Sampling front ends allowed excellent oscilloscope overdrive performance but as you point out, they are no longer produced.
A vast majority of oscilloscope applications do not require more sensitivity, more resolution, or better overdrive recovery. The applications that do are better served with a different test or more specialized instrument. To give a well known example, is it even now possible with 2010 technology to build an oscilloscope useful for measuring the settling time of a high performance operational amplifier? If you can not do that, you sure will not be able to build one for testing a 16 bit DAC directly.
Even testing more than 16 bits of audio ADC performance is tough. How many audio oscillators have part per billion distortion performance? At least you can design and build a part per billion audio source fairly easily if needed.
With all of that said, I would agree that oscilloscopes have fallen into the hole where marketing is more important than engineering.
I would go for some basic electronics building blocks like a folded cascode differential amplifier with current mirror load and differential to single ended conversion. If that is too large just leave out the folded cascode. A Gilbert cell would be another good tattoo.
I believe the reason is so that if someone says they're innocent and is convicted of a crime they can't be convicted of perjury as well.
Actually, they can. You have the right to remain silent but the US Supreme Court resolved the circuit split on the exculpatory "no" doctrine by ruling that an exculpatory no is not protected in the case Brogan v. United States. This is just another reason not to answer questions from law enforcement.
You could weld a valve onto the top, but if you try to close it, the pressure will seek relief elsewhere. If you get really, really lucky, it just blows out the weld and rejects the valve.
This doesn't make sense to me. The key failed element of the blow out preventer was was the "blind ram" that was supposed to pinch the pipe and seal it. They spent the first couple days trying to activate the failed half of the pincher, first by activating it manually and finally by attaching a high-pressure hydraulic pump with an ROV.
If what you're saying is true, that would have simply split the pipe instead of closing off the leak. I seriously doubt they would have even installed the BOP if the most likely outcome of its use was disaster.
The blind ram is suppose to be used while there is still heavy mud in the well and before the pressure becomes too great. That said, I do not know why they were trying to activate it two days later if the well structure was insufficient to contain the pressure.
As RAM gets ever-larger, densities get ever-greater, and the energy requirements for corruption get ever-smaller, the amount of error-correction needed is going to increase.
It is not quite that simple. The struggle to make smaller DRAM storage capacitors without lowering the stored charge means that the charge stored per volume has gone up while the volume has gotten smaller making the cell a smaller target and more resistant. Any given ionization event ends up spreading the same total charge over a greater number of increasingly resistant DRAM cells. The result is that the susceptibility to radiation induced soft errors has leveled off or even gone down per bit of storage in the last couple of generations while the number of bits installed in each system has gone up thereby making up for any improvement.
In contrast to the DRAM arrays themselves, the DRAM sense amplifiers and associated logic have become more susceptible such that they increasingly contribute to the soft error rate. For similar reasons, SRAM cache is an order of magnitude more sensitive to radiation induced soft errors than the DRAM arrays but it have been protected by ECC almost since they were integrated with the CPU.
If there are only 4 levels then it makes much more sense to use comparators.
A comparator is a 1 bit A to D converter. 4 comparators make a 2 bit A to D converter.
Should one require the flexibility of being able to adjust the reference voltage, they could utilize a DtoA as a reference. DtoA circuitry is much simpler/faster then AtoD circuitry.
Internally, something like this is already done. During writes, a reference cell or cells are written which are used during reads to adjust or generate the reference voltage for the comparisons.
The best non-isolated converters can get up to 98% efficiency but only at close to maximum power where the percentage of quiescent power is lowest. You typically see non-isolated converters used in either point of load applications like high current CPU power supplies or in micropower applications where the quiescent power loss can be lowered by operating in bursts. For the isolated designs necessary in off line use, 90% maximum efficiency is possible but only over relatively narrow operating ranges unless a lot of complexity is introduced.
Lower power supply efficiency at lower system power is not usually a problem for heat dissipation limited designs like the Mac Mini. The lower total power draw will more than make up for any decrease in power supply efficiency.
Well, assuming it draws around the 1W power that the laser puts out, and assuming a modest 2200mAH 3.3v battery, it would run for about 7 hours continually. And this thing is BIG (like, maglite big), so it could probably stand to pack more batteries than that!
