean by a small portion of the (AVDD+AVSS)/2? We can say the body has path in series with some impedance to (AVDD+AVSS)/2.

Quote "the upper lead becomes detached from the body thereby is pulled up to the rail AVDD."

Comments: when the upper electrode is detached/off, the body is not pulled up to the rail. It is the electrode as an open circuit node that becomes very close to AVDD rail.

Quote "The second lead remains connected to the body so we now have the body at RL electrode with 1.5 volts on the body and acting on the lower lead and its current source instead of approximate (AVDD+AVSS)/2 that was on that lead prior to the top lead becoming detached."

Comments: Again, it's hard to say the body is at exact 1.5V or (AVDD+AVSS)/2. We can only say that the body, particularly/specifically the body contact to the RL electrode has path in series with some impedance to 1.5V or (AVDD+AVSS)/2. I am not sure what you mean by acting on the lower lead? But, current will tend to go/flow to a lower potential, in this case AVSS if AVSS < 1.5V or AVSS<(AVDD+AVSS)/2.

Quote "there is no worries that the common mode will affect the   DC biasing of the front end?"

Comments: what might be the worry? And, which front end are you referring to? Common mode usually exists but, depends on how the designers/engineers want to deal with it.

Thanks. 

Can you provide the source code for the ads1293 for 12 leads. And if so, can we migrate the code from the existing mcu code to cc2640 to be the master for all three ads1293? Thank you!