08 February 2013

Amps: Retubing Marshall 1923C Combo - Part 4

PI tubes test

During my previous tests it was very clear that this amp loved JJ ECC83S tubes and, even that I liked how it sounded with a Sovtek 12AX7-LPS in the PI position (V4), I wanted to check if it made any difference to use the JJ ECC83S instead.

Well, the JJ is darkening a tad the sound, seems to have more body than the LPS but, the amp went more noisy and was somewhat distorted, even clean.
The LPS brings an opener sound but lacks some body that the JJ had. The LPS was always more defined in speedy riffs. Difficult election.

I guess the issue is that none of those tubes were checked for triode balance and, probably, the LPS has a better balanced pair of triodes, while the JJ should be a bit unbalanced, what introduces crossdistortion in the power stage.

I think, I have to source some triode-balanced tubes to test the PI with better criteria. In the meanwhile, the LPS I've got works better than the JJ I've got.


To me, the best test is to try the amp making some music (or noise) and, therefore, I did a video during my tests with some backing tracks behind.
I've covered "Shine on you crazy diamonds - Part 1" by Pink Floyd (not the whole song!) in first place, to check the goodness of EQ, settings of pedals and, overall loudness. Then, I was improvising over three blues backing tracks, always in the classic / clean channel and, adding Wampler effects during the test.

There is no audio post-processing using Pro-Tools or any other DAW. TSound and image were taken with a Zoom Q3HD, with the mic set to Low Sensibility, to avoid peaks and, therefore, compression.
In the video editor, I've used a compressor to raise the loudness (NOT TO MAXIMIZE), while preserving full dynamics and, being sure that the sound corresponded to what I heard live while playing.
After the compressor, just a brickwall limiter, to avoid overs when the sound will be converted to MP3.
No corrective EQ, delay or whatever else audio effect added. I could make it to sound nicer but, I always work in the same way, I want to be sure that the sound in Youtube goes so close as possible to the sound I am hearing in the room while playing.
Notice that I am not using any Noise Gate, not in the pedalboard and neither in the video editing (as a plugin). I don't like to fake the sound, What you hear in the video is what you would hear in the room.

Update 09/02/2013 - Swapping V3 and V4

Reading my notes about the Sovtek 12AX7-WC and, the information that New Sensor gives in its FAQs page I wanted to give this tube a new oportunity, after that V1 and V2 were clearly determined for JJ ECC83S tubes.

As per Myles Rose comments, the WC is a very close to specifications tube and, as per New Sensor comments, WA, WB and WC tubes work really nice in cathode-follower positions and, the WC specially good in PI positions, since it comes with well matched triodes. So Iwanted to test a couple of 12AX7-WCs fin V3 (Cathode-follower) and V4 (PI). I don't like none of those for tone positions.

So, the pre-amp was configured as:

V1: TAD 12AX7-Cz (tested JJ ECC83S)
V2: TAD 12AX7-Cz (tested JJ ECC83S)
V3: Sovtek 12AX7-WC
V4: Sovtek 12AX7-WC

Results are really good. The amp remains quiet, more or less same sensations than with the last configuration (TAD 12AX7-Cz in V3 and Sovtek 12AX7-LPS in V4). Maybe, slightly darker and, maybe, a bit less tridimensional but good enough. Since my LPSs didn't lasted for long in PI positions, I think this solution can make a very affordable set of pre-amp tubes with great results.

I am happy to see that I've got one more tube type available for PI positions (the Sovtek 12AX7-WC), among with a good NOS JAN/Philips 5751 (but raising the headroom), the Sovtek 12AX7-LPS and the Mullard 12AX7 reissue.

I am also happy to have more options for that cathode-follower positions (The Sovtek 12AX7-WA, -WB and -WC) among the Chinese 12AX7A-C and JJ ECC83S.

The series WA, WB and WC are designed for very low noise and microphonics so, its a good thing, indeed.

No comments:

Post a Comment

Please, feel free to add your comments.