A recent post in the CSL blog about ERGO (Energetic Ray Global Observatory) http://www.ergotelescope.org/ got me thinking about sensors for my Otter Creek Research Station. The ERGO project would like to have a thousand sensors scattered around the world reporting cosmic ray hits. The current design of their sensor uses a Geiger counter but there is interest in using a Resistive Plate Chamber (RPC) with a much bigger capture area. I did a quick web search for information on RPCs and found a lot of articles by researchers building, testing and using RPCs but I did not run across any amateur efforts. This stoked my interest because it is more fun to work on something that not everyone is doing. From all that I read is sure looks like an amateur like myself could build a working RPC. There are a lot of issues that might prevent an RPC from being used reliably as a long term sensor but it is worth investigating.
A resistive film is used to distribute the high voltage field across the chamber. This film must be conductive enough to distribute the charge but not enough to shield the signal of the discharges from the sensor plates. This film needs to have a resistivity of about 1 meg ohm. A variety of approaches to making the film have been used. Plastic antistatic bags, India ink and evaporated metal films have been used. The most popular film seems to be graphite conductive paint. I have a bottle of powdered graphite that is used to lubricate seed planting machines so I did some experiments to see if it could be used to make a conducive film. I mixed some of the graphite powder with lacquer thinner and nitrocellulose lacquer and painted it on a glass plate with a brush. In the real application the lacquer would be sprayed on but this gave me a quick test and allowed a varying thickness sample to be made. After drying, inspection indicates that the graphite powder is probably too coarse for this application as you can see the individual grains and the surface feels rough. Testing with an ohmmeter showed that the thicker parts of the sample were indeed conductive but about 2 orders of magnitude too conductive. A finer powder may allow for a thinner layer with more resistance. When I need a finer powder the thing to do is put it in a ball mill. I placed about 4 ounces of the graphite in a ball mill and set it running. I will let it run for about 24 hours and see what I get.
Above is a picture of the graphite painted on to a glass plate. You can see the grains of graphite in the thinner areas and the probe marks where I measured the resistance. It was not an accurate resistivity measurement but it did give me some idea of how it was going to work. I suspect there is a wide range of particle sizes here as you can see the black specks and the area around it is grey which would be the particles too fine to see at this magnification.
After 24 hours in the ball mill I checked on the results. The first thing I noticed when dumping the balls and graphite out was that most of the graphite was packed into the bottom of the drum. It is possible that it did not get milled too well and spent most of the time in the bottom of the drum. I mixed up a sample with the nitrocellulose lacquer and painted another test stripe on the glass plate. The milled graphite is noticeably finer both visually and by feel. Because of my suspicion that it did not get milled too well I put the batch back into the mill for another day. This time I will check it from time to time to make sure it doesn’t get packed into the bottom.
Each time I checked the ball mill the graphite was all packed into the bottom of the drum. That is kinda weird because the drum rotates horizontally and there is not much difference between the bottom and the top as far as how it should behave when rotating. Also from the appearance of the graphite milling seems to pack it into clumps rather than mill it finer. So I decided to try and mill some of the graphite wet. I placed some graphite and lacquer thinner into a jar with some of the ball bearings I use. I will see how that looks after running overnight.
When I checked the thinner and graphite in the ball mill today I found a thick paste in the jar. Apparently I did not add enough thinner or added too much graphite. Testing it anyway I painted some of the paste onto my glass test sheet. After drying I could see that the graphite had milled up nice and fine and the file measured in the 1K ohm range. The film for the RPC needs to have a resistance in the 1 Meg ohm range. So something needs to be done to increase the resistance. I found that if I added any lacquer to the mixture the resistance was too high to measure. Apparently the lacquer insulates the graphite particles. So it might be possible by using a very small amount of lacquer to adjust the resistance. Another possibility is to make the film thinner. One place on my painted test films I found a thin area that would measure in the Meg ohm range. This may work if in can control the film thickness.
