Just pretend that the reason they threw it away was because the bit that makes this worth the 500 is somehow broken and thus the mistake doesn't look so bad suddenly.
Hey Dave, I'm from Philips Healthcare and thanks for dropping all our IP on the internet :) No really, what a marvel in engineering, wouldn't you agree? So that unit on the screenshot you showed in 1:23 is our Lumify POCUS. POCUS is Point Of Care UltraSound and is a complete US system in a transducer. It's connected to and fed by a normal Android tablet using just one USB connector. Indeed a huge step from the Philips/ATL cart you showed. Mind you, this is an extremely powerful and versatile tool in the field, but does not hold a candle to what is possible nowadays with a more potent sized US unit. For the Hospital we still have those "Wall-E" like unit allbeit way lighter and extremely ergonomic with flat/touch screens and the works. Brilliant 3D processing, it's incredible, really. Thanks for sharing and showing the engineering that goes into devices like this. Learned something myself, because I'm more into the MRi and CT corner of our company. Do you want me to deliver a 6300 kg magnet for teardown? You will need some heavier tools for that...
Thanks for that. Let me check, but I'm fairly sure 6300kg will be too much for the lab building lift... I've got a nice hand trolley though if it fits :-D
All these wires must be co-ax, to minimize crosstalk. 15:51 Outer braids and inner conductors are welded separately. So they used good-old hypertronics connector with 256+ pin for 128 elements. Extra pins are also used for identification EEPROM (3:59, top right corner in PCB). These wires also have some extra loops in the connector cage to ensure all co-ax wires approximately same length. With my experience on industrial ultrasonic testing, ~1ns phase alignment should be enough for
@giangle2849 Dear sir, do you know if there is a table for all the resistor circuit lists inside each of these probe that we need to change in order to make it compatible with different machines? I've been looking into this but to no avail. Say, I want to use the linear probe on the Sonoace X8 and I have with me one probe from the Sonoace X6, which resistor do I need to change inside the probe to make it compatible and interchangable?
Those inductors are not for EMC. They are tuning inductors, one for each transducer element. The PZT transducer elements are basically Type II Ceramic capacitors.
@H Higgins if it is tuning I think they'd make emc worse, not better? If it was resonating with the capacitor it would cause an amplification in the current at the frequency of Interest. Admittedly maybe less at other frequencies.
@@EEVblog The medical imaging people spare no expense! Makes me think of that clip from Jurassic Park you always play. They need to match the ceramic elements as well as the inductors to have all of them resonate at the same frequency and with the same Q. I'm not sure about Philips, but Olympus (I think they are the leader in ultrasonic imaging transducers for medical and NDT&E) has very tight specs on channel match. I think they have to sort through their components and bin them to get ones that match...
@@EEVblog I guess it's more impedance modifying than tuning if they are working over more than an octave of frequency. But it probably makes it easier to get enough current through the transducer at a reasonable voltage.
in fact 128 elements. It is nearly standart for ultrasounds to have 128 elements for US and Europe brands. For Japanese brands, it is generally 192. This is related with analog channels. But now there are matrix probes.
Starting around 15:38 we see the board connections are actually numbered 1 to 128, and a minute later we can see the head itself is numbered likewise (though a bit harder to read).
Ultrasound professional here. The C5-2 used on the ATL/Philips HDI-5000 was carried over to the iU-22 by Phillips that was EOL'd in 2017 and many are still in use today. The C5-2 was superseded by the newer C5-1, a much better probe, but many of these are still in use regardless. The physical technology has changed some since then (crystals are now grown in one single piece rather than cut to size), but the major changes have been made in machine size/power and software.
Dear sir, do you know if there is a table for all the resistor circuit lists inside each of these probe that we need to change in order to make it compatible with different machines? I've been looking into this but to no avail. Say, I want to use the linear probe on the Sonoace X8 and I have with me one probe from the Sonoace X6, which resistor do I need to change inside the probe to make it compatible and interchangable?
I work as a field service engineer for a large medical device manufacturing company. I can really appreciate the quality and engineering that goes into hospital and laboratory instruments. In could watch these type of videos all day long. Would love to send you some failed or broken boards and power supplies for you tare-down and commentary.
@@raymondcourtois67 about like comparing 1/4 inch by 1/4 inch glass fuse to the fuse for the ludicrous mode on Tesla Motors or 2 a utility fuse versus the 1 1/4 in fuse but yes I agree with the knife idea.
I worked for Atl in the late nineties as a field service engineer, great to see inside of the scan heads as we just swapped them over when they went faulty and always wondered what was inside of them. As we use to send the faulty ones back to the USA for repair. There was lots of shielding to stop interference on the signal going back to the ultrasound, hospitals were a very electrically noisy environment. great memories to see the HDi 3000 again. Thanks for the teardown Dave
They dont skimp on cost, either. That grey PTFE tape used under the copper foil is like 600 USD a roll. Remember using it at work and it just about costs more than gold. Even a little skinny roll of it costs like 100 USD.
