Hello, this is my first blog posting. Many, many thanks to all the people who have been doing the excellent reverse engineering work on this series of scopes. I have read through parts of this blog searching for general information on the MSOX3000 series and finally decided to get one. I purchased a new (Keysight demo at 50% off) to me scope from Keysight, and of course it came without any probes or cables. Probes are not so much of a problem, there are many sources for those. But the MSO cable is a problem because it is less common.
After reading through this blog and doing other research on the Keysight web site, I found that the MSO probe cable requirement seens to be the same as used for a series of HP logic analyzers with the same 40 pin pinout, as has been pointed out previously.
An excellent document for this type of cable and its circuit requirements is the Keysight document "Probing Solutions for Logic Analyzers", document number 5968-4632E. This can be downloaded from the Keysight web site. In particular, the section from page 47 onward gives a lot of information about this probe system, such as the typical clip lead nose termination network in Fig. 9.1 and 9.36. All of the networks shown in this document assume use of the 'woven cable' special 40 pin cable to connect the isolation networks for the clip leads to the input of the logic analyzer.
Figure 9.37 on page 71 shows the signal connections as viewed from the end of the MSO cable connector. Therefore looking at the front of the scope, pin 1 is at the top right towards the USB connector (top left square pin in the main board top side photo), and pins 1,3,5, and 39 are not connected inside the scope. You can verify this in some of the main board photos. Note the black 'key' indicator is at the top of the socket, and Fig. 9.37 is as if you are looking at the end of the cable. In the scope, all the grounds are on the bottom side.
The signal paths in this logic probing system are set up the same as normally used in a 10X oscilloscope probe, and require particular conditions to be met to get good bandwidth and transient response at the input of the logic analyzer. Each of the 16 signal paths (the clock path in a cable from a logic analyzer is not used) consists of the isolation network at the clip end, one signal wire in the cable, and a termination network in the logic analyzer or MSO scope input. The signal wire is a special type of pseudo coaxial cable made with very fine resistive wire for the center conductor, so that the 1 meter (~40 inch) long woven cable signal path has an end to end DC resistance of about 180 ohms. The grounds of the paths have very low resistance from end to end. The measured low frequency capacitance of one signal path to ground is about 50 pF, and the transmission line TDR surge impedance Zo measures as about 180 ohms also (coincidence?).
You can see in some previously posted photos of the MSO region of the 3000 series that U1401 to U1404 used as receivers for the MSO signals are MAX9201 quad comparators. Part of the circuitry used with these comparators is visible on the top side of the PC board, but apparently the remainder is on the bottom side. I have not been able to find any high resolution photos of the bottom side of the PC board.
If someone can point to where the backside area under these ICs can be seen, I would greatly appreciate that information. I am reluctant to disassemble the scope I just bought since it is still in warranty. Any information on component values would be a bonus.
My goal is to design a replacement for the standard MSO cable which anyone can duplicate. Obtaining satisfactory frequency response when using a standard ribbon cable for the 'woven cable' requires adding additional compensation networks at each end of the cable. The values of these networks can be obtained from simulations, but the ones already present in the scope, which cannot be altered, must be known in advance.
Making one's own cable fixtures has the feature that specialized cables can be made for common usage situations, such as using a smaller number of channel connections or using a ganged connector to the target breadboard to make usage easier. For instance, a pad pattern for a vertical mounting through hole micro USB connector could be added to a PC board design and connected to otherwise unused pins of a CPLD. In this case, you would have a 5 channel gang connection, which by programming can monitor any signals in the CPLD. Just alter the CPLD program to move your hardware probe points and reprogram without taking it off of the PC board. The connector would be left unloaded on boards after the debugging is complete.
Or you could use an 8 channel connection with a 10 pin header location on the PC board to probe 8 signals as desired, or have only 8 clip leads flying about.
So what I am asking for is any information that someone can give to help determine the termination networks internal to the MSO 3000 series oscilloscopes.
Thanks for your help!