I mean something that could record continuously like a soundcard's line-in port, but much faster, like at the speed of an oscilloscope. I have a Picoscope and it has 2 modes, streaming mode, and block mode. In streaming mode it can only go up to 1MSPS, but in block mode it can go up to something like 20MSPS. Unfortunately block mode doesn't capture multiple consecutive blocks because it waits for the next trigger, and even when using block mode using no trigger (freerunning) trigger method it needs to wait for the last trigger reset before it automatically fires its next trigger. This leaves a gap in the data.
There are also standalone digital oscilloscopes that exist with USB ports, but those ports are only for saving a screencap of the oscilloscope screen and transferring it to your PC. That's because an oscilloscope isn't really built for capturing large ammounts of a signal, streamed to your PC, but rather about only showing small sections of a waveform for analysis.
I need something with the constant streaming of a soundcard, but the wide bandwidth that most oscilloscopes can only use in block mode (like a sample rate of 20MSPS to 50MSPS), something that will generate up to 6 gigabytes per minute of data (assuming the data is 2 bytes per sample due to a ADC having a bitdepth greater than 8 bits per sample) on the harddrive, for the complete storage of the entire length of a several-MHz wide signal that can last several minutes in duration (the harddrive would be a dedicated 100GB harddrive or larger for capturing the signal, preferably a solidstate drive for fast writing speeds). This will allow further processing of that signal using software. Hopefully it will also have a better bitdepth than 8 bits per sample, like maybe 10, 12, or even 14 bits per sample (zero-padded to 16 bits per sample if its native bitdepth is less than 16 bits per sample, so that each sample will end on a byte boundary for easy processing using ordinary computer software).
I was very annoyed a couple of years back when I went searching for something similar. Literally just an ADC that can stream data over USB2.0/3.0 to the computer in any sane, simple
continuous way. Don't give a damn about triggering, I can do that in software after the capture.
Very rough notes of what I recall:
National instruments stuff: can do it all, but locked to their software. Lab I was at was doing 24 hour or more captures at low rates (10's of KHz IIRC? Many channels DC included) and the Windows software would randomly stop capturing or blip with various errors in the night. Tried increasing buffer sizes and other things, no dice. Support told us it was our problem and/or the nature of Windows anyway, can't get around it.
Bitscope: might be able to handle continuous? Their website is very complex to navigate and their specs page is dead (google sheets). Sydney based.
Cheap holteks: Could these do continuous capture? Some models?
Cheap rigol scopes with USB: Can't do continuous.
Sigrok software project: interesting to look at their
list of supported devices but IIRC no info about whether they can continuous capture or not.
streaming mode, and block mode. In streaming mode it can only go up to 1MSPS,
Hello,
this is only true for the (older) entry level models. (PS 2xxxA)
the newer USB3 models (PS5xxxD, PS4xxxA) can go up to 15-20 MSPS in streaming mode.
with best regards
Andreas
Hello,
from datasheet for PicoScope® 5000D Series
"
Maximum sampling rate
(continuous USB streaming into PC
memory)[5]
USB 3, using PicoScope 6: 15 to 20 MS/s
USB 3, using PicoSDK: 125 MS/s (8-bit mode) or 62.5 MS/s (12 to 16-bit modes)
USB 2, using PicoScope 6: 8 to 10 MS/s
USB 2, using PicoSDK: ~30 MS/s (8-bit mode) or ~15 MS/s (12 to 16-bit modes)
[5]Shared between enabled channels, PC dependent, available sample rates vary by resolution.
"
So perhaps a PicoScope 5000D Series scope is a deeper view worth.
Best regards
egonotto
I would send an E-mail to support@picotech.com and/or to sales@picotech.com and ask them for help.
If it is possible, they will help you, with exact data..
They work real hard to help the customers.
this is only true for the (older) entry level models. (PS 2xxxA)
the newer USB3 models (PS5xxxD, PS4xxxA) can go up to 15-20 MSPS in streaming mode.
with best regards
Andreas
Unfortunately those are above my price range right now (over $1000). And I know why too. They still are designed as scopes, with streaming being an "extra feature". As they are scopes (which are far more complicated to manufacture) their price matches that usage intent, and so they cost tons. However the device I described in my opening post here would actually be very cheap to make and sell (could be made for about $200 and sold for about $500, as a bare PCB and for maybe $20 more to have the board mounted in a case). This is because it would literally have just one analog input and just 4 ICs (plus supporting electronics), and these would be high-input-impedance wide-band opamp, a wide-band 2-supply 12bit analog to digital converter, USB interface IC, and microcontroller (to take USB commands to do things like setting the input voltage range on the ADC, and also capturing the output from the ADC and sending it to the USB interface controller for conversion into a USB data stream). The device would depend on a 4-output-pin power supply to give it power, These pins are ground, +12v for opamp +power and ADC +voltage reference, -12v for opamp -power and ADC -voltage reference, +5v for all digital ICs including the digital portion of the ADC.
You can try Chinese SDR receivers like DRAGONFLY RX666:
https://www.aliexpress.com/item/4001186751605.htmlhttps://www.aliexpress.com/item/1005001710002655.htmlBut note, it doesn't have FPGA for hardware frequency down conversion, so your PC should be fast enough to process very fast stream obtained directly from ADC in realtime. Top level desktop i7 CPU is recommended.
Technically this receiver consists of 16-bit high speed ADC LTC2208 and USB 3.0 controller to transfer ADC RAW stream to PC.
This is a clone of a popular ham radio SDR receiver
BBRF103 from IK1XPV?
You can try Chinese SDR receivers like DRAGONFLY RX666:
https://www.aliexpress.com/item/4001186751605.html
But note, it doesn't have FPGA for hardware frequency down conversion, so your PC should be fast enough to process very fast stream obtained directly from ADC in realtime. Top level i7 CPU is recommended.