It was suggested that I look at a commercial product called aquadag. I had avoided looking at commercial products because I wanted to keep as much DIY as possible in the project. But I did have a look at information on aquadag. Now I have known about aquadag for a very long time but never really paid much attention to all of what it is made of. The description of aquadag is an suspension of water and colloidal graphite. The MSDS for aquadag also lists ammonia. The ammonia is there to reduce the surface tension and allow the material to wet the surface. This also tells me that it is undesirable to have any other material in the mixture as ammonia will evaporate along with the water when the coating is applied. Which then leads me to wonder why the stuff seems to stick so well. The information on aquadag gives a resistance in the 1K ohm range which is similar to measurements of my preparation. Based on this I will continue to experiment with my DIY preparation and switch to using water rather than the lacquer thinner.
Next I will investigate both adding a binder and controlling the coating thickness to bring the resistance up to the 1 Meg ohm range. For the binder I will investigate using sodium silicate and get set up to try spraying as a coating method. Sodium silicate sticks very well to glass. The reason I am concerned about how will the coating sticks is that I want to laminate the pickup electrode assemblies onto the coating. If I can’t do that I will have to have a way to hold the stack together mechanically.
Some of the articles on building RPCs talk of the difficulties of spraying a uniform coating. This got me to thinking about another coating method called spin coating. Spin coating is used where one needs to have a very uniform thin coating. It is used in the semiconductor industry to coat the photo resist onto the silicon wafers.
In other areas of the project, today I ordered some copper tape to connect to the graphite coating and a high voltage power supply.
The high voltage power supply and copper tape have been received.
I have given up on using a solvent based binder like lacquer. Almost any amount of lacquer renders the film non-conductive. Some graphite and water were ball milled for two days. When this was applied to a glass plate and allowed to dry it did make a conductive film but it has no adhesion. It will just wipe off. Sodium silicate can be used as a water based binder so I set up an experiment where I would add increasing amounts of sodium silicate to the water graphite mixture and paint test strips on a glass plate. The first test strip was with no binder then one drop, then 2 drops etc. I was hoping to find either a gradual increase in resistance as the sodium silicate was added or some sort of threshold where it became non-conductive. Instead after drying all of the test strips including one with a large amount of sodium silicate were conductive to about the same amount. One good thing though, after three drops the graphite adhered quit well to the glass plate. It will not scrape off using my fingernail. The test strips were all too conductive and it looks like it will be difficult to apply a uniform thin film that will have the desired conductance. Another possible adhesive to explore would be albumen.
In parallel with the water/graphite experiments I tried a different approach. Dave in his comments had mentioned the possible need to rough up the glass to get the desired adhesion. Graphite sticks quite well to paper so it should stick to roughed up glass. A sample piece of glass was sandblasted and I just rubbed the graphite on to the blasted surface. The resulting film had a conductance that was in the ball park of the required amount. For now this looks like the approach I will take. With this approach I will not be able to laminate the signal electrode onto the glass plate as I had planned. The stack will have to be held together with some sort of external clamping means. The sandblasted glass was fairly rough so I am going to try a test with glass wet sanded with 40o grit sandpaper.
I took a break from the RPC for the holidays and work on some other projects. What I know now. Adding any lacquer to the graphite makes it non-conductive. Using sodium silicate works well as a binder and the graphite remains conductive with even large amounts of sodium silicate. It is difficult to make the sodium silicate graphite thin enough to get the proper resistance. Rubbing dry graphite onto sand blasted glass works well to make a film with the proper resistance. The sand blasted glass is the approach I will take for now. Another approach is to use the black plastic film used for anti-static bags. I will save that in case the graphite does not work. I don’t have a blasting cabinet so to avoid making a big mess I have to blast outside. Right now the temperature is in the single digits with 15 mph winds. I am going to wait for more comfortable weather to do the blasting. Next I am working on gathering the bits and pieces needed to make the gas mixture filling for the RPC. I will start another blog post on that subject.
While waiting for better weather to do the sandblasting I have continued to research alternates for the graphite conductive coating. The DIY electrostatic loudspeaker crowd have a similar need. An electrostatic loudspeaker has a thin diaphragm than needs to be conductive with a similar resistivity. As with any DIY group a lot of different materials and coatings have been tried and used. One material stands out as being easy to use and meeting the requirements, that is Licron Crystal ESD coating by Techspray. I have bought a can of the Licron and will give it a try. If it works it will be a lot easier than sandblasting and rubbing on graphite. I have also bought a high voltage probe for my multimeter. It is a bit of overkill going to 40 KV but I do need to measure up to 10 KV that will be applied to the RPC.