Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: sdp1234 on September 09, 2014, 08:07:58 am
-
Tubenit and Geezer,
You two have done so many designs it makes my head spin. The TOS and a bunch of others are taking the preamp into a pentode before driving the output section. Just wondered if you have ever seen the raffler amp from the AX84 site? It was a cool SE design where the Pentode was the last stage in the preamp and also mixed the clean and higher gain channel. I enjoy reading what you come up with and I thought i'd share this because several years ago Carl B and some of the guys at AX84 were coming up with some different things also...
http://ax84.com/raffler.html (http://ax84.com/raffler.html)
-
As long as the circuitry which follows the preamp pentode doesn't need a very large driving signal, then using the pentode is probably fine.
Normally the highest gain stage is placed 1st in a preamp design, to get the desired guitar signal well above the noise floor. All resistors and tubes add some amount of noise; the longer you leave the desired signal small, you increase the "noise-to-signal" ratio because more noise sources are added along with the weak guitar signal.
So there's no "rule" against preamp pentodes late in a circuit. It's just that they normally have the highest voltage gain, so they are placed first in classic designs. Some of the guys here placed pentodes later in the preamp to sidestep occasional microphonic pentodes by giving them a larger signal to amplify; in turn, any microphonic disturbances tend to be smaller relative to the larger guitar signal at the later point in the preamp.
My first statement relates to driving an output stage:
Preamp pentodes need large-value plate loads to achieve high voltage gain. A large resistive load also drops a lot of supply voltage. If the tube also has to swing a big output voltage, there may not be enough total supply voltage to get high gain and a large output signal. The EL34 in the Raffle amp would be easier to drive than a 6L6 in the same spot; an EL84 output tube would be even easier to drive and present fewer challenges.
-
When this tread started I read it with interest,
as per my reading of the preamp books of Merlin, I developed an interest in the use of pentodes as V2
or other alternative position in the signal path that differs from V1
so I discovered amp circuits like the Selmer Zodiac Twin 30 http://www.mediafire.com/view/16wlb6wjd75mo49/selmer_zodiac_twin30_orig_151.pdf (http://www.mediafire.com/view/16wlb6wjd75mo49/selmer_zodiac_twin30_orig_151.pdf)
and the Matchless Clubman http://www.mediafire.com/view/1b67avkipay18pk/matchless_clubman.pdf (http://www.mediafire.com/view/1b67avkipay18pk/matchless_clubman.pdf)
also Tubenit developed some nice sounding amp using a pentode in the path that are very intriguing
Quite accidentally, yesterday I discovered that also Sony used a pentode (6AU6) in the signal path
as the last tube of the preamp, just before of the power tube
(http://i.imgur.com/tRpxHti.jpg)
Full Schematic:
http://www.mediafire.com/view/svt2tvw5lz13k5s/Sony_TC-500a_Schematic.jpg (http://www.mediafire.com/view/svt2tvw5lz13k5s/Sony_TC-500a_Schematic.jpg)
There is a direct connection to the pentode if the Line Input is used and the pentode is preceded by two triodes if the Mic Input is used
Maybe it would be worthwhile to do some testing of a similar architecture as guitar amp as in the AX84 Raffle amp
K
-
All of the old classic Sunn amps and the Ampeg Gemini series amps used a pentode just prior to the cathodyne PI. Some may say that's part of the power amp, but maybe it could also be called the last stage of the preamp.
-
You are right Steve :notworthy:
I focalized my attention to the SE amps and forgot to consider the PP amps
Sunn is a good example and as per the last amp tread of Pete, Dynaco was a precursor in this achitecture
Franco
-
Here's an example in a Portaflex:
http://el34world.com/charts/Schematics/files/ampeg/Ampeg_B12XY.pdf (http://el34world.com/charts/Schematics/files/ampeg/Ampeg_B12XY.pdf)
Jack
-
Quite accidentally, yesterday I discovered that also Sony used a pentode (6AU6) in the signal path
as the last tube of the preamp, just before of the power tube ...
There is a direct connection to the pentode if the Line Input is used and the pentode is preceded by two triodes if the Mic Input is used
A 6AU6 feeding a 6AQ5 (call it a 6V6 for the sake of familiarity). Roughly the same as a 5C1 Champ (http://el34world.com/charts/Schematics/files/fender/Fender_CHAMP_5C1.pdf).
The -10dB line level equates to a maximum signal level of 0.316v RMS input. We don't know the bias of the 6AQ5, but the data sheet condition for ~250v plate and screen has a bias of -12.5v. So for ~0.45v peak input (the line level expressed in peak volts), a gain of 12.5/0.45 = ~28 in the 6AU6 drives the 6AQ5 to full output. I won't bother trying to calculate it, but the naked 6AU6 probably has a gain in excess of 120.
But did you not see that mess of resistors and caps from the plate output of the 6AQ5, through SW1-5, back to the 100Ω at the 6AU6's cathode? That's a feedback loop, which 2 settings depending on whether you're using the line or mic input, which probably throws away most of the 6AU6's gain to increase bandwidth and linearize response.
I don't think the 2x 12AX7's stages are to be considered because they're for mic level signals (-55dB) and therefore ~1.8mV RMS (very, very much smaller than a guitar signal).
The Selmer and Matchless amps are more typical. The Matchless uses a 10kΩ plate load for the parallel-12AX7 input stage, which won't be making very much gain prior to the tone controls. The Selmer is more conventional, but follows its input stage with a bridged-T tone circuit, a regular treble-cut tone control, and a volume control, all before the 2nd stage.
My original point is that it doesn't matter so much where you use a pentode, so much as the input and output signals required at the pentode gain stage. Real circuit design isn't just slapping together circuit blocks, it's slapping together circuit blocks to achieve a known, desired end goal.