ron soyland
There are over a dozen different types of common vacuum gauge types used to measure vacuum from moderate vacuum such as that in vehicles to high vacuum such as that used in tube making. For tube making a set of gauges is necessary both for system functionality monitoring and for testing the newly made vacuum tube.
The chart above shows the ranges of several types of commonly available pressure gauge types. The top scale is in pressure and covers the range from where vacuum tubes operate to atmospheric pressure. The first thing to note is that there is no single gauge type that will cover the entire range. Back in the 50's there was a gauge called an Alphatron that would cover from 10 -6 torr up to atmospheric. This used a fairly strong radioactive Alpha particle source to ionize the gas in the system. This radioactive source required a license to own so the gauge was not particularly successful and was discontinued. Nothing to replace it has been produced since. Note that the pirani gauge can come close, but there is a problem which we will examine.

This range is considered moderate vacuum and it is most accurately measured with direct reading instruments.
In a DIRECT READING gauge the pressure directly affects the gauge mechanism. Examples are bourdon tube gauges, capacitance manometers, McCloud gauges.
In an INDIRECT READING gauge, the pressure causes a change in some parameter that is then measured to indicate the pressure. Examples are pirani gauges and thermocouple, where changes in heat conduction with pressure cause temperature changes that are measured, Ion gauges, where the residual gas in the vacuum is ionized and the current through the ions is measured. In each of these cases the measurement is indirectly coupled to the pressure. Problems are abundant with this kind of gauge.

The most basic of direct reading gauge is the bourdon tube gauge which makes up the common meter type mechanical vacuum gauge. These gauges are only good for rough readings of vacuum, such as in refrigeration systems. They are very nonlinear when the vacuum goes below about 30 torr.
The capacitance manometer is the current choice for direct reading gauges. The baratron (trade name of MKS corp.) is the most common form of this kind of gauge, but many other manufacturers make them as well. The baratron is specified over a three decade range for the basic models, and up to five decades for the precision models. Thus, a baratron that has a full scale range of 1000 torr will read accurately down to 1/1000 of that or 1 torr. This gives an accurate reading of the mechanical pump over that range.
However, in tube making pressures above about 100 torr are not significant, so it is wasting range to have a gauge to go all the way to atmospheric pressure at 760 torr.
A 100 torr full scale baratron would be accurate down to 100 microns, which would give a good range for uses like making nixie tubes.
For vacuum tube making, pressures above 1 torr are not significant so a baratron with a full scale reading of 1 torr would give accurate readings down to 10 microns, which is the limits of most mechanical pumps. Note that the 1000:1 range is for the standard models. The precision models, like the model 121, have a 5 decade, adjustable to 6 decade, range! These only are available in 1 torr full scale range so the lower end would be 5 decades lower or 1/10 micron, 10 -4 torr. These instruments are so accurate that they can be considered secondary pressure standards. They are used to calibrate all other types of gauges, including Mc Cloud gauges. This is the holy grail of gauges for vacuum tube work and is the kind of gauge I have on my system. They are commonly available on ebay for under $100 U.S.
The baratron is an absolute pressure reading gauge. It is not affected by gas type like indirect reading gauges. This means if you read 10 microns, you have 10 microns, regardless of the gas mixture in the vacuum.

With the advent of semiconductor strain gauges a new type of gauge has come into use. The gauge has a small diaphragm with the strain gauge mounted to it. The operation is dependent on the pressure deforming the diaphragm which distorts the strain gauge. This is a direct reading instrument like the baratron. These gauges are commonly specified to about 50 microns because they are used primarily for the refrigeration industry. The readings below 100 microns are rough and not accurate, being off 25 to100% at 50 microns quite common. The instruments are not designed for the low end. They are quite accurate above 100 microns and are ideal for nixie tube gas pressure readings. These gauges are available on ebay for under $100, new.