I work as an engineer at a hospital in Norway and i can tell you that a normal 1 phase probe cost around 6000 USD. And if you run over the wire, just a litte bit, you have to change the whole thing. Also you can get a probe that goes trough your throat and rotates 180 degrees, so you can see the heart without any bones or lungs in the way. (But you have to be very skilled with anatomy because you are basically blind) But Those are very expensive. Because if you have a probe that close to the heart only 60 µA is enough to kill a man. Thats why they used to check for ampere leakage after each use in the old days, but now we only check it every 2nd year.
48k views on the original video is just 48 hours. I'm not surprised, these were badass videos!! Thanks Dave!!!! AMAZING the level of engineering in these things.
Couldn't be arsed counting the stripes on the head like some people, however as a giveaway, there's 128 inductors, at 16:21 there's 1 and 128 written on the head and at 15:38, 1 and 128 written on the PCB where the flexi from the head attaches to the PCB.
How do you get it to work? I downloaded the App and placed my phone against my stomach but there was nothing on the display. Am I doing something wrong?
Yup, pretty much all of the electronics in that beast have been miniaturized to fit in the transducer. On the other hand, if you've got the space and budget for a modern ultrasound cart, it will blow that thing out of the water (to the point where it's often preferred to MRIs and CTs).
Note that all elements have their own tiny coax inside the big coax to minimise crosstalk since the returning signal is so tiny. Btw we just bought the new model at work and it is used with a standard samsung android tablet
I figured that would not be a medical specific tablet. Also did you hear about the medical device that was based on the Game Boy of some description Hardware. Something with cardiac something or other.
16:22 it literally shows the element numbers starting at 1, finishing at 128. Industrial probes usually have a stamp on them to show you which end is element 1 so you don't put it back to front.
There are still companies buying, selling, servicing the HDI 3000 and 5000 series, and are always in need of replacement boards and probes. SonoDepot comes to mind (www.ultrasoundrepair.com/philips/hdi-3000). Some probes go for substantially more than $500 with some systems (which originally cost >$100K) going for (as we say) peanuts. If you are in a third world country (or parts of a country that seam to be part of a third world country) these rehabs can be life savers. Oh! And through the 80s, 90s and beyond I was part of teams developing ultrasound and other medical products (HP, Philips, Siemens).
@1:36 That tablet in the photo you showed looked exactly like my Nvidia Shield Tablet. They had it up-side-down in the photo with it sitting on the buttons! lol It's not a custom tablet or new, it's one of the first gamer tablets featuring an Nvidia GPU. So it looks like the ultrasound probe is now made to work with a custom app in any Android tablet with a micro USB port. Maybe they have one with the more USB-C too. *NOTE* I just read up on the probe and it says "compatible with select Android devices". Maybe the Nvidia Shield Tablet is compatible because of the GPU making it capable of showing a realtime display that would be needed.
Actually had an ultrasound done yesterday, a Toshiba machine, about half the physical size of this one, appears the probes use exactly the same connector though, maybe it's a standard used by every manufacturer.
Anything but. Each manufacturer's probes are proprietary. Some may share similar connectors that are based on some industry standard, but there is ZERO compatibility between manufacturers.
Thanks so much for this. Presently studying for ultrasonography exams (I'm an anaesthetist) - so much better to simply see the inside of a probe, rather than reading numerous slightly confused (and occasionally incorrect) textbook descriptions of it!
I've designed several military test sets using those high pin count connectors, they are very expensive, and are not as good as they look. We had many problems with the pins not making good contact, requiring the operator to un-connect, clean and re-connect, sometimes several times. There are other high pin count connectors that are even more expensive, and some work better than others, but overall with this many connector pins your going to have times then some of the pins don't work as expected. Quite the pain actually!
Back in the '80s the firm I worked at bought several ATE Systems/Wayne Kerr 'Beaver' (no sniggering) functional ATE systems. The test interfaces sat on top of the main system, connecting through several multi-way 'knife and fork' contact connectors from Ericsson IIRC that they'd arrived at after rejecting many others. [Edit] Provided you made sure there was nothing sitting on the base connectors before putting an interface on they worked perfectly.
I think that the cables used here are some form of micro coax. As seen at 5:25, there appears to be a ground bar or elongated solder joint of some sort that is likely connecting the shields. This type of arrangement is also common in laptop display cables but I think that those are thinner than the ones in the video.
I believe you are right other people have mentioned this before as well. Also I would not call them miniature on call them sub micro-miniature miniature miniature.
Or for that matter me at your sub micro Pico if that was a thing. How would you describe something that small if it's smaller manager and sworn subminiature and even micromanager?
Great teardown! The Lumify C5-2 has all the works in the probe, seriously impressive. The current equivalent would be a C5-1. The probe is a linear array with a single row of elements.
it's a zero insertion force connector, used them on a Automatic Test Equipment unit i built back in the 90's i think. I used it to connect the different bed of nails units to the main ATE in it's 19" rack. Turning the handle closes the contacts and locks the connector in place. Also we used Arinc 300 connectors that had about the same number of pins on one Aircraft intercom system. the unit had to be jacked in and out of the tray with screw jacks due to the force required to mate and un-mate 300 contacts.
