EEVblog #970 - Mailbag

[EEVblog]

Mailbag Time
WARNING, Contains Nerd Nudity


SPOILERS:
Lab Gruppen C SERIES 88:4 Power Amplifier Module
ECG Electrocardiogram Module.
http://www.signallattice.com/
PAP2 Portable Gaming Console Tearddown
Nikon D300 Pentaprism
$2 Casio FX-82MS calculator ripoff
1980's Princess Handy TV Teardown
Turf, WTF?
https://www.noworriesturf.com.au/
Vaisala RS80 Radiosonde Teardown with some RF magic.
Oscilloscope shootout, Tektronix MDO3000 vs Keysight MSOX-3000 vs Instek GDS-1104B

[bsudbrink]

Oh no!  Not nerd nudity... OTOH better your nerd nudity than my nerd nudity.

[mtdoc]

Dave-

Much of the noise your were picking up with the ECG looked like skeletal muscle cell electrical activity (likely from chest and abdominal wall).

This is not surprising given the relatively large electrical potentials generated by skeletal muscle cells which are closer to the surface than the deeper cardiac muscle cells.

This is  an issue that is inherent with measuring ECGs. Even high end ECG machines costing many thousands of dollars will show this. ECG techs need to have the patient lying down, not talking, and relaxing all muscles as much as possible to get a clean ECG tracing.


Also FWIW - your heart rhythm looked normal.

BTW - Mathew has been nice enough to send me one of these.  The first one had a small issue and I am looking forward to the replacement so I can have a play with it.

[trophosphere]

Thanks Dave for taking the time to review my ECG project. From the great feedback you provided I want to give my thought-process on the following:

Pin headers on the development board

I wanted to make it so someone can also have the option of hooking up jumpers from the breakout board to a breadboard in case they want to feed the signal into more analog circuitry (additional filtering) or into a micro-controller development board such as Arduino.

Pin headers unlabeled on the other side

I understand the confusion this presents and actually spent some time thinking how to label the other side as the title engraving then take up too much space. So far the best I could come up with is to either place shortened labels on the other side or just remove the title engraving completely thus freeing up space to put the full labels.

Protection Resistors

Very good for taking due diligence on making sure the device won't kill or hurt you.  I was going back and forth in my head on whether or not you were going to hook yourself up to it. Besides the 10K resistors on the breakout board, the inputs and the right leg drive also have 100k resistors in series to the outside world as well. Adding resistance to the sensing input and right leg drive output reduces the CMRR of the module and adds offset (due to mismatch between resistors) but for piece of mind it is a good requirement. If defibrillation/synchronized cardioversion is a possibility then it is a definite requirement as an ECG without input resistance (of at least 10k ) can shunt most of the energy away from the heart and thereby making the act of defibrillating a patient useless.

Input Jacks

It has been a bit difficult to source the right jacks for this application as it's a rather niche product. These are indeed standard for certain bioelectrical monitoring applications such as ECGs and EEGs. They are referred to as 1.5mm safety DINs. One source of 90 degree jacks is here but I found this much later after I sent you the board. Like all things medically related, they are pricey as well. The reason for the cutout was I was trying to use the rings and PCB sort of as a brace for when the user would be pushing and pulling the electrode plugs in/out of the jacks. I never thought about what would happen if the user applied stress orthogonal to the intended insertion direction. Good catch. Will have to think about some sort of brace or just use the 90 degree jacks instead.

LED indicator

I was thinking about incorporating that as well when designing the board but decided against it at the time since I wanted to make it easy to measure the current consumption of just the device itself. I think I will add LED indicator regardless and just have a jumper link to disable it if someone just wants to measure the current consumption of the module. This would then let me add a LED ON indicator and LED heart beat indicator using a simple comparator circuit with hysteresis. I could probably add reverse-polarity protection as well.

Nudity

Colored Input Jacks

The colors on the input jacks follow the American Heart Association color coding for Lead 2 configuration. Of course one can easily reconfigure the module to read Lead 1 or Lead 3 by simply relocating the electrodes. I selected lead 2 since it usually provides the best signal for viewing the electrical activity of the heart. I will add something on the back showing how to hook up the breakout board for lead 2 configuration to mitigate uncertainty.

