OOS – an open operating system for the Rigol Scopes
48 points by qwezxcrty 3 days ago | 13 comments

elevation 2 days ago
I acquired a used Rigol which has the i2c decoder disabled. I tried to contact them for a quote on a license for the decoders but Rigol North America would not respond.

I’m looking forward to software that makes it a decent tool that it should have been!

reply
the__alchemist 43 minutes ago
Hey - derail, and maybe not helpful to you but ... Have you considered a Logic Analyzer I stead for this? $10 for the Chinese ones, can use the Selae software (same chip), and I found it more convenient than using the Rigol. I use the Rigol for analog, but never found it convenient for digital.
reply
elevation 26 minutes ago
A logic analyzer is a better fit. I hadn't purchased a Salea (love them, use them at work) because of the cost, a bit out of my hobby budget which is why I was trying to lean on the equipment I already have. I know Salea prefers that their software only be used with their hardware; how do you prevent their software from detecting a fake and locking you out?
reply
echoangle 19 minutes ago
This is the open source alternative:

https://sigrok.org/wiki/PulseView

reply
the__alchemist 7 minutes ago
This will reflect my own anecdotal experience only: Pulseview crashes frequently (multiple times per session), and requires setting up the view each time you re open it, e.g. post-crash. I don't remember if it was a straight up crash, or just stopping receiving data from the probe until I restarted it.
reply
somat 4 minutes ago
Probably depends on the model of the scope, but I found it fairly easy to find a feature key generator to unlock scope features on mine. cough riglol.
reply
pseudohadamard 2 days ago
This is a really impressive piece of work, a scope isn't just a generic SBC with an ADC attached, it's a huge amount of fully custom analog and digital circuitry that you need to work with, combined with extensive signal processing software requirements, UI code, and so on. Julian (with some AI help) has essentially replicated what an entire dev team at Rigol did over quite a long period of time, it's actually worth getting a Rigol over other similar brands just for this.

Ob-direct link to the repo: https://github.com/deckerjulian/oos

reply
Doohickey-d 2 hours ago
At least from the eevblog post, it seems like the only re-implemented part is the "host" side, so Linux. A scope also has an FPGA, which is where the signal processing, sample memory, and acquisition happens. This one is much harder to reverse engineer, because it's effectively impossible to decompile the bitstream back to code (verilog), it's not possible to debug it in any meaningful way, so you're really starting from scratch in the dark. And it's extra hard, because you need to know what's connected to which pin on the FPGA, which is impossible to do without desoldering it or destroying the board.
reply
torginus 2 hours ago
If he can get a black-box spec for what the FPGA's supposed to do, he can write a cleanroom implementation (or perhaps even add features), or rebuild the whole thing, including what the FPGA actually does and how.

I don't know what FPGA is in these (edit: it's Xilinx), but the usual problem is that toolchains tend to be highly proprietary.

reply
guenthert 2 hours ago
Well, the github repository makes it look like much more was accomplished: Linux on the host was actually already done by someone else, he (and his AI agent) apparently reverse-engineered the custom application of Rigol which controls the FPGA (I suppose he left the FPGA configuration file alone) and is able to perform measurements.
reply
fock 27 minutes ago
and the code looks very similar to all the wideband sdrs we have mostly forgotten now...

Although a 0-dependency approach from-scratch approach is likely sensible here.

reply
fock 2 hours ago
"a lot of AI help" I assume...
reply
destitude 15 minutes ago
What difference does it make?
reply