Why they used metal tape shielding was they need to ultrasonic weld the head case so not to get any bodily fluids in and sterilise it. Interesting to watch on the factory floor as they drop part in a drawer and come out welded. Cool to see that data on how they work looks better than I seen before.
Same element to transmit and receive hence you get aliasing artifacts in Doppler studies if your pulse repetition frequency is too low. Linear probes are phased arrays so beam forming is easily accomplished.
The thin center conductors with the 'loose' fit inside the insulating sleeve serve to minimize the capacitance from center conductor to shield. A big deal for recovering a weak RF signal efficiently.
The tablet-connected ones must do things closer to the transducer head, as it's a much thinner cable - but I guess they then don't need the fancy custom connector, all the inductors and shielding, etc and can convert the signals to digital data?
It would be nice if one of your followers could send you a modern USB one to tear down. Probably looks like any other USB peripheral once you get past the scanning wires, though.
9:31 - I would imagine it wasn't a metal case as that could cause capacitive coupling which the patient would feel as a tingle. Lesser reasons would be weight & temperature (those probes are cold enough with the gel applied - could you imagine how cold it would feel if it was metal!)
I struggle to see how a linear bank of antennae can get a 2D (XY) image. As I see the design it's basically like a version of a swinging radar which would give time/distance/angle reflections...
Those *are* your 2 dimensions: angle/direction and time of flight (i.e., distance). The image you see is a slice through the body in a plane coincident with the PCB of the probe; this is all why the pictures have that classic pie slice shape.
@@TheHuesSciTech Maybe I'm wrong, but the image is XY (ie a photo) with a depth component (density/whiteness). Given my understanding of this(little) the scan is basically 256(whatever) pixels wide on the head by one pixel high on the head. So where does the vertical resolution come from? That's my confusion.
@@PIXscotland The array is 2D, so there's a time element in the waves travelling along the transducer element; by controlling the relative phase of the driving signal, you can reproduce a 2D scan. Dave did mention an article describing how it works.
Huh. Just watched this and got it. ua-cam.com/video/gT0zV0XYKek/v-deo.html Didn't work how I expected/thought. Seeing the image from the curved device gave the clue. The Y in the photo is time/distance. Totally makes sense now.
I guess they choose copper tape instead of metal shielding to avoid having the patient directly connected to ground anywhere. You probably also find the cardiac measurement stuff on the MO disc board having a floating dc servo instead of direct ground connection.
Ah those connectors, DL type very common in the Recording studio world for hooking up Multitrack Machines. those crimp pins are about 45 cents per pin. A PITA to work on.
Dave I should have mentioned that that machine was worth less than the sum of its parts... if you're missing a part for your old machine (as many veterinarians are: we don't have the cash to buy the new latest and greatest) you'll pay a lot for them. However, the nerd in me is delighted that you did this ;)
Now I really want to see a teardown of the modern usb version. I wonder if they've condensed all those electronics down so small it can fit inside of the scanner itself.
At 6:15 that bar looks like a soldier strip that is terminating (bonding) individual cable shields to the PCB. Do you think the cable was stripped and soldered by hand?
I'm thinking that's the case and yes micro some major or whatever coax for sure been brought before and yes I'm pretty sure that's a bonding strip ever seen all the braids ground Connections in something like a big UPS or other high current high power equipment and others or all the ground straps on RF Equipment ohyeah talk about some rain action.
Would love to get the skinny on building a ~500 kHz (tunable / adjustable within 300 Khz) ultrasonic transducer. One fellow mentioned turning "alarms" into such critters.
Woops! These are worth a penny, haha. I really love your videos, Dave. I've recently binged a few, and it's brought back my passion for electronics. Thanks for sharing, and keep them coming.
I would imagine that newer versions of that probe have more circuitry inside in order to combine the wires in the lead into far fewer channels, perhaps via multiplexing or similar techniques.
There should be an acoustic lens covering elements. This is for first focusing mechanism. Then usg beam is focused on recieve by delay. cut size and shape of elements are directly related with image(signal) quality but also they use elements from same single piece crystal for one probe to achieve same quality on each element. The real channel (tx/rx channel)number for for transmitting and recieving may change by multiplexing. USG with more channel is better generally. Multiplexing cause noise and for USG noise is very very very important because RX signals are very weak and if u amplify u amplify noise too. So frontend quality is very important. Some people compares this machine with todays portable machines but believe me that HDI 3000 is a monster about quality though it is old.
Hello, I have always the query about manufacturing process of the ultrasound probes; how all these thin cables can be placed from end to end with the right order? When they have to solder/connect the cables from one side on the connector to the other side of the probe PCB how they know each cable in which pin it has to be soldered (for example pin 1 from connector's PCB must to be soldered at pin 1 at porbe's PCB). All these tiny cables have the same color and there so many… maybe it is stupid what I will say… but there is a possibility that each cable doesn’t have to be soldered at a specific place between two ends and all the cables soldered randomly and the software will re-position in right order all the image sectors?