From the laser diode specifications, it draws about 1 amp at 5 volts.
Best reply so far but no go. We have directional dish type UHF antennas.
Directional dish type UHF antennas? Dish antennas at UHF frequencies would be pretty big. Could you link to an example?
I have seen the type of setup gnu-sucks is describing used for narrowband signals where multipath was not an issue but never for television signals. The idea is to use two antennas,each aligned for the best signal from each station, and then subtract the interfering station from the desired station via phase and amplitude matching.
Yes but you are only telling half of the story here. The lower the frequency the lower the theoretical bandwidth. So yes 700mhz is great for long range communications but it is horrible for high data transmission rates.
It is just easier to use wider bandwidths at higher frequencies. As the bandwidth becomes a larger percentage of the center frequency, antenna and RF design become more complex. A 20MHz channel at 700MHz is the same as a 20MHz channel at 2.4GHz as far as maximum data throughput goes.
Apparently a lot of cars simply don't have the power to make it up, and manual drivers can't do it if they're used to flat land or relatively tame hills. Everyone on that road drove 4wd SUV's.
I do not know why an automatic would have more torque than a manual but I guess the torque converter would allow the engine to operate further up on its torque curve for a given slow speed. With a manual you would have to feather the clutch to get the same effect which is rather hard on it. The Pinto I owned was limited to slopes where the gas did not flow out of the fuel bowl directly into the venturi.
Really, the only reason to drive a manual these days is because it's more fun.
Manual transmissions have the advantage of being less likely to completely fail without warning. I have changed two clutches and one main bearing and none of those events left me stranded.
They also leave you the option of push starting which I have had to do a few times.
I do not remember the details but the Z80 clocking scheme and bus multiplexing were a little weird such that it took about double the clock rate to equal the performance of its contemporaries.
The 68000 based 8 bit external bus equivalent (68008) did not become generally available until well after the 8088. Besides legacy CP/M application ports, using the 8088 allowed access to all of the existing cheap 8080 8 bit peripheral chips as well as narrower main memory.
Did Motorola even have a second source for the 68000 series at the time? When I dealt with them 10 years later, I ended up turning down their embedded processors just because of availability issues in spite of their great development support. I still have a pair of 68HC24 PLCCs for development work that took a year to acquire.
The problem you run into with netbooks and such is that LiIon has about double the energy density of NiMH.
In the best case, lithium chemistry cells have twice the energy density by weight but by volume it is a lot closer. They also have a high cost considering their relatively short life span. A standardized lithium battery size that was ubiquitous and user replaceable would be nice. I would trade any extra energy density for the long life and lower cost of lithium-iron-phosphate.
It is not as simple as adjusting the "seek speed". When the head it tracking a specific track, the dynamic response is limited by the small signal response of the head/actuator/servo mechanism. When the head is slewing across tracks at maximum velocity, its dynamic response is limited by the large signal response. In the later case, vibration probably has no effect at all. It's when vibration interferes with the small signal response as the head reaches the destination track or is tracking that head settling time in increased or the head is pushed out of alignment with the track.
The small signal response is primarily limited by the flexing and moment of inertia of the head and arm. You can make the head lighter and the arm stiffer but there are limits to this depending on the materials and engineering. The best option is to either make the arm shorter by making a smaller drive or add a second actuator closer to the head like the newer WD drives do. Both raise the resonate frequency of the arm and allow higher small signal bandwidth and open loop gain which allows more suppression of external vibration.
That is the system I try to use for all portable electronics. I much prefer swapping good standard AA or AAA cells to fixed lithium cells. My optical mouse goes for more than 2 months and swapping takes about 30 seconds. The only item I regularly use which does not use standard cells is my mp3 player although it does have a user replaceable lithium battery. I am waiting on a good AA using netbook but do not have high hopes.
As far as low discharge NiMH cells, I have had good results from Sanyo's Eneloop cells. Their old low discharge NiCd cells are what I used to use in high reliability applications.