Technically this receiver consists of 16-bit high speed ADC LTC2208 and USB 3.0 controller to transfer ADC RAW stream to PC.
But that says SDR. That's a radio receiver. I don't need it for a radio receiver (other than receiving base-band frequency signals). I don't want an SDR. That includes a tuner IC between the input connector and the ADC, which means you are not getting a baseband signal to the ADC.
Also it has a DC blocking capacitor. Notice the specs don't say DC to 1800MHz. Instead it says 1kHz to 1800MHz.
Also no mention of control software. How do you even operate it? Where do you get drivers? Where do you get the software to control it? Even if it has a driver CD that comes with it, that still doesn't mean you can use it. Do they expect you to write your own software? If so, do they provide an SDK?
Also it has the number 666 in the model number, and as a Christian, I think it would be wrong to purchase any hardware who's model number is 666. That right there pretty much means I will NOT be buying it.
But that says SDR. That's a radio receiver. I don't need it for a radio receiver (other than receiving base-band frequency signals). I don't want an SDR. That includes a tuner IC between the input connector and the ADC, which means you are not getting a baseband signal to the ADC.
Yes, this is DDC SDR, as I said it doesn't consists FPGA for stream preprocessing, it just transfer RAW ADC stream to the PC in realtime. This is what you're asked for.
The main mode for this receiver is direct sampling. It means that ADC input is connected to direct sampling connector and ADC output is transfered to PC as a RAW stream with no preprocessing.
The tuner is used in optional mode to allow frequency down conversion for frequency above 30 MHz. The tuner mode uses separate SMA input connector.
Actually this is not SDR receiver but a high speed realtime ADC acquisition board with two SMA inputs. One is direct sampling ADC and one through onboard tuner for frequency down conversion. If you don't needs frequency down conversion, just don't use it. Use direct sampling input.
Also it has the number 666 in the model number, and as a Christian, I think it would be wrong to purchase any hardware who's model number is 666. That right there pretty much means I will NOT be buying it.
yeah, but there are a lot of different clones which don't have it, you can find for example RX103, RX888, etc...
For example, here is another clone with no such signs:
https://www.aliexpress.com/item/1005001710002655.html
But that says SDR. That's a radio receiver. I don't need it for a radio receiver (other than receiving base-band frequency signals). I don't want an SDR. That includes a tuner IC between the input connector and the ADC, which means you are not getting a baseband signal to the ADC.
Yes, this is DDC SDR, as I said it doesn't consists FPGA for stream preprocessing, it just transfer RAW ADC stream to the PC in realtime. This is what you're asked for.
The main mode for this receiver is direct sampling. It means that ADC input is connected to direct sampling connector and ADC output is transfered to PC as a RAW stream with no preprocessing.
The tuner is used in optional mode to allow frequency down conversion for frequency above 30 MHz. The tuner mode uses separate SMA input connector.
Actually this is not SDR receiver but a high speed realtime ADC acquisition board with two SMA inputs. One is direct sampling ADC and one through tuner.
Is the input using a DC blocking capacitor? If so, it's useless for me. I intend primarilly to monitor not only RF baseband signals (within the bandwidth of the digitizer), but also digitize NTSC video signals which REQUIRE a DC signal to be present (blanking level is 0V, sync level is -0.286V, black level is +0.075V, and white level is +0.714V). These voltages would change based on image content if a DC blocking capacitor were used. A properly working NTSC capture device uses DC coupling, not AC coupling, on its analog input.
Is the input using a DC blocking capacitor? If so, it's useless for me. I intend primarilly to monitor not only RF baseband signals (within the bandwidth of the digitizer), but also digitize NTSC video signals which REQUIRE a DC signal to be present (blanking level is 0V, sync level is -0.286V, black level is +0.075V, and white level is +0.714V). These voltages would change based on image content if a DC blocking capacitor were used. A properly working NTSC capture device uses DC coupling, not AC coupling, on its analog input.
I don't know, needs to ask the seller. In case of need, you can add jumper to bypass input RF frontend circuit.
But anyway DC blocking capacitor is not a problem to digitize NTSC signal, it don't use DC offset. So, I don't think that you will needs to do some tweaking. It should works just out of the box with NTSC signal.
I just recommend to add some 10-20 dB attenuator on the input to avoid ADC overload and for additional protection, because it has high sensitivity. Since it has 16 bit ADC with a high dynamic range, 0.714 V should be OK even without attenuator. But if you're planning to attach some signal more 1 V, it's better to have attenuator on the input.
These voltages would change based on image content if a DC blocking capacitor were used. A properly working NTSC capture device uses DC coupling, not AC coupling, on its analog input.
No, they cannot be changed, because they have pretty high frequency, so even if it has DC blocking capacitor, this is not a problem. All these elements of the signal will be passed through capacitor with no problem.
Some time ago I connected my personal computer NTSC video output to a tube TV. And I was used DC blocking capacitor. It worked good through capacitor
DC blocking capacitor even better here, because it remove possible DC offset, and it will improve dynamic range for ADC capture.
Also note that you're needs fast SSD hard drive which will be able to write realtime stream from ADC. Because it needs about 32 MHz * 2 bytes = 64 MB/sec realtime continuous write speed. Mechanical HDD is not enough for a such speed.
Also you can use AD9226 ADC module, it don't have DC blocking capacitor:
https://www.aliexpress.com/item/33003064732.htmlBut it has just a RAW parallel output from the ADC, so you will need some circuit to transfer it to PC. This is not easy task. And this is what is done with Cypress FX3 USB3.0 controller on these Chinese clones of BBRF103 receiver.