Micsig Tablet Oscilloscope tBook mini TO1104 review (100Mhz 4 channel 'scope)

[fishandchips]

Excuse me. What is mean by "floating" the scope?

Removing the GND in in the AC plug so the oscilloscope ground is not connected to earth.  It is a very bad and dangerous practice.

Thanks. Why people do that? I cannot understand such action.

[paul_ius]

Thanks. Why people do that? I cannot understand such action.

In this video Dave explains the problem with the grounded scopes pretty well:

[frankfan]

Hi,
   I just purchased the TO1104, and I did not found the serial decoder and trigger from the manual. The firmware version is 7.14.0.227. I am just wonder how can I use the i2c decoder. Since I did not see anywhere from the manual or the wedsite mention the support of the serial bus decoder feature. Please help

Thanks

[mikeselectricstuff]

Hi,
   I just purchased the TO1104, and I did not found the serial decoder and trigger from the manual. The firmware version is 7.14.0.227. I am just wonder how can I use the i2c decoder. Since I did not see anywhere from the manual or the wedsite mention the support of the serial bus decoder feature. Please help

Thanks

The FW with serial decode is not officially released yet, but seems they will provide it if you ask.

[I4E]

Just updating our earlier post in the thread.  We are now authorized distributors for Micsig.  If anyone has any feedbacks or suggestions they'd like us to feedback to them by all means please let me know.

 As our website is only a simple site we will probably only put  a few on our website but we can supply the whole range.

Here is a link to the TO1104 on our website. 
https://www.instruments4engineers.com/micsig-to1104-100mhz-tablet-oscilloscope

Have a good day everyone.
Joy Torres
Instruments 4 Engineers
Tel +44 (0) 161 871 7450
info@instruments4engineers.com
www.instruments4engineers.com

[exe]

The FW with serial decode is not officially released yet, but seems they will provide it if you ask.

Micsig said serial decoding will be available on or before 1st of August. But I asked for beta version just in case of delays or something.

[exe]

I received my TO1104 a few days ago and I would like to share my mini review.

1. UI is cool, but a little bit slow. Does not distract, but not "iphone" experience. But, after all, I like this touch interface and I don't want to go back to knobs (never thought I would say this one day).
2. I like the way it displays information about verticals. For example, it shows 10mV/div and right there two buttons to change it. Cool! Although, I wouldn't call this a killer feature, just a good approach to UI.
3. Arbitrary BW limiting. I like it, it is very useful feature in my noisy environment. Is it implemented in hardware or software?
4. A bit of learning curve is present, but after a few hours of playing with it I'm comfortable
5. Probes are 10x only. Well, some people claim this is an advantage (less things to break, less mistakes, no chance to overload circuit with 1x probe). Still, I'll buy a pair of cheap probes with 1x for low-speed low-level circuits. The hook does not work well . Definitely I've seen better probes.
6. Stability. Mine works very stable so far. No problems with uptimes up to 4h. Software version 7.14.0.227.
7. It's floating. Perhaps, this is bad when doing mains and HV stuff, but good for my pet projects, especially because I can put my probes anywhere and don't bother that my signal generator is already grounded.
8. They significantly improved software over last six months or so. It differs quite a lot from what I can see on Mike's video. Very good!
9. Free beta decoding option (at least for me:)). Haven't tried it yet, but from other reviews looks to be more than enough for me, I don't need advanced stuff.
10. Video recording. Perhaps, not that needed feature, but "nice to have".
11. No serious bugs discovered.
12. Portability and size. I don't really need this, but it's good I can take it to work and show to my colleagues.
13. Fast to boot. Not instant, but good.
14. WIFI (haven't tested yet and I don't have this option enabled, but should be good for sharing screen captures)

Things I don't like
1. BNC on top are not very convenient for me. Also not very easy to attach / detach probes.
2. Fan is noisy, esp. after 1h of work. I'll have a look at tear down to see if I can replace fans.
3. Micsig site is not updated and in a very messy state. A lot of broken links, email didn't work for me (I communicate via aliexpress)
5. Angle of screen is bad for my setup. Not just viewing angle is bad, it also perfectly mirrors my table lamp. But I put a piece of wood under the stand, now it's fine.
6. Power button is too easy to press. Not good for a mobile device. But I haven't tried to carry it with me, so may be this not a problem.
7. Zooming does not work when triggering normal. That's annoying because I have to manually "re-arm" trigger after every measurement.

