tideteam/2006_ions_summary contains several graphics that show the results of -3400V on the start MCPs. In no particular order, these are:
1) SC+TOF102506.pdf shows the count rate of the start converts and the direct events for 25 Oct., 2006. The start converts on the left and the direct events on the right. The point here is the high direct events for the low count rates for the start converts. Note that these two panel are not one to one in terms of time but do go together.
2) SC_compare96v06.pdf compares start convert count rates from 31 March, 1996, a time when the mass separation was working, with the rates from 6 October, 2006 shortly after the sart MCP voltage was raised to -3400V. Although the channels through which the ions are entering the instrument are different, the peak counts are about the same.
3) TOF_Compare96v06.pdf compares the direct events for the two dates above, 1996 on the left and 2006 on the right, as above. Note the differences in the levels of the direct events and in the background counts. There may be evidence of more than one type of ion in the 2006 spectrograms around 14:09 UT. There are both cold and warm ions present.
4) t0611190752_0852_2ion.pdf is a spectrogram of stops in energy density showing definitely that multiple (at least 2) ions are present.
5) t0611190752_0852_2ion_Hz.pdf is the same thing as above but showing count rate.
6) t0611190000_2400.sg_de.pdf is the direct event data for 19 November, 2006 as in the previous two graphs. No evidence of multiple ions.
7) de_thumbnail_1996.html is a thumbnail summary of direct events from March to December, 1996, one Spectrogram a month.
8) de_thumbnail_2006.html is a thumbnail summary of direct events from 2006, one Spectrogram a day from 27 September, 2006 to 16 October, 2006.
The daily direct events spectrograms from 27 September to the latest available are posted at tideteam/2006_ions/spectrograms/direct-events. If you scroll through these, it appears that the hydrogen signature has faded away since 6 October and the background appears to have gotten higher.
Given the behavior shown in these graphs, it appears that running the start MCPs at -3400V is showing no positive results.