Zeus Prototype
Zeus High Voltage Project
Updated 01 December 2011. More links added and fixed in the 2nd paragraph per requests.
There are frequently discussions on the GCE list about home-brew high voltage supplies. For various reasons I decided to try to make a good, small, adjustable, relatively inexpensive regulated high voltage supply. I also want to make it an “open source” project with the hopes that you folks will improve it.
With no further ado, the schematic for the Zeus can be downloaded from here, the layout from here, bill of materials (Excel spreadsheet) from here, and the microcontroller controller code from here as a txt file or here as a zipped pbp file. Schematic and layout are in pdf format and the layout is at twice actual size. I'd be happy to provide the Eagle files if anyone is interested. Microcontroller code is written in PICBASIC Pro from microEngineering Labs. Just want the hex file (the compiled code to load onto the microcontroller)? Right here. I program my microcontrollers using the microEngineering Labs U2 programmer. There are many less expensive PIC programmers out there that will work just fine. I confess I do have a bias towards the MELabs stuff which, although a little pricey by some measures, is 100% professional quality. I've never had any trouble with anything from MELabs ever. Well, that wasn't my own doing...
Anyway, with a 5V regulated supply the Zeus can provide up to around 1500V at about 25µA. With no load it draws about 0.75mA at 500V and a little over 1mA at 1000V. It should be sufficient for just about any GM tube. By messing with the microcontroller code the maximum HV can actually be increased well above 2000V but efficiency and regulation suffer.
A *regulated* voltage must be fed into the board. On power up the position of the jumper is read to determine whether to use the settings for a 5V or 3.3V supply. At 3.3V the maximum HV drops to about 1300V. The microcontroller turns on the transistor Q1 and holds it long enough to nearly saturate the transformer. It then cuts off the current and allows the transformer to ring. The output from the transformer is multiplied and filtered. The output is sampled, divided and fed back into the microcontroller’s internal comparator.
So anyway, it occurred to me that before I have a run of circuit boards made I ought to check with the group to see if (1) there is actually any interest and (2) there are initial comments that might push it into a different direction. The current design is mostly surface mount parts to keep it small (1 by 2 inches). I kept the microcontroller a socketed through-hole part so that it can be easily removed and reprogrammed. I can solder all of these parts by hand with no special equipment but am not sure if others would instead rather have a larger board that was easier to solder.
I had a small run (10 pieces) made of a prototype and have 5 remaining. There is a mistake on it that must be fixed by adding a jumper. I can provide these as kits for $30; for an additional $10 I’d be happy to provide it assembled. A cost of $30 or $40 sounds like a lot to me but I assure you I will lose money on each one…8^) Presuming interest in this exists, the cost will hopefully come down.
There are frequently discussions on the GCE list about home-brew high voltage supplies. For various reasons I decided to try to make a good, small, adjustable, relatively inexpensive regulated high voltage supply. I also want to make it an “open source” project with the hopes that you folks will improve it.
With no further ado, the schematic for the Zeus can be downloaded from here, the layout from here, bill of materials (Excel spreadsheet) from here, and the microcontroller controller code from here as a txt file or here as a zipped pbp file. Schematic and layout are in pdf format and the layout is at twice actual size. I'd be happy to provide the Eagle files if anyone is interested. Microcontroller code is written in PICBASIC Pro from microEngineering Labs. Just want the hex file (the compiled code to load onto the microcontroller)? Right here. I program my microcontrollers using the microEngineering Labs U2 programmer. There are many less expensive PIC programmers out there that will work just fine. I confess I do have a bias towards the MELabs stuff which, although a little pricey by some measures, is 100% professional quality. I've never had any trouble with anything from MELabs ever. Well, that wasn't my own doing...
Anyway, with a 5V regulated supply the Zeus can provide up to around 1500V at about 25µA. With no load it draws about 0.75mA at 500V and a little over 1mA at 1000V. It should be sufficient for just about any GM tube. By messing with the microcontroller code the maximum HV can actually be increased well above 2000V but efficiency and regulation suffer.
A *regulated* voltage must be fed into the board. On power up the position of the jumper is read to determine whether to use the settings for a 5V or 3.3V supply. At 3.3V the maximum HV drops to about 1300V. The microcontroller turns on the transistor Q1 and holds it long enough to nearly saturate the transformer. It then cuts off the current and allows the transformer to ring. The output from the transformer is multiplied and filtered. The output is sampled, divided and fed back into the microcontroller’s internal comparator.
So anyway, it occurred to me that before I have a run of circuit boards made I ought to check with the group to see if (1) there is actually any interest and (2) there are initial comments that might push it into a different direction. The current design is mostly surface mount parts to keep it small (1 by 2 inches). I kept the microcontroller a socketed through-hole part so that it can be easily removed and reprogrammed. I can solder all of these parts by hand with no special equipment but am not sure if others would instead rather have a larger board that was easier to solder.
I had a small run (10 pieces) made of a prototype and have 5 remaining. There is a mistake on it that must be fixed by adding a jumper. I can provide these as kits for $30; for an additional $10 I’d be happy to provide it assembled. A cost of $30 or $40 sounds like a lot to me but I assure you I will lose money on each one…8^) Presuming interest in this exists, the cost will hopefully come down.