Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Toxophilite on November 19, 2015, 08:59:11 pm
-
I'm hoping to tweak the preamp on this Akai Roberts conversion I'm doing
I've done many of these and they work great for quite recording or getting distorted sounds at low volumes
I rework the signal path, get rid unnecessary bits , ad in a simple baxandall Eq but I've never messsed with the B+ supply nor the bypass caps and resistors on the 12ax7 cathodes
right now I have about 105vdc on pin 1 and 125 on pin 6 of the 12ax7
I'm wondering if there is anything I can do to increase the gain of the preamp 12ax7
I'll post a schematic of the conversion (minus the tone stack I put in between the 1st gain stage and the volume control) I'll also post the tone stack I'm using
On the inout to the 12ax7 'channel' I have 1 33k resistor and a 1meg to ground like a classic fender setup
THanks for looking
-
You can try lowering the value of the 1st 12AX7 section's cathode resistor from 3kΩ to as low as 820Ω.
You can't tinker the 2nd 12AX7 section's cathode resistor without disconnecting the feedback loop from the OT (though that alone would increase gain/distortion from that section through to the secondary). If the feedback loop was broken (disconnect R21, C18) and you still didn't have enough gain, you could lower that stage's cathode resistor as in the 1st stage and/or install a cathode bypass cap across the resistor.
There's quite a bit of tone-shaping on the EL84 plate. Did you leave that stuff intact or remove it?
-
The power amp is currently how it is on the schematic.
How could I monkey with that?
And would it affect the preamp out (which is just prior to the EL84)
I usually put a 25-50K pot in feedback circuit to give me more gain and tonal differences. It seems to impact the preamp out volume too.
I was wondering about the cathode resistors, I'll give that a try.
I use this unit for late night and quiet apartment recording.
-
This was an Akai product? So it was some kind of tape amplifier, maybe?
I'm not familiar with the function of this stuff, but there are a bunch of resistors & caps on the EL84 plate: C12, C13, C14, R18, R19 and C19, C20, R25, R26. These look like passive, fixed EQ to undo the EQ applied to mass-produced recordings.
- I don't know much about the EQ used on tape, but records had one of several pre-EQ curves applied (usually RIAA in the later days) to minimize hiss and keep the bass from making record grooves so big they cut into the adjacent groove. The pre-EQ boosts highs and cuts lows, so a phono preamp has an opposite-shaped fixed EQ curve to restore largely-flat response (this also cuts record hiss as a handy side-effect)
C17 may/may not be part of this tone-shaping as well. It would cut highs; often such a circuit is used across an OT primary, but with a resistor in series with the cap, to offset the rising impedance of the speaker with increasing frequency (keep things from getting too bright).
I mentioned the need to break the feedback loop to change R14, the 2nd 12AX7 section's cathode resistor: the complete feedback loop is from OT, through R21/C18, through R14 to ground. R21 & R14 form a voltage divider to set the feedback amount, while C18 reduces the feedback for higher & higher frequency signals (probably included more for phase correction given all the stuff hanging off the EL84 plate more so than for actual tone shaping). If you slap a bypass cap across R14, you immediately kill all the feedback (because it's now shunted to ground through the new bypass cap).
-
I'm actually working on a Roberts but same company as Akai, just a more US market friendly name post WW2 :icon_biggrin: I kid you not!
It's one of two modular preamp/single ended amps that came out of a stereo reel to reel deck.
Akai M-7 is the one pictured here.
There are about 3-4 models under both names that work with this conversion.
The Ef86 ends up bing a balanced mike pre or a straight line preamp separate from the 12AX7 'channel'
the 12ax7 I usually set up as a guitar pre and low power single ended amp
I did try putting a 1.5K resistor instead of a 3k on pin 3 with little or no discernable difference, if anything it lowered the voltage to the plate on pin1
-
Gotcha. The 250kΩ plate resistor is already a "high gain" change, stock. Looking at tube charts, I suppose the best gain increase you could get with the cathode resistor change is roughly an increase from 48 to 58. That may not be a "Wow!" change...
Still looks like your best shot is removing feedback loop, EL84 components and adding a bypass to the 2nd 12AX7 stage.
-
For fun I tried detaching the EL84 plate from that whole network of caps and resistors and there really wasn't a notice-able effect...and I pride myself on having good ears
I might have a listen in the morn when I can turn it up a little more
I did what I usually do and put a pot (in this case a 50K) between pin 8 of the 12ax7(2nd gain stage) and the feedback loop, enabling me to reduce the amount of feedback and that works well as it always has.