Does that mean an element transmits in pulses and then receives using the same element, or the element has a trans and receive in it, or is there a large trans element behind and the small elements receive.
Interesting. I learned in this video the plane of the ultrasound signal on the screen goes away from the head of the transmitter. Seems obvious now but I was trying to figure out how they got a top down 2d picture from a single row of transducers!
50 V at 2 A max is 100 W. How do they power modern probes from battery powered tablet? I assume that you really need this power to penetrate body tissues.
thank you for this awesome video , i'm working on ATL HDI 1500 Ultrasound system , i'm trying to update the pc motherboard , but i have some problem with the old connectors , do you have any advice ?!
Those connectors are garbage, really bad. THey are used a lot on ATE's for testing aircraft equipment. I have been working and troubleshooting them a lot. They have so many bad connections all the time and resulted in so many false errors on DUT.
It's probably not too hard to make those connections. All the individual wires just run through parallel guides before they put those brown kapton pieces on. Then after that it gets braided and the insulation is put on
That's an awesome video I'm watching it from Pakistan and i work on ultrasound machines it's really a nice video kindly make another videos like this and tell more about element fittings ,, but it's a nice & amazing practical video.
Seems like potentially a LOT of power across that head. What was it, like 300 watt max? I know the entire point is to convert that energy into ultrasound, but sure those heads get super hot, right?
Don't know, but likely not a lot, as there is very little thermal mass there. And modern one in the same model work from USB-C, so it's more like 10's of watts than 100's.
That time when you destructively tear down something old and then find out they sell for up to $500 on ebay...
Wow, ain't that a bastard
Still, we are so grateful for you taking it apart! Very interesting ;)
well.. selling something for $500 doesn't necessarily mean it will be sold for $500 :D
Just pretend that the reason they threw it away was because the bit that makes this worth the 500 is somehow broken and thus the mistake doesn't look so bad suddenly.
All usg stuff are overpriced on ebay (i repair them as daily work), this one is worth $50 max but you need to sell it in Africa.
Hey Dave, I'm from Philips Healthcare and thanks for dropping all our IP on the internet :) No really, what a marvel in engineering, wouldn't you agree?
So that unit on the screenshot you showed in 1:23 is our Lumify POCUS.
POCUS is Point Of Care UltraSound and is a complete US system in a transducer. It's connected to and fed by a normal Android tablet using just one USB connector. Indeed a huge step from the Philips/ATL cart you showed. Mind you, this is an extremely powerful and versatile tool in the field, but does not hold a candle to what is possible nowadays with a more potent sized US unit. For the Hospital we still have those "Wall-E" like unit allbeit way lighter and extremely ergonomic with flat/touch screens and the works. Brilliant 3D processing, it's incredible, really. Thanks for sharing and showing the engineering that goes into devices like this. Learned something myself, because I'm more into the MRi and CT corner of our company. Do you want me to deliver a 6300 kg magnet for teardown? You will need some heavier tools for that...
Thanks for that.
Let me check, but I'm fairly sure 6300kg will be too much for the lab building lift...
I've got a nice hand trolley though if it fits :-D
@@EEVblog ua-cam.com/video/HRLHtnza2jM/v-deo.html
Oooh ide love a big magnet like that.. you can come to my shed and do an autopsy on it dave
@@northcoastaudiovisualncavc9706 ua-cam.com/video/0ZMEVW7AfF4/v-deo.html
This a great video but the companion video should be ‘how has Phillips actually manufactured this beast
All these wires must be co-ax, to minimize crosstalk. 15:51 Outer braids and inner conductors are welded separately. So they used good-old hypertronics connector with 256+ pin for 128 elements. Extra pins are also used for identification EEPROM (3:59, top right corner in PCB).
These wires also have some extra loops in the connector cage to ensure all co-ax wires approximately same length. With my experience on industrial ultrasonic testing, ~1ns phase alignment should be enough for
They are indeed coax wires. And a pain to hand solder.
I need to learn how to fix probe problems, could you give me training??
@giangle2849
Dear sir, do you know if there is a table for all the resistor circuit lists inside each of these probe that we need to change in order to make it compatible with different machines?
I've been looking into this but to no avail. Say, I want to use the linear probe on the Sonoace X8 and I have with me one probe from the Sonoace X6, which resistor do I need to change inside the probe to make it compatible and interchangable?
Hello Sir, could you check my post if you please about a query about probe cable manufacturing that I have (I posted today)?
Those inductors are not for EMC. They are tuning inductors, one for each transducer element. The PZT transducer elements are basically Type II Ceramic capacitors.
@H Higgins if it is tuning I think they'd make emc worse, not better? If it was resonating with the capacitor it would cause an amplification in the current at the frequency of Interest. Admittedly maybe less at other frequencies.
Wouldn't that cause huge tolerance issues in the resonant frequencies between channels?
