EEVblog Electronics Community Forum
EEVblog => EEVblog Specific => Topic started by: EEVblog on October 14, 2014, 10:52:22 am
-
Mailbag Monday.
Altium Circuit Maker:
http://circuitmaker.com/ (http://circuitmaker.com/)
Professional USB 2.0 Protocol Analyser:
https://www.kickstarter.com/projects/itic/professional-usb-20-ls-fs-hs-protocol-analyzer-mod (https://www.kickstarter.com/projects/itic/professional-usb-20-ls-fs-hs-protocol-analyzer-mod)
from International Test Instruments:
http://www.internationaltestinstruments.com (http://www.internationaltestinstruments.com)
Cypress USB Microcontroller: http://www.cypress.com/?rID=38801 (http://www.cypress.com/?rID=38801)
http://www.geppettoelectronics.com/ (http://www.geppettoelectronics.com/)
EEVblog #673 - Mailbag (https://www.youtube.com/watch?v=4L6JZ5g4Wa8#ws)
-
That checkpoint board looks interesting: There seems to be some rf power stuff on the right side. Maybe this board drives those anti-theft protection things found at the exits of department stores or some other long range rfid antennas.
It is probably impossible to make it do something without a lot of other stuff connected to all the connectors, but it could be interesting to probe the outputs.
-
Multi-hour Mailbag episodes are fine with me, love the 2-5min mini teardowns.
-
Its spring in .au, and for the first time I hear a hum on an eevblog video. Air con, maybe? :)
-
Its spring in .au, and for the first time I hear a hum on an eevblog video. Air con, maybe? :)
Yep, next doors aircon. I turn mine off during shooting.
-
If anyone's curious, that clock had the "whacky" firmware in it - it ticks once a second, but on a random tenth-of-a-second.
-
I have laying around some solar cells from all stages of the production line. From ingot-chunks (lumps of poly-silicium) over thin sheets of silicium cut with a wire-saw to the finished cell.
If you like, I'll send you a sample of every production state and some with different production errors on it.
The lumps make nice paperweights ^-^
AmaG
-
If anyone's curious, that clock had the "whacky" firmware in it - it ticks once a second, but on a random tenth-of-a-second.
A while ago someone did a much more subtle clock hack - it gradually sped up in the half-hour before lunchtime and slowed down for the next hour!
-
The UT513 at 500v has full scale of 20G and only 10% accuracy up there. The 20G resistor was probably more than 20G or the UT513 thought it was. OL is just a not very good over range indication.
It showed OL on the 10G resistor till the voltage got high enough to produce the minimum measurable current.
-
Dave you got set up to fail with that solar cell, it was VERY POORLY packed. You dont pad the shit out of things that might shatter when bend, you pack them in something STURDY first (like a sheet of metal for example,or a plank of wood), not cardboard and bubble wrap :/
That checkpoint board looks interesting: There seems to be some rf power stuff on the right side. Maybe this board drives those anti-theft protection things found at the exits of department stores or some other long range rfid antennas.
I dont think its rf, it looks like Power Over whatever_serial_interfaces_they_used so all the sensors hanging off this thing dont need separate supplies
ps: not long enough!
-
Dave, you might consider giving our Hungarian friend a free uRuler for destroying his precious solar panel! ;)
Just to show the world what a great guy you are! :-DD
-
I dont think its rf, it looks like Power Over whatever_serial_interfaces_they_used so all the sensors hanging off this thing dont need separate supplies
At least the board transmits something at a higher power level and processes the received signal: On the right side is a frequency jumper and a TXGND label. The parts to the left seem to be a analogue RX chain (GAIN jumper, and some Analog Devices parts, maybe opamps and adc).
Those ferrite parts on the top side look similar to those used in rf power amplifiers (for push-pull amplifiers or differential-single ended conversion).
Dave, maybe you could take some high-res fotos of the part numbers?
-
If anyone's curious, that clock had the "whacky" firmware in it - it ticks once a second, but on a random tenth-of-a-second.