Interference

This is where most of the innovation lies is in the electrophysiologic world. There are many implementations (mostly digital) out there to try to deal with noise in the form of line interference, respiratory, and muscle. Each one has varying degrees of success but they all suffer from the requirement of limiting filter related distortion. An ECG machine is no good if the filters it uses to eliminate interference results in distorting the waveform to a degree that it causing diagnostic errors either in adding something that is not there or removing something that would be important. It's always a balancing act and most manufacturers rely on the side of preserving signal integrity hence for diagnostic-mode ECGs they require the patient to lay still. For monitoring-mode ECGs, such as Holters, signal distortion is more relaxed and so the filters are better. I designed the ECG module with diagnostic modality in mind. The obvious way to reduce interference caused by muscle would be to decrease the bandwidth of the module but it would start distorting the QRS complex (that spike you see on an ECG signal) as it looks like muscle interference. Even more so, the J-point and accompanying ST-segment would become even more distorted as well since they follow right after the QRS complex and that is dangerous to mess with because the morphology of the ST-segment in relation to other features forms one of the core findings for diagnosing a heart attack*.

*There are lots of other things on just an ECG alone for hinting at a heart attack such as T-wave peaking/inversion/QRS widening/conduction delay resulting in left-bundle branch block to name a few but the ST-segment is classic.

[f4eru]

Vaisala  teardown :



(sensors explained)

[Andrew McNamara]

Having used those Lab Gruppen amps, I can say they are hard-core touring amps, not audiophool snake oil - the mosfets are a switching stage that produce a waveform within about 10V of the target voltage, then the bipolar transistors are used in class B for linearity (which they call class TD, or "tracking D").  The result is very little dissipation when run at any power level - it's quite surreal how little heat comes out of them compared with most high-power touring amps. That board is just the power output stage (obviously) - there's a lot more going on inside a complete unit - smart power management, signal processing, remote management, etc.

They have a lot of other smarts in them, such as modelling the upstream circuit breaker and modulating their total power demand to avoid trips when running near the limit (you only run one of these per circuit - they draw 16A continuous at full power - although you can program lower limits if needed). They also cope gracefully with saggy mains voltages and other abuse - they are designed for touring use, so they're almost indestructible/idiot proof. They also have (proprietary) remote monitoring and management over ethernet.

There's some superficial information about their technology here: http://labgruppen.com/technology. There's more documents at http://labgruppen.com/download-all.

As noted in the YouTube comments, there are also counterfeits units - a tribute to how successful Lab Gruppen are, I guess.

[jonovid]

 the ECG is interesting! so how much medical red tape do you need to cut to get a medical electronics project in to a hospital ?
their maybe $ in new types of medical electronics.  like an improvement on an existing medical design.

the radiosonde altimeter or an altitude meter  the diaphragm and the encoder is missing.
humidity and temperature sensors are on the white strip sticking out of the box.  altimeter / barometer is meant to be inside the box.  no battery

here is the little I know about weather balloon radiosondes.
the radiosonde has a basic radio transmitter circuit and the modulator circuit the chip. the altimeter / barometer  by air pressure or altitude meter switches the modulator chip Hz as the radiosonde package climbs in altitude attached to the weather balloon along with a parachute and a radar reflector for tracking. this is released at the weather radar ground station. the greater the altitude, the lower the pressure, so eventually the weather balloon will burst then the parachute opens as the radiosonde descends in altitude.  humidity and temperature sensors are switched in & out along with the altitude data in a pattern of specific order of modulation. like packets of data  or tone multi frequency signaling system. so the ground station has a minute by minute report. of humidity ,temperature & altitude along with the radar that gives wind speed & direction. radiosonde are expendable disposable units. 

Australia had weather balloon radiosondes made under the AWA & Astor trademarks in the 1960s many by Electronics Industries Ltd (E.I.L.) Centre clayton in Victoria.& used in Australia & overseas. the 1960s 72MHz VHF radiosondes had military radio batteries made by Eveready-union carbide that powered two radio valves. one a VHF transmitter circuit the other the modulator circuits.

https://www.eevblog.com/forum/rf-microwave/australian-made-awa-1960's-vhf-twin-valve-radiosonde/

[Vgkid]

That was a rather enjoyable mailbag episode. Looking forward to the camera.

[johnh]

Had a note to go with the TV. Finished taping the box up and then put the label on it. Then I noticed I hadn't put the note.
Couldn't be bothered opening it all up and putting it in and resealing it.

