Created attachment 236 [details]
full code exhibiting the problem

Hi,
I study propagation of low energy (20 keV) electron in the 0.5T magnetic filed in the volume filled with low pressure (1/25 atm)  H2 gas.

To illustrate the issue I took TestEm4 example and made only minor modifications:
- set volume to be larger box
- change material to be low density H2 gas
- added uniform B-filed of 0.5T parallel to the Z-axis
- Particle guns fires e- at large angle vs. Z axis

To test all works well for high energy electrons I set the gun to fire 3 30 MeV e-,
and got reasonably curved tracks (see fig 1).

Next, I change only energy of electrons to 30 keV and got this nonsense, shown in fig 2. That low PT tracks should spiral tightly around the z-axis (blue).
However, the want to jump transversely by many cm, many times. They seems to spiral only  between jumps.

Fig 3 shows zoom in of 30 keV e- transport. You can compare the size of jumps with the 10 cm long axis.

The whole code is also attached. I noticed that if I disable:
 ph->RegisterProcess(new G4eIonisation,         particle);
the problem is gone. But I do want to have ionization losses in my model

Please advice how to force Geant to propagate those low energy electrons correctly.