What is the best way to reduce gain on a JAS 2.1 amp.
a) Increase negative feedback
b) Remove 12ax7 from circuit
c) some other method???
Some history... So I now have a preamp that is very high gain, it is a very quiet preamp however it still injects some noise "hiss", this coupled with an amp I'm running as a power amp (JAS) which already has a slight "hum" into high efficiency speakers (98db) needs tweaking. So I now have a bit of hiss and a bit of hum, I figure a reduction of gain in the amp is the way to go because it's the last component before the highly efficienct speakers.
I think it might be easiest to increase the negative feedback, but wouldn't mind removing the 12ax7 as I see it as unecessary.
The negative feedback curcuit has an 18k resistor and a 47uf cap in parallel, what factor does the gain decrease by with changing resistance and would I need to recalculate the capacitor in parallel?
I've attached schematics.
a) Increase negative feedback
b) Remove 12ax7 from circuit
c) some other method???
Some history... So I now have a preamp that is very high gain, it is a very quiet preamp however it still injects some noise "hiss", this coupled with an amp I'm running as a power amp (JAS) which already has a slight "hum" into high efficiency speakers (98db) needs tweaking. So I now have a bit of hiss and a bit of hum, I figure a reduction of gain in the amp is the way to go because it's the last component before the highly efficienct speakers.
I think it might be easiest to increase the negative feedback, but wouldn't mind removing the 12ax7 as I see it as unecessary.
The negative feedback curcuit has an 18k resistor and a 47uf cap in parallel, what factor does the gain decrease by with changing resistance and would I need to recalculate the capacitor in parallel?
I've attached schematics.
Attachments
The current gain is 18K/1K( feedback resistor/cathode resistor ). The 47pF cap in that is in parallel with the 18K determines the high frequency cutoff frequency and that will not change unless the 1K cathode resistor is changed in value. Your gain right now is 18K/1K ( the ratio of the feedback resistor to the cathode resistor ) it you reduce the 18K to a 10K then your gain would be 10 instead of 18 and your HF cutoff would be the same as there would be no change in the 47pF cap and 1K cathode resistor You would then have a gail of 10 instead of 18 with the same amplifier stability
Great, I'll start with a value of 10k then. That seems like a good place to start and half the gain might be in order.
Note*
If I remove the 12ax7 I'd of course reroute the negative feedback to the 6922. I'd also like to point out that this amp sounds better with the added negative feedback of going direct and bypassing the volume knob. Also pluging into the integrated circuit doesn't reduce the slight hum, but it woud allow for a reduction of the slight hiss, I'd like to minimize both, something I think negative feedback might do.
Note*
If I remove the 12ax7 I'd of course reroute the negative feedback to the 6922. I'd also like to point out that this amp sounds better with the added negative feedback of going direct and bypassing the volume knob. Also pluging into the integrated circuit doesn't reduce the slight hum, but it woud allow for a reduction of the slight hiss, I'd like to minimize both, something I think negative feedback might do.
Without a comprehensive stability analysis adjusting GNFB is the worst possible way to "adjust" gain of an amplifier.
If you think you have too high gain you should first analyze what the sensitivity now is ?
And you also should know the amount of GNFB used at the moment ?
According to the schematic the gain looks really high.
Also the number of amplifying stages inside the GNFB loop seems too high.
It would improve the amplifier if 12AX7 stage would be outside of GNFB loop.
What I do not fully understand is why one (6922) buffer stage is driving an other (300B) buffer stage.
Last edited:
I'll remove the 12ax7 too. I have an idea how to do this.
The 300b has a large voltage swing which is good for driving the 805, the 6922s drive the 300b by means of a cascode circuit. The 12ax7 is just useless.... too much gain, maybe good if I didn't have a preamp.
The JAS 2.2 uses a 15kohm and 47pf in the negative feedback, I'll follow suit and make adjustments as necessary. This without the 12ax7 will probably do perfect just like the 2.2 circuit.