@@EEVblog The medical imaging people spare no expense! Makes me think of that clip from Jurassic Park you always play. They need to match the ceramic elements as well as the inductors to have all of them resonate at the same frequency and with the same Q. I'm not sure about Philips, but Olympus (I think they are the leader in ultrasonic imaging transducers for medical and NDT&E) has very tight specs on channel match. I think they have to sort through their components and bin them to get ones that match...
@H Higgins Sure. By making the transducer much higher Q at the transmit frequency, they are also filtering out other frequencies and harmonics.
@@EEVblog I guess it's more impedance modifying than tuning if they are working over more than an octave of frequency. But it probably makes it easier to get enough current through the transducer at a reasonable voltage.
10:42 it's 266 stripes /133 elements
The connector has 260 pins on it, too
in fact 128 elements. It is nearly standart for ultrasounds to have 128 elements for US and Europe brands. For Japanese brands, it is generally 192. This is related with analog channels. But now there are matrix probes.
Starting around 15:38 we see the board connections are actually numbered 1 to 128, and a minute later we can see the head itself is numbered likewise (though a bit harder to read).
Ultrasound professional here. The C5-2 used on the ATL/Philips HDI-5000 was carried over to the iU-22 by Phillips that was EOL'd in 2017 and many are still in use today. The C5-2 was superseded by the newer C5-1, a much better probe, but many of these are still in use regardless. The physical technology has changed some since then (crystals are now grown in one single piece rather than cut to size), but the major changes have been made in machine size/power and software.
Dear sir, do you know if there is a table for all the resistor circuit lists inside each of these probe that we need to change in order to make it compatible with different machines?
I've been looking into this but to no avail. Say, I want to use the linear probe on the Sonoace X8 and I have with me one probe from the Sonoace X6, which resistor do I need to change inside the probe to make it compatible and interchangable?
Can I ask you questions about ultrasound probes for a senior design project my group and I are working on?
I work as a field service engineer for a large medical device manufacturing company. I can really appreciate the quality and engineering that goes into hospital and laboratory instruments. In could watch these type of videos all day long. Would love to send you some failed or broken boards and power supplies for you tare-down and commentary.
I'm sure he'd love to receive them maybe you should get on a private message or whatever they call it and do it.
Thats not a connector, THIS is a connector!
He should have pried it open with the giant knife..
@@raymondcourtois67 about like comparing 1/4 inch by 1/4 inch glass fuse to the fuse for the ludicrous mode on Tesla Motors or 2 a utility fuse versus the 1 1/4 in fuse but yes I agree with the knife idea.
@@raymondcourtois67 YESSS
I worked for Atl in the late nineties as a field service engineer, great to see inside of the scan heads as we just swapped them over when they went faulty and always wondered what was inside of them. As we use to send the faulty ones back to the USA for repair. There was lots of shielding to stop interference on the signal going back to the ultrasound, hospitals were a very electrically noisy environment. great memories to see the HDi 3000 again. Thanks for the teardown Dave
The amount of engineering that goes into medical devices never fails to blow my mind. Very cool!
Same
You know it's good when the shielding is actually interesting!
@Lassi Kinnunen True words, some of the medical "you could by a house instead" devices is shockingly bad (pun intended).
They dont skimp on cost, either. That grey PTFE tape used under the copper foil is like 600 USD a roll. Remember using it at work and it just about costs more than gold. Even a little skinny roll of it costs like 100 USD.
I work as an engineer at a hospital in Norway and i can tell you that a normal 1 phase probe cost around 6000 USD. And if you run over the wire, just a litte bit, you have to change the whole thing. Also you can get a probe that goes trough your throat and rotates 180 degrees, so you can see the heart without any bones or lungs in the way. (But you have to be very skilled with
anatomy because you are basically blind) But Those are very expensive. Because if you have a probe that close to the heart only 60 µA is enough to kill a man. Thats why they used to check for ampere leakage after each use in the old days, but now we only check it every 2nd year.
8:47 you could say there are now two thermosingles lmao
That was a great one!
48k views on the original video is just 48 hours. I'm not surprised, these were badass videos!! Thanks Dave!!!! AMAZING the level of engineering in these things.
Couldn't be arsed counting the stripes on the head like some people, however as a giveaway, there's 128 inductors, at 16:21 there's 1 and 128 written on the head and at 15:38, 1 and 128 written on the PCB where the flexi from the head attaches to the PCB.
To think, that entire massive machine has basically been reduced to "there's an app for that".
Ha ha
oof haha
Still pretty amazing demonstration of the evolution of tech
How do you get it to work? I downloaded the App and placed my phone against my stomach but there was nothing on the display. Am I doing something wrong?
Nah, last time I had an ultrasound they where still large-ish. I think that they have gone up in resolution and features rather than down in size.
Yup, pretty much all of the electronics in that beast have been miniaturized to fit in the transducer. On the other hand, if you've got the space and budget for a modern ultrasound cart, it will blow that thing out of the water (to the point where it's often preferred to MRIs and CTs).
This is one of the most impressive teardowns in recent memory if only because I had no idea how these worked!
Note that all elements have their own tiny coax inside the big coax to minimise crosstalk since the returning signal is so tiny.