Found the TV and portable short wave radio and some other bits and pieces( sata cables) and a State Electricity Commission   linesman handbook in the skip, on walk home from the train station to home. 
There was group of people rummaging through the content of skip. Spoke to one of women and said the lady had died recently and husband had died a while ago and  they were clearing the house of all the contents and this was the 4th skip (small).  Over the next noticed another 2 skips.  The last two seemed to have kitchen? cupboards and some broken chairs

[Windfall]

Dave unpacking some astroturf must be the most hilarious mailbag moment yet.

[128er]

Dave unpacking some astroturf must be the most hilarious mailbag moment yet.

I remember a pretty old episode where Dave received a black negligee. Fortunately he was not wearing it    

But the turf is also very strange.

[trophosphere]

the ECG is interesting! so how much medical red tape do you need to get a medical electronics project in to a hospital ?
this is $ in new types of medical electronics.

I want to be upfront and say that the red tape is huge. Getting a medical device into hospitals requires so much time and money that it can often exceed the total time and money spent on R&D. For a Class III Type BF device, the time required would be on the order of years.

[mmagin]

WRT ECG interference, is this partially an issue of skin conductivity?  If the skin is not moist enough (or electrode gel is not applied), won't the increased source impedance make the signal to noise ratio a lot worse?

[mtdoc]

WRT ECG interference, is this partially an issue of skin conductivity?  If the skin is not moist enough (or electrode gel is not applied), won't the increased source impedance make the signal to noise ratio a lot worse?

Yes, absolutely, conductivity between skin and electrode is important. But electrode gel is not used.  ECG electrodes have a gel-like adhesive designed for that purpose and if needed the skin is shaved...  ECG techs routinely have to replace electrodes if they are getting too noisy of a trace.

Scalp electrode placement for EEGs is even more critical and in that case conductive gel is often used. People generally don't like their heads shaved.

[Windfall]

Dave unpacking some astroturf must be the most hilarious mailbag moment yet.

I remember a pretty old episode where Dave received a black negligee. Fortunately he was not wearing it    

Clearly a gift for Nicole (and, in that respect, for Dave too  ). But astroturf ? That's so delightfully off-topic it's almost creative.

[PA0PBZ]

This must be the first time I heard someone calling feedthrough capacitors "penetrators". Is this aussie-talk or just Dave?

[thm_w]

Dave has some serious traps.

It has been a bit difficult to source the right jacks for this application as it's a rather niche product. These are indeed standard for certain bioelectrical monitoring applications such as ECGs and EEGs. They are referred to as 1.5mm safety DINs. One source of 90 degree jacks is here but I found this much later after I sent you the board. Like all things medically related, they are pricey as well. The reason for the cutout was I was trying to use the rings and PCB sort of as a brace for when the user would be pushing and pulling the electrode plugs in/out of the jacks. I never thought about what would happen if the user applied stress orthogonal to the intended insertion direction. Good catch. Will have to think about some sort of brace or just use the 90 degree jacks instead.

This might be the source: http://www.plastics1.com/Gallery-CCS.php?FILTER_CLEAR&FILTER_FNAME=36619 but no price is listed (and http://www.plastics1.com/Gallery-CCS.php?FILTER_CLEAR&FILTER_FNAME=41829).
They don't look as good as yours though. Could tighten up the PCB tolerance and drop a little glue on the edges if its an issue.

[trophosphere]

@thm_w

Thank you for the link. I just e-mailed them to see if they could give me a quote for a small quantity.

[Brumby]

This must be the first time I heard someone calling feedthrough capacitors "penetrators". Is this aussie-talk or just Dave?

I have heard of "Feedthrough caps" before, but cannot recall hearing about "penetrators" until this - but I must admit I haven't been taking in much RF material since Electronics Australia bit the dust.

[TheSteve]

This must be the first time I heard someone calling feedthrough capacitors "penetrators". Is this aussie-talk or just Dave?

Must be a special dirty feedthrough cap.

[Smokey]

Dave has some serious traps.
...

Admiral Ackbar approves this message!

[piranha32]

I found once a radiosonde, just a newer model. It was powered from a small lead-acid battery. Unfortunately it did not survive impact on landing, the shell plastic was cracked, and the acid spilled inside the styrofoam battery compartment.

Re the tuning caps: it's interesting how they hacked the second one. It is mounted in reverse, so the ceramic body is just a mechanical mount, and the shaft is the active element. It looks like a tuned inductor, where the bend in the copper strip is one turn coil, and the capacitor shaft is used as a tuning slug. Looks just right for 433MHz.