BTW the preamp I'm using sounds fantastic and has a world class phono stage that's why I'm using a preamp with this amp, both amps are worth making work together and they sound great with the speakers too, just need to calm the noise floor down a touch.
Sometimes I think 45W of SET power is too much for these speakers. I think the amp was designed to run a less efficient speaker.
The 300b has a large voltage swing which is good for driving the 805, the 6922s drive the 300b by means of a cascode circuit. The 12ax7 is just useless.... too much gain, maybe good if I didn't have a preamp.
The JAS 2.2 uses a 15kohm and 47pf in the negative feedback, I'll follow suit and make adjustments as necessary. This without the 12ax7 will probably do perfect just like the 2.2 circuit.
BTW the preamp I'm using sounds fantastic and has a world class phono stage that's why I'm using a preamp with this amp, both amps are worth making work together and they sound great with the speakers too, just need to calm the noise floor down a touch.
Sometimes I think 45W of SET power is too much for these speakers. I think the amp was designed to run a less efficient speaker.
If you remove the 12AX7, you will need to move the feedback to the second state, which will require removing the cathode bypass capacitor on the second stage and attach the feedback there. However, this introduces a 180 degree phase shift so you need to change the tap on the opt by swapping the ground to the 8 ohm tap and take the feedback from the 0 ohm tap. Then you have to bypass the first stage.
Not a good plan overall but doable.
Not a good plan overall but doable.
Swapping the AX for an AU should work fine, 100 voltage gain vs 17 of the AU.
might look to decrease the plate resistor to 100k aswell while you are at it.
12au7,12ay7,12at7 and 12ax7 are pretty much interchangeable with different gain, the lower gain ones having a slightly lower voltage tolerance and higher current capability, with some added linearity. i run 12au7's in place of 12at7's and it works nicely.
might look to decrease the plate resistor to 100k aswell while you are at it.
12au7,12ay7,12at7 and 12ax7 are pretty much interchangeable with different gain, the lower gain ones having a slightly lower voltage tolerance and higher current capability, with some added linearity. i run 12au7's in place of 12at7's and it works nicely.
Now I'm concerned, what about changing the tap on opt??? 0ohm tap??? Now I'm confused because I don't know where the output to 0ohm speakers exists unless you mean ground and that makes even less sense to me. Why would I change from where I'm drawing negative feedback now?
please before you try anything, put an 12au7 there.
and yes he means "ground" on the output.
I recommend going with a tube change to reduce gain. If that is not enough, you can remove the cathode bypass cap on the second stage to further reduce gain. This will improve distortion as well.
If you swap tubes, you might need to change the plate (anode) resistor to correct the bias point. Going to a lower mu tube will most likely require a lower value, which will further reduce your gain.
If you swap tubes, you might need to change the plate (anode) resistor to correct the bias point. Going to a lower mu tube will most likely require a lower value, which will further reduce your gain.
Ok, I see the phase shift that occurs with running grids off cathodes... never thought or realised that before (I'm still learning
)
Yes my 0ohm tap is grounded (double checked) and that seems confusing to me, if I where to run the GNF signal off that would I need to "lift" ground first?
Also tried the 12au7, and it's a step in the right direction, I found one to best my sylvania 5751 bp. So Thanks for that tip.
Comments on Negative feedback are; Sound becomes harder and more two dimensional but crispier, amp designer has found sweet spot already. I ran a variable resistor and when I dialed it back from 10k I came to about 15-16k to regain some sweetness.
)Yes my 0ohm tap is grounded (double checked) and that seems confusing to me, if I where to run the GNF signal off that would I need to "lift" ground first?
Also tried the 12au7, and it's a step in the right direction, I found one to best my sylvania 5751 bp. So Thanks for that tip.
Comments on Negative feedback are; Sound becomes harder and more two dimensional but crispier, amp designer has found sweet spot already. I ran a variable resistor and when I dialed it back from 10k I came to about 15-16k to regain some sweetness.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.