I'm not sure what you guys are doing but both my last 2 x AA wireless keyboard and my current 2 x AAA wireless keyboard last months between charges. I used normal NiMH AAs for the earlier one and now use low discharge AAAs for the new one. I've never actually run down the 2 x AAA one yet. My 2 x AA low discharge NiMH mouse goes about 2 months.
I do not buy any new NiMH cells that are not low discharge. The earlier cells were all pretty poor replacements for good NiCd cells.
One of the side effects of a stateful firewall (which includes NAT since it requires state to be kept) is that routing can be significantly faster. Once an IP connection is added to the state table, the routing table no longer has to be traversed for each packet of that connection. For this reason, many routers maintain IP connection state just for performance reasons.
Slashdot Mirror
Methinks you doth protest too much. :)
Sampling front ends allowed excellent oscilloscope overdrive performance but as you point out, they are no longer produced.
A vast majority of oscilloscope applications do not require more sensitivity, more resolution, or better overdrive recovery. The applications that do are better served with a different test or more specialized instrument. To give a well known example, is it even now possible with 2010 technology to build an oscilloscope useful for measuring the settling time of a high performance operational amplifier? If you can not do that, you sure will not be able to build one for testing a 16 bit DAC directly.
Even testing more than 16 bits of audio ADC performance is tough. How many audio oscillators have part per billion distortion performance? At least you can design and build a part per billion audio source fairly easily if needed.
With all of that said, I would agree that oscilloscopes have fallen into the hole where marketing is more important than engineering.
I would go for some basic electronics building blocks like a folded cascode differential amplifier with current mirror load and differential to single ended conversion. If that is too large just leave out the folded cascode. A Gilbert cell would be another good tattoo.
Actually, they can. You have the right to remain silent but the US Supreme Court resolved the circuit split on the exculpatory "no" doctrine by ruling that an exculpatory no is not protected in the case Brogan v. United States. This is just another reason not to answer questions from law enforcement.
http://www.tjtaxlaw.com/tjn19980126.htm
The blind ram is suppose to be used while there is still heavy mud in the well and before the pressure becomes too great. That said, I do not know why they were trying to activate it two days later if the well structure was insufficient to contain the pressure.
It is not quite that simple. The struggle to make smaller DRAM storage capacitors without lowering the stored charge means that the charge stored per volume has gone up while the volume has gotten smaller making the cell a smaller target and more resistant. Any given ionization event ends up spreading the same total charge over a greater number of increasingly resistant DRAM cells. The result is that the susceptibility to radiation induced soft errors has leveled off or even gone down per bit of storage in the last couple of generations while the number of bits installed in each system has gone up thereby making up for any improvement.
In contrast to the DRAM arrays themselves, the DRAM sense amplifiers and associated logic have become more susceptible such that they increasingly contribute to the soft error rate. For similar reasons, SRAM cache is an order of magnitude more sensitive to radiation induced soft errors than the DRAM arrays but it have been protected by ECC almost since they were integrated with the CPU.
If there are only 4 levels then it makes much more sense to use comparators.
A comparator is a 1 bit A to D converter. 4 comparators make a 2 bit A to D converter.
Internally, something like this is already done. During writes, a reference cell or cells are written which are used during reads to adjust or generate the reference voltage for the comparisons.
The best non-isolated converters can get up to 98% efficiency but only at close to maximum power where the percentage of quiescent power is lowest. You typically see non-isolated converters used in either point of load applications like high current CPU power supplies or in micropower applications where the quiescent power loss can be lowered by operating in bursts. For the isolated designs necessary in off line use, 90% maximum efficiency is possible but only over relatively narrow operating ranges unless a lot of complexity is introduced.
Lower power supply efficiency at lower system power is not usually a problem for heat dissipation limited designs like the Mac Mini. The lower total power draw will more than make up for any decrease in power supply efficiency.
From the laser diode specifications, it draws about 1 amp at 5 volts.
I draws about 1 amp at 5 volts so 20% efficient.
Directional dish type UHF antennas? Dish antennas at UHF frequencies would be pretty big. Could you link to an example?
I have seen the type of setup gnu-sucks is describing used for narrowband signals where multipath was not an issue but never for television signals. The idea is to use two antennas ,each aligned for the best signal from each station, and then subtract the interfering station from the desired station via phase and amplitude matching.
