Many thanks to Nickolay (nsh) for creating an updated Sphinx 8kHz Acoustic model based on  the VoxForge audio corpus:

voxforge-en-r0_1_3.tar.gz 16-Feb-2010 13:34 15.3M

From the README file:

This is English Voxforge acoustic model for CMUSphinx decoders -
pocketsphinx and sphinx4 version 0.1.3. It was trained with SphinxTrain.
The process of training is described in SphinxTrain tutorial.

This is 8kHz model, it could be used both for desktop and telephone
recordings.

To setup the files use build.sh script as a base. It should download
required files from Voxforge, setup structure and extract features.
Scripts are located in scripts subfolder.

Testing is performed with sphinx3. Language model for testing is built
from test prompts with SRILM. Result of the testing should be similar to

TOTAL Words: 44025 Correct: 40336 Errors: 4388
TOTAL Percent correct = 91.62% Error = 9.97% Accuracy = 90.03%
TOTAL Insertions: 699 Deletions: 930 Substitutions: 2759

To use this model with pocketsphinx, just provide it's location with
-hmm pocketsphinx option.For information on how to use this model with
sphinx4 see the documentation

http://cmusphinx.sourceforge.net/sphinx4/doc/UsingSphinxTrainModels.html

Remember that this model uses MLLT and it requires both frontend
modification. You need to use transform with LDA component and you need
to adjust melfilter parameters.