My wishlist (let me know if any of these are already present and I just was not able to find):
1. May be a bit faster UI update rate? I'm not sure this slownes of UI. It might be they don't want to, e.g., load CPU too much to extend battery life / thermal performance. I'll try to check if it's possible to do something from the shell.
2. Gain measurement? Phase measurement is already here, lovely.
3. Force trigger. Cannot find how to do this when triggering normal.
4. Cursors and scales on FFT
5. Multiple math channels
6. A few spelling mistakes. Don't really annoy me, just reminds the origin of the scope.

Haven't tested yet:
1. Networking and web UI
2. HDMI
3. USB
4. Video recording
5. Saving / exporting data. Hope it saves not just waveforms, but on-screen measurements as well.
6. FFT, just had a quick look.

Overall impression. So far so good. Doesn't feel like a super-professional scope, but suit my modest needs (characterizing opamps and power supplies, phase and gain measurements). I don't want to pretend I'm a big specialist in scopes and I'm very biased. And at this price range you have to make priorities. But I definitely recommend this unit for consideration. ALthough, I didn't spend much time with this scope yet, so I may change my opinion (hope not).

OK, now I'll try to get shell access to see if I can do something useful from inside. Who knows, may be it allows writing own plug-ins to implement new measurements or something.

PS thanks nctnico for suggesting this unit.

[lukier]

exe: I have similar impressions, although I don't mind fixed 10x probes.

One thing that I think is a missed opportunity is the lack of gigabit ethernet. Even NanoPi boards based on the same CPU have it. With that Micsig could do remote desktop like R&S RTB2004.

One idea for Micsig came to my mind. Serial decoders are just first step, but later one could use a terminal. Adding a terminal app (launched from the desktop menu) and enabling usb-serial dongle kernel modules could be very useful in the field.

[exe]

One thing that I think is a missed opportunity is the lack of gigabit ethernet. Even NanoPi boards based on the same CPU have it. With that Micsig could do remote desktop like R&S RTB2004.

One idea for Micsig came to my mind. Serial decoders are just first step, but later one could use a terminal. Adding a terminal app (launched from the desktop menu) and enabling usb-serial dongle kernel modules could be very useful in the field.

I think so . Now I'm trying to re-enable telnet again (so far I managed to unpack firmware file using encryption key published here). Once this is done we can discuss further steps. I don't know how much bandwidth for remote desktop is needed, but even 100mb/s is enough if done properly. I'm more worried about CPU and memory resources required. It looks like it has just 256MB RAM available, this might not be enough for many ideas (like real-time screen capturing and video encoding). But loading kernel module is certainly possible. And enabling wifi (just a matter of launching the script).

What I would like to try is to cooperate with MicSig. May be they won't mind enabling some of the features in official firmware. I'll write if we achieve something with new features.

[nctnico]

One thing that I think is a missed opportunity is the lack of gigabit ethernet. Even NanoPi boards based on the same CPU have it. With that Micsig could do remote desktop like R&S RTB2004.

Gigabit ethernet is not necessary for this kind of equipment (not even for remote desktop). The typical processors inside can't handle this kind datarate. For example: The RTB2004 reaches speeds of (approx) 12Mbit over it's network interface.

[Micsig_support]

Hello, Your wishlist 3,  you can see it in the picture.

and we also have the screen mask to make the screens displays better, pls contact our sales to get it. I

[Micsig_support]

To lukier.

Do you mean to use the PC software to control scope by PC, we have it already, and you can download it on our official website. it can supports WIFI connection also.

[Micsig_support]

Hello EXE,

3. Arbitrary BW limiting. I like it, it is very useful feature in my noisy environment. Is it implemented in hardware or software?---It is hardware.

Gain measurement,Multiple math channels,  and gigabit ethernet, RD reply me they do not have the plan to do them right now.