It is just easier to use wider bandwidths at higher frequencies. As the bandwidth becomes a larger percentage of the center frequency, antenna and RF design become more complex. A 20MHz channel at 700MHz is the same as a 20MHz channel at 2.4GHz as far as maximum data throughput goes.
I am sure they can just buy more electrical power from Arizona. Oh, wait . . .
I do not know why an automatic would have more torque than a manual but I guess the torque converter would allow the engine to operate further up on its torque curve for a given slow speed. With a manual you would have to feather the clutch to get the same effect which is rather hard on it. The Pinto I owned was limited to slopes where the gas did not flow out of the fuel bowl directly into the venturi.
Manual transmissions have the advantage of being less likely to completely fail without warning. I have changed two clutches and one main bearing and none of those events left me stranded.
They also leave you the option of push starting which I have had to do a few times.
Do you mean like an atomic bomb pumped x-ray or gamma ray laser?
I just remember that there was something weird about the Z80 clocking which led to it running slower than the clock speed itself would indicate.
I do not remember the details but the Z80 clocking scheme and bus multiplexing were a little weird such that it took about double the clock rate to equal the performance of its contemporaries.
The 68000 based 8 bit external bus equivalent (68008) did not become generally available until well after the 8088. Besides legacy CP/M application ports, using the 8088 allowed access to all of the existing cheap 8080 8 bit peripheral chips as well as narrower main memory.
Did Motorola even have a second source for the 68000 series at the time? When I dealt with them 10 years later, I ended up turning down their embedded processors just because of availability issues in spite of their great development support. I still have a pair of 68HC24 PLCCs for development work that took a year to acquire.
In the best case, lithium chemistry cells have twice the energy density by weight but by volume it is a lot closer. They also have a high cost considering their relatively short life span. A standardized lithium battery size that was ubiquitous and user replaceable would be nice. I would trade any extra energy density for the long life and lower cost of lithium-iron-phosphate.
It is not as simple as adjusting the "seek speed". When the head it tracking a specific track, the dynamic response is limited by the small signal response of the head/actuator/servo mechanism. When the head is slewing across tracks at maximum velocity, its dynamic response is limited by the large signal response. In the later case, vibration probably has no effect at all. It's when vibration interferes with the small signal response as the head reaches the destination track or is tracking that head settling time in increased or the head is pushed out of alignment with the track.
The small signal response is primarily limited by the flexing and moment of inertia of the head and arm. You can make the head lighter and the arm stiffer but there are limits to this depending on the materials and engineering. The best option is to either make the arm shorter by making a smaller drive or add a second actuator closer to the head like the newer WD drives do. Both raise the resonate frequency of the arm and allow higher small signal bandwidth and open loop gain which allows more suppression of external vibration.
That is the system I try to use for all portable electronics. I much prefer swapping good standard AA or AAA cells to fixed lithium cells. My optical mouse goes for more than 2 months and swapping takes about 30 seconds. The only item I regularly use which does not use standard cells is my mp3 player although it does have a user replaceable lithium battery. I am waiting on a good AA using netbook but do not have high hopes.
As far as low discharge NiMH cells, I have had good results from Sanyo's Eneloop cells. Their old low discharge NiCd cells are what I used to use in high reliability applications.
I'm not sure what you guys are doing but both my last 2 x AA wireless keyboard and my current 2 x AAA wireless keyboard last months between charges. I used normal NiMH AAs for the earlier one and now use low discharge AAAs for the new one. I've never actually run down the 2 x AAA one yet. My 2 x AA low discharge NiMH mouse goes about 2 months.
I do not buy any new NiMH cells that are not low discharge. The earlier cells were all pretty poor replacements for good NiCd cells.
I still use a Pentium II based router you insensitive clod!
I want to hear the Leroy Jenkins version of this story.
One of the side effects of a stateful firewall (which includes NAT since it requires state to be kept) is that routing can be significantly faster. Once an IP connection is added to the state table, the routing table no longer has to be traversed for each packet of that connection. For this reason, many routers maintain IP connection state just for performance reasons.