| Author |
 Topic  |
|
|
ndrewtp
2 Posts |
 Posted - 08/27/2025 : 08:38:12
|
If you feel adventurous, as of yesterday there is a RadPro Firmware upgrade for the GMC-800 that makes it a best in class machine!
https://github.com/Gissio/radpro |
Edited by - ndrewtp on 08/30/2025 21:39:33
|
|
| Reply #1
ullix
Germany
1220 Posts |
Posted - 08/27/2025 : 23:39:17
|
Wow! This is some very impressive reverse engineering!
GeigerLog does support the RadPro firmware alreay. Though in this moment I can't tell whether this also applies to the GMC-600 variant.
If any users can tell, I am very interested!
|
 |
|
| Reply #2
Kaninbjerget
Denmark
28 Posts |
Posted - 09/01/2025 : 10:50:40
|
Wow nice! Use it on the FS5000.
Wonder though if I have the guts to try to install it on the 800. If there's a way to get the original firmware back without having to send it to GQ - something I don't expect they'll do for free. |
|
|
| Reply #3
ndrewtp
2 Posts |
|
| Reply #4
Kaninbjerget
Denmark
28 Posts |
Posted - 09/02/2025 : 12:13:52
|
Thanks! I overlooked that. I have an updater which I saved as my 800 had some bugs when I first got it. Perhaps it's time to check for an update again as there's still some bugs although not as severe as when it was released. Back then readings made erratic jumps from time to time. Thankfully that isn't the case anymore and it's fast and seems accurate - like all their other units with newer firmware.
But even though I'm not sure I'd go with Radpro in the 800 - might try it then but I have my doubt I'd go with it in the long rung. Actually I think the algorithm in the 800 is better than Radpro. Giving a stable reading faster. Up until at least 2.1.1 which I currently have on my FS5000 - will be upgraded soon when I don't have carpenters running around the house so I'll have my workbench back - it only updates display when it received a count. It doesn't do a more intelligent average like the 800. Beside personally I'd like a little more eye candy and I really love the main screen on the 800 which regardless of which unit you want displayed the largest also show all the others - no constant switching unit depending on what you want to see.
Beside the stock firmware on the 800 is really good and except for a game and dice I don't think Radpro offers much that the stock doesn't have. Quite the opposite on the FS5000. It's stock firmware is really horrible and has downright wrong conversion of CPM to uS/h which varies greatly even at very low counts. At low counts it's 170 CPM er uS/h and not much higher it can be down to 80 CPM per uS/h. Way to low to even start taking deadtime into consideration. So for that the Radpro it not only nice - more a less a necessity if you want readouts in uS/h (yes I know it's only valid for gamma and right isotope but still for gamma only it'll not be that far from real values in most cases).
Edit: uS/h couldn't be written with the most correct greek micro |
Edited by - Kaninbjerget on 09/02/2025 12:16:11 |
|
|
| Reply #5
ullix
Germany
1220 Posts |
Posted - 09/03/2025 : 00:12:22
|
I am thinking about trying Radpro. But first I'd like to review the performance of the GMC firmware. Here is a first shot.
I am running the "GMC-800Re1.18" counter at background, and using GeigerLog 2.0 to measure the duration of a CPM and CPS call.
Most calls take around 25 ms, which is quite good! However, at rather regular intervals of about twice per day, the duration builds up, up to about 115ms, in a rather strange pattern. This is inconenient for reliable recording.
It looks like some firmware issue. Perhaps related to garbage collection?
|
|
|
| Reply #6
ullix
Germany
1220 Posts |
Posted - 09/03/2025 : 02:00:27
|
This awkward, ill-conceived FET ruins the day once again!
This is a GMC-800 with a Thorium source, creating something around CPM=900.
Don't use FET !!! You are creating Fake data!
Set it to FET=60 to switch it off. |
|
|
| Reply #7
ullix
Germany
1220 Posts |
Posted - 09/05/2025 : 00:28:11
|
More on this hilarious FET:
With FET=60 it is switched off. Then - in this example - you see a random fluctuation between CPM=800 up to 950. Reasonable.
However, with FET=3 - at its default, worst setting - it fluctuates between CPM=400 and 1400!
FET is the worst feature stuffed into the GMC firmwares! It renders your data to fake data.
|
|
|
| Reply #8
ullix
Germany
1220 Posts |
Posted - 09/09/2025 : 02:50:21
|
What a surprise! Using GeigerLog to look at the difference of "calibration" between a GMC-800 counter with either Firmware 1.09 and 1.18:
With firmware 1.09 each of the 6 calibration points codes the very same sensitivity of 154 CPM/(uSv/h). This value has been used by GQ for all tubes in their GMC-300, 500, 800 counters since the very beginning. And no matter what tube M4011, J321, J305 has actually been used in the counter (which is very strange as well).
However, now on FW 1.18 we find completely different values, but for the VERY SAME tube, how comes?
In particular, the new values signal that the sensitivity is only HALF of what it had been! So, from one firmware to the next, the tube has been downgraded about 2 fold, has become 2 fold less sensitive?
Let's see how this is explained in the manual. Looking into version Re. 1.04 May-2025 I find:
Converted into proper units this is: Gamma-Sensitivity=132 CPM/(uSv/h).
Why is this not matched with neither the old nor the new firmware? The new firmware numbers still suggest a 40% lower sensitivity than what is specified in the manual?
What shall we conclude out of this? You had wrongly overrated your counters' sensitivity by approx. a factor of 2? And we have to downgrade all measurements by this factor?
|
|
|
| |
Topic |
|
GMC-800: Custom firmware
