NetBurner Community Forum

Something you may want to know if you use the MOD5270 email feature. The hosting company for my website apparently won't let you send email to yourself. I was trying to get the Scintillator to send me an email and I kept getting a DATA_END_SERVER_REPLY_FAILED (-7) error. I tried a few things and nothing worked. I then changed the email address to my gmail (forwards to my main email) and could then send from the Netburner. Others may have this policy - so note well. I am going to send myself an email with each hours data. Once that is proven I will send every 48 hours and start putting 48 hour historical data in a data base.

Keep in mind that the email feature may require TLS in the future...

dciliske wrote:Keep in mind that the email feature may require TLS in the future...

Ok.

I decided that email is annoying, abandoned that idea, and am now FTPing incremented file numbers to the main server. A test file every minute for now and when I am done testing - I will send a file every hour. I then hope to use PHP to extract the data and put it in a MySQL database. Historical data will then be automatic and the bandwidth available from that server will allow graphs and so forth.

Pic shows a test earlier today.

http://www.pahrumpgeigercounter.com/

Now have the database updating on the "More Information" page. FTP works great and I am now just overwriting the same file name. PHP reads it on the top of each hour and updates the database. The "More Information" page then displays it. After I accumulate some data I will probably make tables for monthly averages and yearly averages. Graphs too.

http://www.pahrumpgeigercounter.com/info.htm Improved the information page. I added a 24 hour average and limited the hourly count data to the last 24 hours. This for the sake of neatness. Don't want a 1000 hour scrolling list. I will eventually have another data page with the last 30 days of 24 hour averages. That will be followed by a rolling 12 months and so on. So there will be lots of data in the data section. Soon I will add another table to the database for daily average values.

http://www.pahrumpgeigercounter.com/info.htm

Added a link "CLICK FOR HISTORICAL DATA" on the info page below the data tables. That takes you to yet another page which has the daily data for the last 30 days. When more data is available I will add months and year. That is it will have that daily data when it is up for 30 days. Tonight at midnight it will do the first actual day. The numbers there now are from a test today and will eventually be deleted.

Nate Silver now gives Trump a 57% chance of winning and having his little fingers on the nuclear button!

I am stoked! Stoked! Just think of thousands of scintillators, MOD5213s, and MOD 5270s needed to monitor fallout patterns. This is a YUUUGE opportunity!

To that end I am building a new high voltage PMT power supply. The schematic is what Ludlum uses now and I don't like it. It is based on a switching regulator chip normally used in generic low voltage power supplies. They have a 500 meg resistor as part of a voltage divider for the feedback and then a 1.2 meg adjustment. Way too much resistance in my view. Temperature changes the bias current at the regulator chip input, that moves the setpoint around, any surface humidity changes the 500 Megs, and a change of a couple of volts on the output changes the photopeak. Not good. Also the attenuation of the voltage divider reduces the loop gain. Mine is similar except I add an instrumentation amp to recover the gain lost in the voltage divider and use lower value resistors. I also use a voltage doubler and not a multi stage Cockcroft Walton. The voltage multiplier is for 40KV, not 1000. I can add a dominant pole to keep it stable while I run as much gain as it will take and still be stable. I want 900.00 VDC 24/7. More in a couple of weeks after I (hopefully) get it working.

I have an animated GIF with the Cesium 137 dispersion over Europe after Chernoybl on my web site. Watch the bottom of the front page, it will change, and wait.

http://www.huffingtonpost.com/entry/sca ... 3164c347d0

In case you think we don't really need CS-137 monitoring! I am going to put the existing scintillator in a temperature controlled cooler to increase accuracy. Have other work I have to get done - but will do this soon.

Got the HV PMT power supply together and running. Great regulation and low noise. Maintains 900.1V day and night with an adjustment range with the current voltage divider of 800 to 930V. I am going to temperature compensate the power supply. Since increasing temperature reduces the output of both the PMT and the crystal - and increasing high voltage has an exponentially opposite effect - I am going to install a NTC-MF52-103 / 3435 10K thermistor, in parallel with a 2.2K resistor, at the bottom of the voltage divider. This should almost exactly compensate for temperature. 5C rise will get me a ~ 1.6V increase. The PMT I plan to use is happy at 850V - perfect for this supply. The top picture shows the "background" noise with everything off so one has to subtract that. The second picture is the HV noise at 900.0V at 50 millivolts per division. The spikes are the MOSFET switching off, and I could add a snubber, but I am just going to use brute force filtering on the amplifier board. The amplitude of the spikes varies depending on small variations in the error signal - the longer the MOSFET stays on the more energy in the inductor and the bigger the spike. Subtract the existing noise and you get about 100 mv peak to peak. Just as good as a Hamamatsu (but cheaper). The preamp is next!

View of the bottom of the power supply board showing the 1 OHM current sense resistor (it needs a 1/4 watt!) and the inverter transformer. The inverter transformer is a surplus item that was part of a defunct disposable film camera flash system. They made too many transformers, film camera sales went dead, and so one can now get them almost free. I used one in the Geiger counter and now this HV power supply. Don't have any specs on it - but I have figured out what it likes. It would probably get max efficiency at about 20KHZ but is doing really well at 12.5KHZ. I tend toward a lower switching frequency because every time you toggle the MOSFET power is wasted during the transition. I don't go any higher than needed and don't need a heatsink! Transistor runs at 87F with the room at 79. And this is with the 10 Meg Fluke as a load. The PMT divider is 60 Megs.

Circuit board made in China - about $4 each if you make 10. Just upload the Gerber and Drill files and DHL brings it. Allows all sorts of creativity.

Trump could win and you will need an on line scintillator to watch for CS-137! Little fingers, big red button. Not good folks, not good!