GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Audio Discussions
In the Thread: Initial thoughts about new/old Lamm ML2s

Post Subject: 6C33C matching: first proposalPosted by Lx_ on: 12/7/2009

Paul S wrote:

I also wonder about the supposedly simple but ever mysterious procedure for matching 6C33Cs for gain. Presently, my amps are ~ 2 dB apart.

To do gain matching we need a way to measure gain. To be precise, we need to measure a value that is representative of the tube's gain. We don't need the absolute gain value (though having it would not hurt and would do the job perfectly, it is not strictly necessary). Then by using the representative value for several tubes we can select two that match "closely enough" (quotes used to underline that this is too vague to be used as is).

We have a quite sophisticated measurement tool at our disposal, albeit an expensive one. It's made by Lamm himself and is called ML2(.1).

Here is a naive suggestion: why not use it to evaluate gain? We could apply a sinusoidal signal at the amplifier's input. Measure the voltage at the speaker binding posts (output signal) and divide by the input voltage. This gives the overall gain of the amplifier. Not the gain of the 6C33C, but if we change V2 tubes, we have a function (overall amp gain) of only one variable (the V2 tube) and that should be enough for our needs.

Some problems I can see with this procedure:
- I am not sure the output voltage can be used (the amp acts not only as a voltage amplifier but also as a current amplifier)
- I don't know what tolerance we can use to determine that two tubes match "closely enough"
- not repeatable in time: if we measure a batch of tubes at some point in time, there is no guarantee that it will measure the same later on, because other elements change in time (other tubes...). This change is probably minimal and hopefully negligible. It that's the case, when receiving a new batch of tubes, it is sufficient to measure only the new tubes, and they can be compared to ones that were measured previously. Otherwise an old tube should be measured again and old measurements corrected accordingly.

Regards,
Laurent

Rerurn to Romy the Cat's Site