I was just curious if some monkeying with the B+ would help
But if it won't I can just stick to my normal thing
I really appreciate you taking the time to have a look at the schematic , It helped me do so as well which gave me a better understanding and eliminated a few questions I had.
If there are any other ideas out there it'd be neat to here them
These units make quite good bedroom amps/recording pres in case anyone comes across one
-
I was just curious if some monkeying with the B+ would help
But if it won't I can just stick to my normal thing
Good point; I didn't address that.
Look at page 2 of a 12AX7 data sheet (http://www.mif.pg.gda.pl/homepages/frank/sheets/127/1/12AX7WA.pdf), at the "Resistance Coupled Amplifier" chart. Look at the 90v column, 100kΩ plate resistor (listed as "0.1" MΩ) and 2000Ω cathode resistor. Gain is 38. Now skip over to the 180v column, same cathode resistor, and 0.24MΩ (240kΩ) plate resistor. Gain is now 54 (up 16, or 33%). So it appears if you double plate voltage and double plate resistance, you can get a 1/3 increase of gain.
Going up 120v to 300v plate, 240kΩ plate resistor, and 1.8kΩ cathode resistor only bumps gain up to 64 (an increase of 10, or ~19%). So it looks like small payback for big changes to plate voltage.
Your best best is killing the feedback by breaking the loop, then adding the bypass. Gain in the 12AX7 stage will double (or a bit more), while gain reduction due to feedback will be removed.
-
For fun I disconnected the feedback from the speaker...LOUD as expected no feedback louder, more mids!
Then for experiments sake I bypassed R14 I used a 22/25
It got way louder but also buzzed like crazy at even the slightest volume level, the nasty bzzz, not a hum.
At heart I like a clean scooped sound I was just wondering if I could get a little more poop out of this unit and was curious about B+ voltage
I think I'll stick to a pot on the negative feedback control essentially doing half of what you were talking about (removing some feedback)
However that's interesting about the relationship between gain and voltage on the plates.
Somehow I thought it was a more dramatic difference, especially when it's often said that one should remove the 12ax7 from the normal channel of a Fender amp to increase the headroom on the vibrato channel as you get more voltage to the plates, seems like it would be an infinitesimal difference. There's a lot of 'funny' information out there.
I really have to do more studying of theory
Thans again
-
... it's often said that one should remove the 12ax7 from the normal channel of a Fender amp to increase the headroom on the vibrato channel as you get more voltage to the plates ...
You're right there is some funny information out there, but mainly because people don't accurately quote the source they're citing. Gerald Weber first put that in print as far as I know. However, the point wasn't plate voltage (exactly), but cathode voltage.
For the Fender amps affected, the 2nd gain stages of the Normal and Vibrato channels share a cathode resistor (820Ω), which is about half the typical value (1.5kΩ) because 2-tubes' current flows through it. To bias the preamp stage the same, but with double-current, you need half-resistance. That lands at the same cathode voltage.
What Weber was saying was if you pull a preamp tube for the unused Normal channel, the bias is now different for that 2nd-stage still left in the Vibrato channel (true 820Ω instead of "effective 1640Ω"), which increases the current in that stage, lowering the plate voltage (because more current through the plate resistor = more voltage drop = less left at the plate), which is supposed to result in less headroom and more distortion.
When I tried it in old blackface Fender amps, I never really noticed a difference worth bothering with.
-
Yeah I don't think the blackfaces are like that, (or at least the ones I've look at the schematics for) I tried it on my 70s super and same thing, little or no dis-cernable effect
I've only seen the shared cathode resistor and bypass cap on the reverb recovery and gain stage after, but my experience is very limited!
This page puts forth the concept as well and suggests it for pretty well every fender amp I've ever looked up http://fenderguru.com/amps/vibrolux (http://fenderguru.com/amps/vibrolux)
-
I've only seen the shared cathode resistor and bypass cap on the reverb recovery and gain stage after, but my experience is very limited!
Most, if not all, blackface and silverface amps that have a Normal and Vibrato preamp use a shared cathode resistor/cap on the second preamp stage triode. Take a look at the schematics. It will be hard to find one that doesn't share. That's a lot of resistors and caps that Leo saved! Enough for an early retirement! :icon_biggrin:
-
duh! I was only looking at the 1st stage
It's the voice of inexperience talking