A while ago someone did a much more subtle clock hack - it gradually sped up in the half-hour before lunchtime and slowed down for the next hour!
One of the available firmwares for mine is one that will run 10% fast for 12 hours, then 10% slow for another 12. Makes the days fly by.
The problem you run into if you want to get much more sophisticated is that the controller has no idea where the actual hands are pointing. All it can do is make the second hand move.
-
Yes the right hand side is for rf. The board was installed in one of those 6ft anti-theft gates and had two coils wrapped around the gate structure vertically. There where some things dangling off the other side that where chopped off, presumably the checkout scanner and another section of gates.
The controller must be their in-house hammer for every problem as its massivly overpowered for such a task. I added up the blackfin, memory and fpga and was over $200/qty100
-
:D you got a Vetinary clock! Awesome!
As to that resistor, I'm guessing it's rated MUCH higher than 500V - these kinds of resistors were/are used for amongst other things for high voltage dividers. I'd look for the error somewhere else.
-
:D you got a Vetinary clock! Awesome!
Actually, while a Vetinari firmware is available for that controller, that one isn't running it. The Vetinari firmware ticks ostensibly normally most of the time, but actually ticks imperceptibly slow. Once it has "gathered" enough time, it does a special "double tick" to catch up. It winds up doing that approximately every 30 seconds or so (there's random distribution to make it unpredictable).
The firmware Dave got just ticks once a second, but on a random tenth-of-a-second, so it's an arhythmic clock.
-
i liked the video,
i think it better now that it more about opening mails, and less about reviewing them.
so you open the item, do a quick 2 minutes review, and go to the another one.
and if you need to do a full review, you can do it on teardown tuesday.
Also, maybe you can do Mailbag Monday "The Marathon" with 1 hour or so video,
so you can open everything :)
And about the new µRuler, i would like to buy one.
-
The UT513 at 500v has full scale of 20G and only 10% accuracy up there. The 20G resistor was probably more than 20G or the UT513 thought it was. OL is just a not very good over range indication.
It showed OL on the 10G resistor till the voltage got high enough to produce the minimum measurable current.
A real "Megger" 5kV tester measures about 100G at 500V.
Also, I have seen some Welwyn resistors of about that size that were good for 100G 15kV element voltage (when in oil or potted)
-
Mailbag Monday.
Professional USB 2.0 Protocol Analyser:
https://www.kickstarter.com/projects/itic/professional-usb-20-ls-fs-hs-protocol-analyzer-mod (https://www.kickstarter.com/projects/itic/professional-usb-20-ls-fs-hs-protocol-analyzer-mod)
from International Test Instruments:
http://www.internationaltestinstruments.com (http://www.internationaltestinstruments.com)
Thanks Dave! Yes, the Kickstarter campaign was a flop but I have determined KS is simply the wrong forum for such projects. KS is, of course, all about hobby projects and, since the pledge levels were in the "laptop-price", few hobbyists could help out.
Anyways, the PCIe connector you mentioned is for attachment to a debugging board that allows the FX2 firmware to be downloaded and debugged interactively via serial port (SRAM and RS232 line drivers are on the debug board). Also, the debug board routes out a bunch if LVCMOS FPGA pins to MICTOR connectors that allow me to hook up to my Agilent 16700 Logic Analyzer for deep LA capture. Finally, the ground strips along the edges of the board are something i normally use to avoid ESD damage - better to dischange static electricity into the GND plane than into individual components on the board. The top fill is actually 3V3, not GND, so this is why is it is not connected to the edge GND strips.
By the way, I also have a PCI Express Protocol Analyzer available for sale on ITIC's web site. Ans, a USB 3.0 Protocol analyzers is being designed!
Cheers,
/John.
-
The CheckPoint stuff is theft protection device used in stores. It must have a 58 kHz RF loop as antenna attached to be able to detect armed RF tags.
And it's sold in millions as it can be found in almost any store these days.
(http://www.glove-tech.com/wp-content/uploads/2011/05/mono-antennas.jpg)
(http://www.eas-labels.net/EAS/antenna/101123/2_060S5c3.JPG)
-
The CheckPoint stuff is theft protection device used in stores.
