Glass Tube Shootout

 

This is a comparison of 8 glass walled GM tubes, with a SBM-20 as a control. I did this some months ago, but hadn't pulled all the data together until now.

Of the eight tubes, four are of the J305 flavour - M4011, J305 (2021 - light sensitive), J305 (2023 - not light sensitive) and one that appears to be a copy, with no markings and came in a cheap FS-2011 Geiger counter. The other four are more interesting, or less, depending which way you look at it. The HH614 is from a pen type counter and is quite small and thin. The remaining three are very low sensitivity tubes designed to be used in high radiation areas - SI-3BG, Dextray 3G8B and DOB-50. The J305's and M4011 get easily form AliExpress or your favourite Chinesium outlet. Same for the HH614, but seemingly less available for just the tube on it's own. May be easier to buy a pen type counter, with the tube. SI-3BG - there are usually some of these knocking about on ebay for cheap. The Dextray 3G8B was a one-off ebay purchase. Quite a rare tube, maybe. There is certainly next to no info about it. And finally, the DOB-50 came from a (parts machine) DP-66M.

The tests were performed in a fairly consistent manner on the same counter. Connect the tube and make sure anode resistor and voltage are correct. Shove one source up to the tube, measure over 5 minutes to get an average. Then move on to the next source, and the next and so on. Once the 5 sources and background have been counted, change the tube. Lather, rinse, repeat. The charts are in both linear and logarithmic as the background counts simply weren't showing on the linear chart. Quite surprisingly, the genuine 2019 J305 came out the most sensitive of all, even more than the SBM-20 control, and the Dextray hardly gave any clicks whatsoever, to anything. The SI-3BG (or multiple thereof) would be great for high radiation monitoring, but feels like it doesn't work with not getting a single background count in 5 minutes, but they are cheap, and available. Hmm, after a quick look, they aren't as cheap as when I bought them. They were going for around £5.00 each. I got five for aroun £16. Now it seems they are around £18 each. I guess that's what war does.

Test results:
 

 

 

Forget the series resistor network on the CAJOE for measuring voltage. It simply doesn't work properly. Or maybe does for a specific meter. Use a HV probe or make a voltage divider, which you are already on with.

With regards to the SBM-20/J305 I do think the SBM-20 is far superior to the 305. I have about 10 SBM-20's and 3 J305's. All of my SBM-10's when on the same counter perform roughly the same. The J305's do not. One of my J305's is light sensitive, the other is not. One is more sensitive (especially to just background counts) than the other. The J305 seems quite sensitive to voltage changes, the SBM-20 does not and has a nice long plateau. I wouldn't use a J305 in anything I was serious about as they just seem inconsistent to me. Every one of my SBM-20's really do perform very similar, but there seems to be inconsistency with the J305. Not long ago, I went  halves with another user on 8 SBM-20's for about £200, so ~25 each. I tested each one and they were all roughly the same, so I threw on a couple of my older SBM-20's, and they were pretty much the same. As some J305's are light sensitive, and some not, and of the ones I have, all seem slightly off to one another. I just see them as being inconsistent. I also see a lot of them being used in stations here that tend to give erroneous readings from time to time, but that could be the counter more than the tube.

Of the J306, J307 and J308 tubes (all glass walled) I tested the J306 seems the most stable, the J307 the least stable and the J308 somewhere in the middle. The following graphs are all 1 month graphs, from Radmon. The testing was all done on the same counter and the tube positioned in the same place, and not touched. It is still sitting there now with a SI-21G on test. I have seen similar inconsistencies with the J305 tubes. Then there are the specs. Is there a proper data sheet for the J305 or others? I can't find one, only bits of info on various sites like AliExpress and Alibaba, and some of that just seems plain wrong. Such as the supposed specs state the J305 is 107mm long, with an effective length of 87mm, yet all of my J305's are all ~88mm long with an effective length of around ~65mm. For me, the inconsistency is just too great for me to rely on.

 
This is the J306

 
This is the J307

 
This is the J308

 
And this is a SI-21G - decent NOS Russian tube from the cold war period. Maybe even soviet era.

Notice how solid the SI-21G is nice and even though it is more sensitive than all of the J tubes. For me personally, the J type tubes just seem less consistent than the SBM-20, or other metal walled tubes. I personally won't ever be putting a counter out in the wild for background monitoring with anything less than a SBM-20. And then the cost.... The J305 was cheap at one point. I could get them for around £9 each (at least that's what I saw them on sale for) a few years back. Now they are £20+ each, so not far off a SBM-20. If a J305 was £20 and a SBM-20 was £30, I'd get the SBM-20 all day long.

You do have to be careful buying them as it appears the market has been flooded with SBM-20's that don't work. They are usually listed as 'for parts' or 'not working' or something, but you have to read every single word on the descriptions. Even the broken ones are going for good money! These are probably what is left of the stock once all the good ones have been sold so it is a bit of a mine field. You can still get good ones though, and I haven't bought a bad one yet. I did have to wait some time though for the 8 x SBM-20s to come up at a reasonable price. And when paying £20-25 for a J305, I don't think £25-30 is a bad price for the SBM-20.

You have one at 107mm. I have three at ~88mm. One from 2021, one from 2023 and one that I think is a copy and has no markings. Size difference, light sensitivity differences, no real datasheet that I can find. These are the inconsistencies that really put me off using the J305. If their production varies, then I'm sure so do the specifications. They just don't cut the mustard for me. I'm also not much of a fan of the CAJOE boards. I have had one intermittently fail on me although I fixed it. It was dirt on the carbon.wiper of the voltage control pot. The other one came with the tube and has just sat in my drawer. I don't particularly like the CAJOE. It's super cheap, as in doesn't come much cheaper (not that that may be an issue, though, pots needing cleaning suggests very cheap parts) and they are very rudimentary design. Not something I would use out in the wild for anything serious.

It does beg the question; are they reliable? As in, is the PSU reliable, is there any drift, does the detector pick up every count, does the voltage bomb with high CPM, are they sensitive to temperature changes? I may build some kind of voltage logging system and put one on test for several days and see how the PSU performs. I have a feeling it may drift somewhat, but uncertain until I test. I know a lot of stations with the J305's that submit some bonkers readings, use the CAJOE. That is going from the station description.