Re: [Linux-gpib-general] Attack of the clones - Keysight 82357B edition

[Philip P. <li...@ph...>]

On 11/08/2025 02:36, Hernán Freschi wrote:
> Hi,
> mine certainly does not look anything like that video. Here's a photo of 
> the insides of my 82357B https://i.imgur.com/iJXaqzg.jpeg <https:// 
> i.imgur.com/iJXaqzg.jpeg>. It's well built but it doesn't really look 
> like an official product. Certainly the part number lasered off is 
> rather unusual. I think in HP era a major hardware change like this 
> would probably be released as an 82357C hehe.
> 
> I did try it with Keysight's drivers and it worked "more or less". It 
> seemed to work with read/write commands. But I needed to open it in 
> Board mode to send a GET (Group Execute Trigger) command and it 
> segfaulted when I tried opening GPIB0::INTFC.
> 
> My guess is that this is a "functionally equivalent" clone that 
> implements good enough compatibility and that's it. It doesn't even try 
> to use Cypress chips and copy Keysight's firmware.
> 
> I tried shorting pins on the unpopulated header to see if I could make 
> it boot in bootloader mode, to see if it's at least detected as an FX2 
> chip, but nothing changed.

I'd be willing to bet that the lasered chip is some kind of ARM Cortex 
microcontroller. The 10pin header is probably an ARM JTAG one, in that case.

If you can connect a JTAG pod up to it, you can probably do an ID read 
and identify the chip.

Either way, I've never known Agilent/Keysight to laser the markings off 
of chips -- they usually use 18xx-xxxx house numbers but even that's 
rare these days.

I'm willing to bet this is a clone, and someone's decided to "clone" the 
protocol after the firmware has been loaded, and only implemented the 
bare minimum.

-- 
Phil.
ph...@ph...
http://www.philpem.me.uk