4. Cursors and scales on FFT---we have it. see pictures here.

[lukier]

Gigabit ethernet is not necessary for this kind of equipment (not even for remote desktop). The typical processors inside can't handle this kind datarate. For example: The RTB2004 reaches speeds of (approx) 12Mbit over it's network interface.

I guess that after video encoding the bitrate required is not that high, not sure about the impact on latency though. I'm not sure about the statement that SoC can't handle such datarate. I'll run an iperf test with the NanoPi 2 Fire board and report here. I know some SoCs (AFAIR iMX6) can have poor gigabit performance because of bugs.

It still seems silly for me that manufacturers go to such great lengths to save $1-2 or similarly insignificant amount - GbE is standard nowadays. Keysight of course beats Micsig and other companies on penny pinching, as they ask almost 400 GBP for what is essentially a 100Mbit MagJack - DSOXLAN.

This is a minor complaint, actually I like that Micsig integrated a fairly modern SoC platform into the scope, back when other low-cost brands used very outdated Blackfin or lowest-end iMX2 CPUs. Sometimes, the situation is very similar in the high end scopes, when one pays many 10s of thousands $ and gets Core2Duo with USB 2.0, seriously come on. PCs are ridiculously cheap.

Do you mean to use the PC software to control scope by PC, we have it already, and you can download it on our official website. it can supports WIFI connection also.

I'll try the app, but what RTB2004 does is live view and remote control from the HTML5 compatible web browser, no app required, see how fast it is here:
https://youtu.be/mcgJSKxj0i0?t=281

TO1074 web interface is pretty poor in the comparison. AFAIR the screen shot image updates every few seconds. I thought it might be the limitation of the Ethernet interface, but maybe the CPU is already under heavy usage from the scope app. But S5P4418 has a hardware video encoder, supporting for example H264, so in principle it should be possible to encode and transmit the live screen-stream without bothering the CPU cores too much.

Even with 100Mbit Ethernet the hardware (SoC) should have some headroom, so I bet many features can be fixed/implemented in the software 

[exe]

It still seems silly for me that manufacturers go to such great lengths to save $1-2 or similarly insignificant amount - GbE is standard nowadays.

What's the use case for this scope requiring 1Gbps link? I'd be worried about 256Mb ram...

[alm]

I also recall 1000baseT phys to use a lot more power than 100baseTX phys. That would be an additional consideration for a device that can be battery-powered.

I am actually positively surprised they did not include gigabit Ethernet. Putting gigabit Ethernet on a device that does not need it is the typical big, fancy features over functionality that I have come to expect from the Chinese test equipment designers.

[lukier]

What's the use case for this scope requiring 1Gbps link? I'd be worried about 256Mb ram...

Yup, another 256 MB is just few $.

I also recall 1000baseT phys to use a lot more power than 100baseTX phys. That would be an additional consideration for a device that can be battery-powered.

Hmm, good point. It seems Gigabit PHY requires four times more power. But I doubt it would be a common scenario to run on the battery while attached to the Ethernet.

I am actually positively surprised they did not include gigabit Ethernet. Putting gigabit Ethernet on a device that does not need it is the typical big, fancy features over functionality that I have come to expect from the Chinese test equipment designers.

My point is that the SoC is already fairly decent and with few extra $ here and there the functionality of the device could be extended with software. More complex on board waveform analysis (quad core is there, just space in RAM for buffers), live data streaming (GbE)  like a DAQ/Digitizer etc.

Look how rarely there is display output (on low-end scopes) somehow and Micsig did the right thing when they've decided it is worth it to spend a dollar or so on the micro-HDMI connector and some misc components - nice touch almost for free. Here one could also argue that HDMI peripheral block and LVDS transmitters consume so and so mW more.

Luckily, there is the USB host port (alas not 3.0), so one could enable support for various USB accessories. I've mentioned serial terminal app with FTDI dongles somewhere here before - I could see that useful in the field. Another possibility is a semi-MSO functionality - low cost USB logic analyzer with trigger in and out BNCs that would connect to the Aux output and one of the channels. LeCroy did something similar for their early non-MSO scopes, it was called MS-32. Apparently, it wasn't great but mostly because of buggy design by the 3rd party, not by the afterthought-like nature of the device.