Ah, right, of course, Guess I should have googled that.
-
The UT513 at 500v has full scale of 20G and only 10% accuracy up there. The 20G resistor was probably more than 20G or the UT513 thought it was. OL is just a not very good over range indication.
Doh, that explains it.
-
A real "Megger" 5kV tester measures about 100G at 500V.
PASS will sell you Megger's entry level MIT515 for £1865, they will sell you a UT513A for £189.
-
I just happened to be watching Shariar's teardown Calibration etc of a Keithley 220 and that unit appears to contain the same high value wire wound resistors. Skip to around 10:39. I suppose that figures.
https://www.youtube.com/watch?v=pMYK5qoQvYo (https://www.youtube.com/watch?v=pMYK5qoQvYo)
-
Re the ferrous battery correction on the Sanwa VOM...
D'arsonval meter movements can have an external magnetic field and hence the meter accuracy can be degraded with nearby magnetic objects.
If you look at old catalogs of panel mount meters, you can sometimes find two versions, one calibrated for non-magnetic panels and one for magnetic panels. I believe most modern D'arsonval movements are shielded internally, at least the more expensive ones, and hence are suitable for magnetic or non-magnetic panel mount.
If you remember, some dry cells included a steel outer jacket - I think Ray-o-Vac featured that construction in the USA, claiming it reduced damage from leaking cells and I assume you had something similar 'down under."
I believe that Sanwa has intentionally added a ferrous battery holder so that whether a steel jacketed or non-jacketed cell is used, the meter movement always sees the same magnetic environment, with the leakage magnetic field dominated by the ferrous battery holder.
It's possible the battery holder is something more exotic, such as mu-metal, but I suspect it's steel.
It's a nice touch to maintain accuracy where the meter movement is unshielded, or imperfectly shielded.
-
At least that was a good example of why solar panels should be put in a box before bubble wrapping them because you're not as likely to bend the box while struggling with the tape.
-
+1 for the µRuler campaign.
-
Love the 2-Minute Teardowns™.
Also, another +1 for the uRuler campaign!
I wonder why the checkpoint thing is so powerful. To my knowledge, there's a lot of different systems for theft detection RFIDs in frequency ranges between 10kHz and microwave bands, but almost all of them are dumb as a rock and the sender/detector units could easily be built fully analog - that is some serious overkill hardware they have there. And it's not like they manufacture hundreds of it, it's probably more like hundreds of thousands, so why not optimize for cost? Am I underestimating modern anti-theft RFIDs?
-
Could be for a library or other place that rents stuff, where you have to check if all leaving items are in fact authorized to leave, as opposed to removing or disabling the tags on purchased items.
-
Re the ferrous battery correction on the Sanwa VOM...
D'arsonval meter movements can have an external magnetic field and hence the meter accuracy can be degraded with nearby magnetic objects.
Interesting, thanks.
-
Love the 2-Minute Teardowns™.
Also, another +1 for the uRuler campaign!
I wonder why the checkpoint thing is so powerful. To my knowledge, there's a lot of different systems for theft detection RFIDs in frequency ranges between 10kHz and microwave bands, but almost all of them are dumb as a rock and the sender/detector units could easily be built fully analog - that is some serious overkill hardware they have there. And it's not like they manufacture hundreds of it, it's probably more like hundreds of thousands, so why not optimize for cost? Am I underestimating modern anti-theft RFIDs?
The simplest anti-theft tag that I'v seen is few metal strips stacked together, so the tags are dumb as a rock. Security gate has transmitting coil that makes tag resonate and second coil detects it. The tags are passive so transmitter must be powerful and receiver sensitive and the gate has to detect different types of tags at different frequencies. Large market might have 10 gates at cash registers so 10 powerful transmitters and 10 sensitive receivers very close together. The gates cant take much space (few inches wide) which limits shape of antennas and shielding. Store owner wont accept many false alarms, once a day is probably too many. Does that hardware still look like overkill? Just imagine what kind of problems signal reflections from store structures must cause (has to work in any shape store that gates are sold to) or gates at next door store made by different manufacturer.
