Hi all.
Just remembered that back on 4HV I documented a partial fix for “Bad” tubes.
http://4hv.org/e107_plugins/forum/forum_viewtopic.php?50094.post
http://4hv.org/e107_plugins/forum/forum_viewtopic.php?150298.post
The fix is based on a well known effect called a Zobel network, in this case a series resistor and HV
capacitor allows counting even if the tube is prone to oscillation and AFAIK is used in high end counters to reduce saturation at abnormal radiation levels by minimizing dead-time.
A neon in parallel with a resistor also works, interestingly (tested a week ago)
I’ve tested it with the SBM21 and 20, as well as CBT11A (ebay special, supposedly faulty) and all three work fine
notably the smaller tubes work far better if this is done.
It seems that adding this circuit allows the discharge pulses to bypass the resistor but also adds a measure of damping so
sub-threshold pulses which would normally be ignored are counted.
An additional modification is to add a radio frequency sensor around the tube, in the form of an AM RX coil.
This is tuned to around 65 KHz and can do double duty as an extra pulse detector as well as serve as an MSF/whistler antenna.