Btw we just bought the new model at work and it is used with a standard samsung android tablet
I figured that would not be a medical specific tablet. Also did you hear about the medical device that was based on the Game Boy of some description Hardware. Something with cardiac something or other.
16:22 it literally shows the element numbers starting at 1, finishing at 128. Industrial probes usually have a stamp on them to show you which end is element 1 so you don't put it back to front.
There are still companies buying, selling, servicing the HDI 3000 and 5000 series, and are always in need of replacement boards and probes. SonoDepot comes to mind (www.ultrasoundrepair.com/philips/hdi-3000). Some probes go for substantially more than $500 with some systems (which originally cost >$100K) going for (as we say) peanuts. If you are in a third world country (or parts of a country that seam to be part of a third world country) these rehabs can be life savers. Oh! And through the 80s, 90s and beyond I was part of teams developing ultrasound and other medical products (HP, Philips, Siemens).
@1:36 That tablet in the photo you showed looked exactly like my Nvidia Shield Tablet. They had it up-side-down in the photo with it sitting on the buttons! lol
It's not a custom tablet or new, it's one of the first gamer tablets featuring an Nvidia GPU. So it looks like the ultrasound probe is now made to work with a custom app in any Android tablet with a micro USB port. Maybe they have one with the more USB-C too.
*NOTE* I just read up on the probe and it says "compatible with select Android devices". Maybe the Nvidia Shield Tablet is compatible because of the GPU making it capable of showing a realtime display that would be needed.
maybe the graphics needed for gaming made it desirable ..
Wow. Complex piece of kit! Thanks for opening it up.
I think the transducer array very pretty. Something about super precise, repeating complex objects is just satisfying.
15:33 there is 128 writen in the board, there is 128 wires but I counted 134 points at the tip where is soldered to the transducers.
Thank You. Opening the transducer actually helps me to visualize the physics component for my ultrasound program.
Actually had an ultrasound done yesterday, a Toshiba machine, about half the physical size of this one, appears the probes use exactly the same connector though, maybe it's a standard used by every manufacturer.
Anything but. Each manufacturer's probes are proprietary. Some may share similar connectors that are based on some industry standard, but there is ZERO compatibility between manufacturers.
Thanks so much for this. Presently studying for ultrasonography exams (I'm an anaesthetist) - so much better to simply see the inside of a probe, rather than reading numerous slightly confused (and occasionally incorrect) textbook descriptions of it!
I've designed several military test sets using those high pin count connectors, they are very expensive, and are not as good as they look. We had many problems with the pins not making good contact, requiring the operator to un-connect, clean and re-connect, sometimes several times. There are other high pin count connectors that are even more expensive, and some work better than others, but overall with this many connector pins your going to have times then some of the pins don't work as expected. Quite the pain actually!
Back in the '80s the firm I worked at bought several ATE Systems/Wayne Kerr 'Beaver' (no sniggering) functional ATE systems. The test interfaces sat on top of the main system, connecting through several multi-way 'knife and fork' contact connectors from Ericsson IIRC that they'd arrived at after rejecting many others.
[Edit] Provided you made sure there was nothing sitting on the base connectors before putting an interface on they worked perfectly.
Language translation: "tick" in Ozzyland translates to "chooch" in Canuckistani, as per AvE.
I think that the cables used here are some form of micro coax. As seen at 5:25, there appears to be a ground bar or elongated solder joint of some sort that is likely connecting the shields. This type of arrangement is also common in laptop display cables but I think that those are thinner than the ones in the video.
I believe you are right other people have mentioned this before as well. Also I would not call them miniature on call them sub micro-miniature miniature miniature.
Or for that matter me at your sub micro Pico if that was a thing. How would you describe something that small if it's smaller manager and sworn subminiature and even micromanager?
I'm a big connector fanboy and I had a good time !
Always wanted to teardown one of these myself ! Really glad !
Olde 24trk tape machines/mixers used same connectors... so 24*3 (24 tracks and +/-/gnd) *2 (input/output) = 144 pins... used to take a long day
Great teardown! The Lumify C5-2 has all the works in the probe, seriously impressive. The current equivalent would be a C5-1. The probe is a linear array with a single row of elements.
it's a zero insertion force connector, used them on a Automatic Test Equipment unit i built back in the 90's i think. I used it to connect the different bed of nails units to the main ATE in it's 19" rack. Turning the handle closes the contacts and locks the connector in place.
Also we used Arinc 300 connectors that had about the same number of pins on one Aircraft intercom system. the unit had to be jacked in and out of the tray with screw jacks due to the force required to mate and un-mate 300 contacts.
The number of pins are different but the connection seems to be the same as the connector I use with my deep-fryer .. no pun-intended
Wow! Even the release handle has Rf shield spring on that ZIF connector.
Cheers Dave - I've always wondered how the head of one of those probes worked. Thanks very much - great video.
Why they used metal tape shielding was they need to ultrasonic weld the head case so not to get any bodily fluids in and sterilise it. Interesting to watch on the factory floor as they drop part in a drawer and come out welded. Cool to see that data on how they work looks better than I seen before.
