ELEKTOR FORUMS

Hi JJL,
The problem is the voltages. If your serial port uses TTL levels, it will be no problem. Otherwise (which is more likely) the voltages of the COM port are too high and will damage the microcontroller. If you want to use a 'standard' COM port, you should use a max232 (or similar) to protect the circuit from too high voltages. See also here and here. The latter also has a pinout for DB9-COM ports.

I have the COM port set to 19200 baud, but I can not see anything on the chart, and I can not set the level. I noticed that in the lower left display, the second moving very slowly. Right? or, can it be that the error?

Thanks for your attention.

alefbe... What I was not able to do is change the L threshold level: no matter which value I choose and button I push (level or eeprom). L always stays at 10.

Hi Alefbe,

Did you order a kit or did you program the controller yourself? If the latter, could it be a fuse setting?

Also, in steps:
- On the oscilloscope observe a noisy signal about 800mVpp.
- For the resistors, I have found that in the scheme and the list of materials, R5 and R7 are 10Khom in editing the journal in English, but Spanish version, R5 = 3K3 and R7 = 1K Could be such an error?
- On the other hand, my transistor is a BF245C, Should be a BF245B or A?
- Another issue is that when setting the level to 100 with the software, turning on the meter again, mark "L154"

Thanks for your attention. Greetings: Juanjo.

I ordered a full kit, I do not know how to program the ATmega88 on the board.
I do have some experience with programming the Arduino UNO board but I think that is a totally different environment.

By the way, just for fun, I connected the "probe" circuit to Arduino
and then adapted your program for Arduino and it works great and was also able to use the VB program as is to collect data on the PC.
However I still would like to use the original board with its nice display.

Hi alefbe,

We haven't seen this problem before. Could you please check that your RX connection from BOB to processor board is working properly?

Hi,
Is it possible to improve the meter to handheld version, so that test objects must not be inside the can with the meter?
I did built the radiation meter from the kit. It is very sensitive indeed and shielding is a challenge.
It should work if to drill small hole to the shielding and place sensor directly behind and then cover the hole with some material that lets the radiation pass but stop the light.

Anyone tested some filter materials already?

I propose we make some tests with: 1. thin black plastic that is used packing photographic film. 2. Mica painted with black paint.

Wiki says that some Geiger–Müller tubes are covered with mica and are still able to detect alfa radiation. So we sould at least be able to detect beta radiation through mica and paint.

But first I will get some radioactive samples to test with.

Hi everybody!

I am making this radiation meter too, in the source code, it display only the CPM number. I would like to know if someone know how to convert this value in Sievert/h. Thanks by advance.

Hi Alfbe,

Very strange indeed! Hold on, I will check with the author. I'll get back to you.

thijsbeckers800 mVpp seems too much to me. My prototype only shows about 50 mVpp noise.
R5 and R7 should both be 10k. There were some last minute corrections in the schematic, that perhaps couldn't be entered anymore/magazine was already at the printer.
You should use a BF245B. As written in the article, the operating point is shifted when you use an A or C version.
I didn't notice the strange L154 indication yet, but I also didn't set the threshold to 100. I guess this will be ok when you have your hardware running properly.

Good luck and let us know if it works out!

Well, I have 10K resistors and BF245B, and the results are good, with a level of 60 gives an average of 0.15 counts / min.

Thanks for everything.