[HighVoltage]

It has been a bit difficult to source the right jacks for this application as it's a rather niche product. These are indeed standard for certain bioelectrical monitoring applications such as ECGs and EEGs. They are referred to as 1.5mm safety DINs. One source of 90 degree jacks is here but I found this much later after I sent you the board. Like all things medically related, they are pricey as well. The reason for the cutout was I was trying to use the rings and PCB sort of as a brace for when the user would be pushing and pulling the electrode plugs in/out of the jacks. I never thought about what would happen if the user applied stress orthogonal to the intended insertion direction. Good catch. Will have to think about some sort of brace or just use the 90 degree jacks instead.

That is indeed a very nice and interesting ECG project
Can you give a source to those straight 1.5 mm connector you are using on the PCB?
Thanks

[trophosphere]

That is indeed a very nice and interesting ECG project
Can you give a source to those straight 1.5 mm connector you are using on the PCB?
Thanks

Thanks for the comment. I have been getting my connectors from MVAP. Scroll down and you will see the connectors in different colors. Note that they had to validate my account on their side after I registered. I called their customer service line to expedite it and it went through smoothly but it was kind of annoying to be honest.

[TheAmmoniacal]

(...)

Did the version you sent to Dave include the 50 Hz or 60 Hz notch filter?

[trophosphere]

Did the version you sent to Dave include the 50 Hz or 60 Hz notch filter?

It included a 50Hz notch filter. I made sure I sent a 50Hz version as I had made a blue mark on the module with a sharpie after I had set the notch frequency. In fact, I used Dave's sample to derive the frequency response graph that is present in the module's datasheet.



The actual trough of the of the notch filter is around 49.7Hz (because of component tolerances) but I didn't want to specifically trim it as I won't be doing that to other units and I wanted a fair representation of what others will be getting. In essence, there is no cherry-picking as I have another 50Hz version sitting on my desk that has a deeper notch depth around -34dB as its component values are more closer to ideal.

[TheAmmoniacal]

Alright, a starter kit has been ordered Thanks.

[SeanB]

Those radiosonde batteries are a special design, an 18V battery using a stack of dry electrodes, zinc sheet anode, manganese dioxide and ammonium perchlorate with sawdust powder blend electrolyte and a graphite cathode on a zinc sheet, which is also the anode of the next cell in the series pack. These are in a waxed paper housing with a small vent on top, and a hole at the base that allows you to activate them. Activation is simple, you take it out of the sealed plastic pouch it comes in, in the base of the sonde, and dunk it in water for 5 minutes to wet the electodes, which then wet the electrolyte and activate the battery.

Then you put it back in, plug in the power lead and close the case, and take the small strip of teletype tape ( or for some the calibration data printed on a strip of paper, for the particular device from the manufacturer, but the tape is machine readable) and put it in the recorder to get the calibration data for the particular unit, as they can have a large unit to unit variation, and there is no cal before flight. Then you attach the yellow handle to the inflated balloon, which also seals the rubber fill section, unwind the cord and let it go.

[ornea]

Incase it has not been mentioned,  i believe the batteries are water activated.  As the ballon is readied for deployment,  the battery is sprayed with water.

[Regulus]

Just a few more facts about the radiosonde.

The round metal thingy shown separately (with the text "humidity sensor") is actually a barometer aneroid (i.e. a sensor for atmospheric pressure). There is vacuum inside the flexible metal capsule. Atmospheric pressure pushes the sides closer together and the distance is measured. In the case of this particular sensor the distance is measured by sensing the capacitance of two plates inside the capsule, attached to the two sides. Atmospheric pressure decreases with altitude, sonde altitude is calculated based on that measurement.

Judging by the antenna this radiosonde might have been a 1.6 GHz band version. I am not an RF guy so don't take my word for it... Flight of the sonde can be tracked with a radiotheodolite or a tracking radar. Wind speed and direction at different altitudes can be calculated from the radiosonde track.

This sonde has lost it's temperature sensor, which has been at the end of the flexible "arm", the chip remaining there is the humidity sensor. This version of a radiosonde has used only capacitive sensors (including the temperature sensor), the mysterious IC is an ASIC used to measure the capacitance values.

Most of the real magic in this device is in the sensors and how they are measured, not really in the radio transmitter (although it appeared to have rather unique construction). A radiosonde is a disposable precision instrument with e.g. temperature measurement range from +60 degC down to -90 degC with accuracy of 0.2 degC, quite a special device.