The Mc Cloud gauge is one of the oldest types of vacuum gauge. They have been designed to measure from below 1 micron to fifty torr or greater. They are direct reading instruments but have a fatal flaw that can throw the readings off by 100% or greater. The Mc Cloud gauge works by compressing a calibrated volume of the gas from the vacuum being measured into a closed chamber that is carefully calibrated. The method of compresson is the use of mercury. The fatal flaw with the gauge is at the low end of the readings. If there is water vapor present in the system, (usually is) then when the gas sample is compressed, the water vapor may condense causing a nonlinear pressure indication, off by easily 50% or more. Laboratory Mc Cloud gauges have a silica gel dryer cartridge that is in the connecting line to extract the water and keep it out of the gauge. This also helps to keep the mercury vapor from migrating out of the gauge into the vacuum system. Mc Cloud gauges are absolute measurement instruments that can be used to calibrate other types of instruments, but they are not suitable for routine use because of their clumsy method of operation.

The pirani gauge is an indirect gauge that uses the temperature change in a hot wire to indicate relative pressure change. If a piece of wire with a high coefficient of resistance change with temperature is placed in the vacuum sample, the denser the gas in the sample the more heat that is conducted away from the wire, thus causing a resistance change that is directly proportional to the pressure. More or less. The problem is that all gasses do not have the same heat conduction coefficient. Thus you get a different reading for the same pressure for hydrogen and air, or neon and air. This can be a serious error, upwards of 50% depending on what gasses are in the vacuum being measured. This is the main disadvantage of this kind of gauge. Another disadvantage is that the gauge is affected by ambient temperature. Since the gauge is directly reading the temperature change of the wire, any cause for the change causes the reading to shift. Elaborate correction schemes are used in modern instruments to correct for ambient temperature changes.

These operate identically to the pirani gauge except that a thermocouple is fastened to the wire to sense the temperature change. This doesn't in any way correct for the gas difference problem so these gauges are not accurate if the gases in the vacuum are anything but pure air. The readings also are seriously affected by ambient temperaure changes in spite of correction schemes.

The ideal gauge for the vacuum tube system is a MKS model 121 baratron 0-1 torr range.
This gauge gives an excellent accuracy in the 1 micron to 1 torr range, which is the range of operation of gas tubes you would make.
 The second choice would be a model 221, 0-1 torr unit which is accurate to 10 microns. This would still be accurate enough to make gas tubes, but it is not a pressure standard and will thus drift with ambient temperature.
There are other models of baratrons that are equally good and can be used if available at a good price. Ebay is the source. Be sure you can return the item if it turns out to be defective, because many sellers are selling junk from company storerooms that is in unknown condition and can be bad.

When the pressure gets below about 1 micron, direct effect of the gas on a physical sensor like a diaphragm becomes too small to be effective. A different approach must be used to measure the vacuum.
The most common is the measurement of ion current. The vacuum is composed of freely floating gas molecules at pressures below about 1 micron. That is, the gas molecules are far enough spaced that they are considered not touching each other.
If the gas molecules are ionized (given an electrical charge) then the flow of the charge to a sensor probe that has the opposite charge will create a current that is in proportion to how many molecules are in the area. This is assuming that the charge on each molecule is the same, which it is at the voltages used in the ion gauge tubes.

There are two types of ion gauge tubes commonly available. One is the Bernert Alpert type and the other is the Penning gauge. They both ionize the gas in the vacuum and measure the ion current, but they operate slightly differently. This difference is not important here.
The most widely used is the Bernert Alpert type gauge. It is accurate from approximately 1 micron down to 10 -6 torr without bakeout. This is completely satisfactory for vacuum tube making.
The Penning gauge uses a magnet to cause the ions to travel in a longer path before getting to the sensor probe. This gives a linear response over a wider range of pressure. However, the Penning gauge uses high voltage to ionize the gas. This has a serious problem of creating an electrostatic field that attracts molecules of all types. This is what it is supposed to do but the side effect of this is that it also attracts diffusion pump oil vapor that has backstreamed into the system. This oil is a heavy molecule that is disintegrated by the high voltage field, leaving a coating of silicon on the gauge elements.This causes severe inaccuracy to develop. These gauges are not suitable for use in a diffusion pump type system. They would have to be disassembled and cleaned every few hours!