Preamp Gain stage Impedances

[lowell]

If I'm running 1st preamp tube 12ax7 w/ standard value resistors - 1.5k and 100k, then through a coupling cap and into another stage without a tone stack, what have you all found the difference in frequency and feel when the 2nd stage input impedance is played with. Say goin into a 1M grid leak vs a 100k grid leak. From what I've learned impedance matching is good, but sometimes it's ok to go low to high too. Technically the output z of stage 1 is around 35k so even 100k is high when considering z matching. Just interested to hear some thoughts.

[tubeswell]

Hi lowell

If you get Merlin Blencowe's chapter on Gain Stages from the Valve Wizard site, you'll see what he has to say about impedance bridging (as opposed to matching). He also has some formulae for calculating input and output impedance, and a recommendation is that the source impedance should be about 1/5th of the amount of the input impedance of the next stage to preserve a good signal. (If you get his book, the next chapter talks about coupling between stages and AC impedance etc.)

Having read and re-read this material about 20 times, my take on it is that the output impedance of a particular stage is like the 'numerator' on a voltage divider, and the input impedance of the next stage is like the 'ground leg'/denominator of a voltage divider. If the value of the numerator is low compared to the ground leg, then attenuation between the stages will be less. What is important is the impedance to AC, because it is the signal that is being affected. So you need to calculate the AC load line (as opposed to the DC load line) of each stage to understand what each stage is putting out, and hence what the attenuation will be going into the next stage (and thus be able to understand what the next stage would be putting out).

The frequency of each stage is a product of the input impedance in combination with the miller capacitance of each stage, which you can roll-off with various amount of cathode bypass capacitance, and 'comb' up with various amounts of coupling capacitance. However the coupling capacitance of AC coupled stages also interacts with the impedance bridging between stages to produce a particular frequency response.

The coupling cap also contributes to excess grid current in the following stage if the coupling capacitance is too high to allow the grid leak resistor of the following stage to dissipate excess grid current. So you can either deal with this by reducing the value of the coupling cap, or by reducing the size of the grid leak resistor. The former will cut some of the bass frequency band from being passed to the next stage, and the latter will attenuate the signal going into the next stage. As to whether 100k is 'alright', depends on the freq you have from the earlier stage and what the miller capacitance is of the second stage etc. It will be less 'noisy' than 1M

Don't know if that is any help.

[Robert M. Martinelli]

....From what I've learned impedance matching is good, but sometimes it's ok to go low to high too...

What tw ( and Merlin ) said.

Basically, a correct impedance matching is essential only when transferring power, to ensure that the maximum possible power is transferred from one point to another with the lowest possible losses.

When transferring a signal, which has voltage but practically no current capability ( = we're not transferring "power", at least not with its most common meaning ) impedance bridging is the way to go, its main purpose usually being to keep the preceding stage from being loaded too heavily ( this is especially true with passive, high impedance guitar pickups, which find their ideal "companion" in a tube amp's input stage and its very high input impedance ).

Cheers

Bob

[MWJB]

You know the value is going to be <1Meg, so use a 1Meg trim pot, wired as a variable resistor, for the grid load and set by ear. Be sure to play at a variety of settings & volumes. When you're happy that you have the right value, measure pot (without disturbing the setting) & replace with a fixed resistor.

[Old Tele man]

Basically, a correct impedance matching is essential only when transferring power, to ensure that the maximum possible power is transferred from one point to another with the lowest possible losses.

...+1, and an example of non-matched impedances are simple filters and waveshapers.

[lowell]

thanks for the replies guys. i brought this up cause it seems that in/out z is another variable that can be tweaked in a tube preamp and cause different tonal and feel results. it's something to consider when modding/tweaking/designing an amp for a certain sound, ya know aside from the power supply, Ra, Rk, coupling caps, grid stoppers, tube type, plate voltages, output class etc etc....