-
Does it bother anyone else that there appears to be a DSO upside down on the shelf behind Dave?
-
Does it bother anyone else that there appears to be a DSO upside down on the shelf behind Dave?
Nah, it looks fine down under.
Since he build that new bench and the new mailbag format that scope is been that way since ever. He knows but I guess its to attract attention for us OCD people :)
-
I wonder why the checkpoint thing is so powerful. To my knowledge, there's a lot of different systems for theft detection RFIDs in frequency ranges between 10kHz and microwave bands, but almost all of them are dumb as a rock and the sender/detector units could easily be built fully analog - that is some serious overkill hardware they have there. And it's not like they manufacture hundreds of it, it's probably more like hundreds of thousands, so why not optimize for cost? Am I underestimating modern anti-theft RFIDs?
There is a couple of different security types, not only RFID (That needs a cpu, because it uses digital communication).
Some of the old analog types needs a rather powerful sending coil (A CRT within a few meters of a gate would be affected) and the receiver would either be a couple of analog filters or a DSP. I have seen this type used in libraries, because the security can be activated again.
Another type of security is a resonance circuit, where the gate scans for the resonance circuit. These tags are either removed or burned out, before the goods can be moved through the gate.
-
Love the 2-Minute Teardowns™.
Also, another +1 for the uRuler campaign!
I wonder why the checkpoint thing is so powerful. To my knowledge, there's a lot of different systems for theft detection RFIDs in frequency ranges between 10kHz and microwave bands, but almost all of them are dumb as a rock and the sender/detector units could easily be built fully analog - that is some serious overkill hardware they have there. And it's not like they manufacture hundreds of it, it's probably more like hundreds of thousands, so why not optimize for cost? Am I underestimating modern anti-theft RFIDs?
The simplest anti-theft tag that I'v seen is few metal strips stacked together, so the tags are dumb as a rock. Security gate has transmitting coil that makes tag resonate and second coil detects it. The tags are passive so transmitter must be powerful and receiver sensitive and the gate has to detect different types of tags at different frequencies. Large market might have 10 gates at cash registers so 10 powerful transmitters and 10 sensitive receivers very close together. The gates cant take much space (few inches wide) which limits shape of antennas and shielding. Store owner wont accept many false alarms, once a day is probably too many. Does that hardware still look like overkill? Just imagine what kind of problems signal reflections from store structures must cause (has to work in any shape store that gates are sold to) or gates at next door store made by different manufacturer.
Hmm...I would have thought the amount of power these resonant systems put out isn't all that great. Because of the size of the coil(usually something in the area of a coil wound a couple of times around the entire detector, maybe 1.2-1.6m tall) and the very short range it's designed for(distance sender-detector ~1m) they should be able to get away with very little power, shouldn't they? Would also avoid all the problems involved with stuff like interference from the shop next door...
-
The simple strips are actually a very complex system of a magnetic strip sandwiched between a mumetal strip, that is magnetised and when the external field is coupled with it it saturates and makes a harmonic of the excitation frequency, which is detected by the receive coil and then is processed by a DSP to get the signal out. Degauss the strip and it no longer generates the harmonics, and it can then have a magnetic field applied to it to reset it to operate again.
The drivers use a large strip aluminium coil, typically 1in wide strip, wound into a coil spaced with foam tape. Around 20 turns about 2ft wide and 5ft high, and this is driven with a high power audio signal of around 15kHz, generated by the DSP and amplified with a class D amplifier, and then filtered by brute force LC filters to drive with a very low distortion sine wave, low distortion as you are detecting harmonics in the presence of a very strong excitation. The receive side uses a few bandpass filters that remove the transmit signal and amplify to feed the DSP that then does the heavy lifting. The systems switch between each coil acting as transmitter and receiver alternately, and use 2 or more frequencies at the same time to provide tbetter discrimination. Processor there is pretty simple, mostly generating the control signals, making the PWM transmit drive and doing comms to the built in counters and displays, along with driving external communications.