[exe]

Here we come to the fundamental problem: there is simply not enough Micsig resources to implement all these features. Especially considering the price of the unit.

I think, only with the help of community it is possible to deliver such a rich functionality you mentioned (although, some features are fairly easy to implement). So, I offered my help to Micsig to implement a better remote control not requiring Windows and NI-VISA (I'm on Linux, but even on Windows I'd hesitate installing such a big package), but they politely rejected. Too bad .

Meanwhile I'm still trying to get telnet access and try to add more features unofficially. But no luck so far .

PS I hope NI-VISA uses standard protocol. This way I still can try to implement own remote control and measurement program. Or even a web-server to access the scope from any browser.

[alm]

I do not know the connection string they use for VISA, but I imagine it is just LXI (based on VXI-11). If they were just using sockets, then NI-VISA seems like a complicated way to do that. So connecting with something like python-vxi11 might be possible over Ethernet. Getting the correct commands may be a different matter. They may or may not be SCPI compatible. Sniffing the traffic with the original software might give some insight into this. Did anyone try to send *IDN? via VXI-11?

[nctnico]

I tried some port sniffing but TO1000 doesn't seem to support LXI (plain text over telnet) which makes it extremely hard to control the TO1000 over the network. To me it is useless if I need NI VISA. Why the hell are manufacturs still using VXI? Even Siglent switched to LXI on most of their test equipment 

[alm]

Your statement about LXI being telnet and VXI not being LXI does not make sense. LXI does not require the telnet service. The bare minimum required (see page 66) is a web interface and a discovery service. Telnet SCPI is optional (page 68), as is VXI-11 (same page). Hence my suggestion to try VXI-11, since telnet does not work and using NI-VISA for bare sockets seems odd.

I seem to recall NI-VISA shipping with some sort of VISA sniffer (NI Spy?). If it shows the connection string, you should be able to figure out the protocol NI-VISA uses (see page 68 of the LXI standard for examples of connection strings). You would need some sort of compatible Windows machine/VM to run their software.

As for why they are still using VXI-11 over telnet, because it has more features like group triggers and asynchronous SRQ. HiSLIP is a more modern standard built on top of  VXI-11.

[nctnico]

My point is: if you need a whole lot of extra bloatware to send some simple text commands to an instrument then something is wrong. NI-VISA is a major pain in the ass to setup and (for example) the version required for the TO1000 isn't compatible with the Windows 7 version I have. So how to go from there on a random PC at a customer??
When I write software which talks to instruments I use RS232 ports and let my software do it's own detection. That way I can ensure my software works plug&play and there is no third party layer which can screw things up. Just keep things simple and they will work.

[alm]

If you want to rant rather than solve the problem, go ahead.

While I do not love the various VISA frameworks, it is the industry standard and software from pretty much all instrument manufacturers, including Keysight and Tektronix/Keithley, uses it. For developers, the advantage is that it abstracts the interface to the instrument. You could be using serial, or USBTMC, or VXI-11, and the only thing you have to change is the connection string. So it is inevitable that any computer that does instrument control ends up with at least one copy of VISA installed, unless you exclusively run your own software. There are also (equally bloated) NI-VISA versions for MacOS and Linux if you prefer that to the Python VXI-11 implementation I suggested.

[nctnico]

If you want to rant rather than solve the problem, go ahead.

The problem can be solved if manufacturers have some kind of SCPI over telnet (let's call that LXI and I really don't want to go into semantics). Sending data to a serial port or over a telnet socket is equally trivial in a modern programming language. Also there is very little difference if I need to configure a serial port, IP address or connection string so all in all the NI-VISA platform doesn't really solve any problem for me. It only adds more places where things can go wrong, discomfort to download the right version (after registration on NI's website) and additional installation time.

[exe]

It only adds more places where things can go wrong, discomfort to download the right version (after registration on NI's website) and additional installation time.

Same here, plus don't want tons of bloatware on my computer. I prefer simple solutions if they work.