Same element to transmit and receive hence you get aliasing artifacts in Doppler studies if your pulse repetition frequency is too low. Linear probes are phased arrays so beam forming is easily accomplished.
The thin center conductors with the 'loose' fit inside the insulating sleeve serve to minimize the capacitance from center conductor to shield. A big deal for recovering a weak RF signal efficiently.
Dave's obsession with test gear now extends to human test gear apparently...
Notice all the inductors are 90 degrees to one another, similar to twisted pair wiring to minimize crosstalk.
I live an hour away from Bothel WA! Wild seeing that on a bit of kit being torn down by ya!
it looks like they can do all the processing in software now on a standard tablet with that new probe...makes sense when ya can already do SDR..
The elements have to be grounded to the shield on one side to avoid the possibility of shocking the patient.
The tablet-connected ones must do things closer to the transducer head, as it's a much thinner cable - but I guess they then don't need the fancy custom connector, all the inductors and shielding, etc and can convert the signals to digital data?
Disassembling old-school convex probe with mechanical beam deflection would be interesting too.
Only ultrasonic transducers ive seen looked lke mini piezo elements.. pretty neat
Thats a probe for abdomen ultrasound exam.
Its from from start of 2000-s.
It was around 4k $ new
Nah that's from 1992/95, 2000 models looked differently.
@@brainthesizeofplanet Same probe with only a slightly different ZIF connector is still in use today.
It would be nice if one of your followers could send you a modern USB one to tear down. Probably looks like any other USB peripheral once you get past the scanning wires, though.
9:31 - I would imagine it wasn't a metal case as that could cause capacitive coupling which the patient would feel as a tingle. Lesser reasons would be weight & temperature (those probes are cold enough with the gel applied - could you imagine how cold it would feel if it was metal!)
Would agree with the cold part for sure.
I struggle to see how a linear bank of antennae can get a 2D (XY) image. As I see the design it's basically like a version of a swinging radar which would give time/distance/angle reflections...
Those *are* your 2 dimensions: angle/direction and time of flight (i.e., distance). The image you see is a slice through the body in a plane coincident with the PCB of the probe; this is all why the pictures have that classic pie slice shape.
@@TheHuesSciTech Maybe I'm wrong, but the image is XY (ie a photo) with a depth component (density/whiteness).
Given my understanding of this(little) the scan is basically 256(whatever) pixels wide on the head by one pixel high on the head.
So where does the vertical resolution come from? That's my confusion.
@@PIXscotland The array is 2D, so there's a time element in the waves travelling along the transducer element; by controlling the relative phase of the driving signal, you can reproduce a 2D scan. Dave did mention an article describing how it works.
Huh. Just watched this and got it.
ua-cam.com/video/gT0zV0XYKek/v-deo.html
Didn't work how I expected/thought. Seeing the image from the curved device gave the clue. The Y in the photo is time/distance. Totally makes sense now.
The ultrasounds as they penetrate the body change speed with the material density. That's why the unit has a doppler shift detection component
I guess they choose copper tape instead of metal shielding to avoid having the patient directly connected to ground anywhere. You probably also find the cardiac measurement stuff on the MO disc board having a floating dc servo instead of direct ground connection.
Is this a Zero-Insertion-Force-Connector?
Ooh, I didn't think of that. Will have to check...
@@EEVblog It should be, I used to play with ITT Cannon connectors like that and they were ZIF
@@EEVblog yeah, it looks like the contacts move when you turn the locking part.
It is. Toshiba MRI coils use the same design
Ah those connectors, DL type very common in the Recording studio world for hooking up Multitrack Machines. those crimp pins are about 45 cents per pin. A PITA to work on.
They are Beam forming signals, you would get a sector any electronically steered beam!
Wow, this is incredible! Thanks for this tear down. Very informative!
It actually says ZIF on the board behind the connector. 4:54
Dave I should have mentioned that that machine was worth less than the sum of its parts... if you're missing a part for your old machine (as many veterinarians are: we don't have the cash to buy the new latest and greatest) you'll pay a lot for them. However, the nerd in me is delighted that you did this ;)
17:00 - "If it ain't broken, don't fix it" ... I think you "fixed" that one really well, it'll never work again :-D
Why aren't these showing up on eevblog.com? I keep missing videos.
Excellent review
i didn't know those head cables have 100 conductors. also how does it not flex and break off i wonder. that's what loose sleeve is for?
Please let me know the type of oil for refill and where I can get bladder for oil thanks
Thanx for the teardown Dave...
Now I really want to see a teardown of the modern usb version. I wonder if they've condensed all those electronics down so small it can fit inside of the scanner itself.
At 6:15 that bar looks like a soldier strip that is terminating (bonding) individual cable shields to the PCB. Do you think the cable was stripped and soldered by hand?
I'm thinking that's the case and yes micro some major or whatever coax for sure been brought before and yes I'm pretty sure that's a bonding strip ever seen all the braids ground Connections in something like a big UPS or other high current high power equipment and others or all the ground straps on RF Equipment ohyeah talk about some rain action.