Power is around 150VA, with massive amounts of capacitors to provide very low ripple on the supply rails.
Yes, I did take one apart............. ;) Still got a lot of the boards knocking around.
However the one Dave got is more likely a controller for an access gate, like you get in large buildings where you clock through gated with a pass, and where you need a card to use things like lifts and to enter offices. Thus the need for a Lotus database back end, and the smaller contactless card driver, along with the motor drive systems for things like gates, semaphore lights and to interface to things like door releases and door position detectors and most likely also interfaces for alarm inputs and outputs, and with most likely inputs for occupancy sensors, temperature and light sensors and more outputs to control lighting, air damper valves and possibly even automatic blinds or window shades. Having all that on board having such a high power processor is starting to look like it might be needed.
-
Power is around 150VA, with massive amounts of capacitors to provide very low ripple on the supply rails.
Holy shitsnacks, that's a lot more power then I thought...any chance you'd do a video on or post some photos of the boards out of the system you took apart?
-
As to that resistor, I'm guessing it's rated MUCH higher than 500V - these kinds of resistors were/are used for amongst other things for high voltage dividers. I'd look for the error somewhere else.
Yes, this are High-Voltage Resistors made by Ohmcraft
http://www.ohmcraft.com/ (http://www.ohmcraft.com/)
Series CR KOBRA.
The big one is made for 32kV and 12W of max. Power,
the small one is made for 16kV and 7W max. Power.
BTW, 500V @ 20GOhm makes 12,5µW Power.
-
Dave, I happened to be listening to some Amp Hour episodes on a flight overseas and one of them mentioned Chris trying out layout on a plane. Your thought was that you're so tied to the internet that it's not worth bothering. If that's the case, what is the inherent downside of a program using the cloud? Obviously a bad program in other respects will ruin the experience, but given two identical programs where one is "cloud based" and the other stores files to your local hard disk is there any reason not to like this? Seems a bit contradictory.
-
Dave, I happened to be listening to some Amp Hour episodes on a flight overseas and one of them mentioned Chris trying out layout on a plane. Your thought was that you're so tied to the internet that it's not worth bothering. If that's the case, what is the inherent downside of a program using the cloud? Obviously a bad program in other respects will ruin the experience, but given two identical programs where one is "cloud based" and the other stores files to your local hard disk is there any reason not to like this? Seems a bit contradictory.
its lovely, especially if you want chinese/competition with friends at nsa/cia/Altium to have your company design before your company does
its ok as long as you are also ok with mounting Dropcams (cloud based webcam) in your bathroom
-
Power is around 150VA, with massive amounts of capacitors to provide very low ripple on the supply rails.
Holy shitsnacks, that's a lot more power then I thought...any chance you'd do a video on or post some photos of the boards out of the system you took apart?
Found the only board still around, I stripped the caps out of it though, to reuse on a UPS to lower battery ESR. This was the power amplifier board, with the 2 class D power amplifiers on the left. Another board had the massive LC output filters and the huge film capacitors that were used to make a pure sine wave drive for the coils.
-
At least that was a good example of why solar panels should be put in a box before bubble wrapping them because you're not as likely to bend the box while struggling with the tape.
Reading the letter first would have helped. As would cutting the tape. Speaking from experience, and not hindsight.
I was starting to think that amount of bubble wrap was a joke.
Sorry, I normally try to restrain myself, but:
The amount of bubble wrap was a joke. That was just an absolutely stupid way to package that cell. What breaks a solar cell? Even moderate bending forces. What does nothing to dissipate bending forces? Bubble wrap. What is likely to cause lots of bending forces during opening? Craploads of tape that achieves nothing useful. What protects a solar cell? A nice sturdy box and a little bit of foam/bubble sheet with just enough tape to hold the lid shut. If you think for half a second about what breaks a solar cell, and were to go out of your way to package it in the worst possible way, the sender of that package succeeded. Dave should feel no guilt whatsoever, this is a valuable lesson for people that think that bubble wrap is just a magic potion that doesn't require thought to use. Facepalm at anyone who defends the sender.