I wonder if the 3D ones have segments instead of lines... i.e. stacked lines? There's a lot of material science in that gizmo.
No they are motorised they move inside.
great great video dear Dave, i hope to see more of these medical equipment >>>
Would love to get the skinny on building a ~500 kHz (tunable / adjustable within 300 Khz) ultrasonic transducer. One fellow mentioned turning "alarms" into such critters.
Mylar is clear, at least in my experience. The color you showed on the tape is normally Kapton.
Woops! These are worth a penny, haha. I really love your videos, Dave. I've recently binged a few, and it's brought back my passion for electronics. Thanks for sharing, and keep them coming.
I would imagine that newer versions of that probe have more circuitry inside in order to combine the wires in the lead into far fewer channels, perhaps via multiplexing or similar techniques.
There should be an acoustic lens covering elements. This is for first focusing mechanism. Then usg beam is focused on recieve by delay. cut size and shape of elements are directly related with image(signal) quality but also they use elements from same single piece crystal for one probe to achieve same quality on each element. The real channel (tx/rx channel)number for for transmitting and recieving may change by multiplexing. USG with more channel is better generally. Multiplexing cause noise and for USG noise is very very very important because RX signals are very weak and if u amplify u amplify noise too. So frontend quality is very important. Some people compares this machine with todays portable machines but believe me that HDI 3000 is a monster about quality though it is old.
Hello, I have always the query about manufacturing process of the ultrasound probes; how all these thin cables can be placed from end to end with the right order? When they have to solder/connect the cables from one side on the connector to the other side of the probe PCB how they know each cable in which pin it has to be soldered (for example pin 1 from connector's PCB must to be soldered at pin 1 at porbe's PCB). All these tiny cables have the same color and there so many… maybe it is stupid what I will say… but there is a possibility that each cable doesn’t have to be soldered at a specific place between two ends and all the cables soldered randomly and the software will re-position in right order all the image sectors?
he is so happy
Does that mean an element transmits in pulses and then receives using the same element, or the element has a trans and receive in it, or is there a large trans element behind and the small elements receive.
Using the same piezoceramic bar for both tx-ing and rx-ing is the most common practice.
Thermocouple? You mean that orange LED that left in casing? =)
I was wondering why something with no active components would get hot. Would those wires potentially get hot with the amount of power being sent in?
@@sbalogh53 it does have to produce enough acoustic power to make out anatomical details that are fairly deep.
Easier to solder when they are tracks and not a big plain? Better solderablity
@17:49 lol that's me right there getting all geeky at the ultrasound room lol
Interesting. I learned in this video the plane of the ultrasound signal on the screen goes away from the head of the transmitter. Seems obvious now but I was trying to figure out how they got a top down 2d picture from a single row of transducers!
Dave,
Stitching vias so electrons don't escape ???
Did not understand that concept:)
Cool teardown.
Cheers!!!
50 V at 2 A max is 100 W. How do they power modern probes from battery powered tablet? I assume that you really need this power to penetrate body tissues.
Some modern ultrasound imagers actually have a rotating element inside a translucent spherical head. Not a clue what's in there.
thank you for this awesome video , i'm working on ATL HDI 1500 Ultrasound system , i'm trying to update the pc motherboard , but i have some problem with the old connectors , do you have any advice ?!
10:33 Geordi La Forge eat your heart-out.
sir can you describe how does ct scan unit flat panel detector work , how does it transmit analog image instantly to a digital image
Those connectors are garbage, really bad. THey are used a lot on ATE's for testing aircraft equipment. I have been working and troubleshooting them a lot. They have so many bad connections all the time and resulted in so many false errors on DUT.
was wondering if it would be too difficult to remove a single element from the array
what material do you use for the acoustic lens rubber replacement?
Loved and enjoyed your video really. Thanks
Well done!
What would be the pitch of the pins out the transducer ?
I'm curious what would happen if you simply held one of these near an object such as a pop can if it would show something on screen.
I was always curious, what's inside of an ultrasound probe. Well, a nice connector :) - now USB version is available too
It's probably not too hard to make those connections. All the individual wires just run through parallel guides before they put those brown kapton pieces on. Then after that it gets braided and the insulation is put on
Wait, was this a 260-pins connector?
That's an awesome video I'm watching it from Pakistan and i work on ultrasound machines it's really a nice video kindly make another videos like this and tell more about element fittings ,, but it's a nice & amazing practical video.
Seems like potentially a LOT of power across that head. What was it, like 300 watt max? I know the entire point is to convert that energy into ultrasound, but sure those heads get super hot, right?
I believe it was a 199 date code..
How are they do it today?
Any idea how much power is dissipated in that probe? Doesn't look like much heatsinking at the rear of the array, if any.
Don't know, but likely not a lot, as there is very little thermal mass there. And modern one in the same model work from USB-C, so it's more like 10's of watts than 100's.
sir if you can please give a link to down load that pdf file describe multiple elemant transducers