I actually haven't read the AX84 posting about the paralleled 12AU7s but I'm trying to figure out why paralleling an already high current(for 12a*7) tube needs to be paralleled. You only gain mA output when paralleling and I see you have only a gain of 12 for the stage. You can have the loss in the tone stack if you make up for it in the next stage. A higher mu tube would do that. What is the reasoning for paralleling a 12AU7? Usually, you don't need that much current in the preamp. That's what the power tube is for. Maybe I'm missing something here. I have never seen this anywhere else.
I don't think there is any technical reason for paralleling the 12AU7 or using it in that position, other than perhaps adding a bit of "seasoning", it may very well fatten up the tone a bit, but a small turn on the T/S does the same thing...
If you really want to take advantage of the paralleling effect, you can use the pair as the CF, since their output impedance would be even lower than the single 12AX7, then you can use the 12AX7 post-T/S for more gain. OTOH, if you like the soundclip of the Blues preamp (please post a link), you may just want to build it as is...
If you really want to take advantage of the paralleling effect, you can use the pair as the CF, since their output impedance would be even lower than the single 12AX7, then you can use the 12AX7 post-T/S for more gain. OTOH, if you like the soundclip of the Blues preamp (please post a link), you may just want to build it as is...
It's not a need for current. The explanation I saw was that it "thickens" the tone. My interpretation of that explanation is this:
The two are biased slightly differently (one with a red LED, one with a green LED) and thus their voltage drops will be a bit different. I think this will result in two very similar but slightly different tones being produced and thus the perception of a "thicker" tone.
That's what makes sense in my head, though I'm not sure that's the correct explanation.
The two are biased slightly differently (one with a red LED, one with a green LED) and thus their voltage drops will be a bit different. I think this will result in two very similar but slightly different tones being produced and thus the perception of a "thicker" tone.
That's what makes sense in my head, though I'm not sure that's the correct explanation.
Jazbo, I was trying to think of how to handle the tonestack issue after boobtube's post. I like the idea of swapping the position of the 12AU7 with the 12AX7. I think that will work much better with the voltage gains I'm trying to achieve.
As for clips, they're hard to track down because a lot of them are old and the links no longer exist. Here's one that's an ES335 with the blues preamp and 20W SE (I'm guessing a KT88). That's the only one I can seem to track down that's still up. I had heard one other a while back that was really terrific.
As for clips, they're hard to track down because a lot of them are old and the links no longer exist. Here's one that's an ES335 with the blues preamp and 20W SE (I'm guessing a KT88). That's the only one I can seem to track down that's still up. I had heard one other a while back that was really terrific.
Yep, you can achieve thickening of tone and add more signal recovery after the tone stack with a 12AX7 paralleled using the same biasing scheme.
Since you are willing to veer off the Blues preamp design, you may want to check out the Matchless Clubeman 35, which kinda reverses the order of the Blues preamp, it has parallel 12AX7 up front follow by a less lossy T/S then a pentode gain stage. I think it sounds pretty good.
Matchless Clubman 35 Demo
Matchless Clubman 35 Demo
Last edited:
It is a point to consider, having the pentode later on. If it is noisy, as pentodes can be, then the noise will not be amplified as much. I think this was part of the reason Vox changed the AC15 and 30 later on was because the partition noise was amplified more because the EF 86 was at the input.
Another way to get more gain would be to reduce R19 from 220k to 20k and/or R22 increase to 680k.
That's the problem, there is a thousand ways to do these tube amps.
Another way to get more gain would be to reduce R19 from 220k to 20k and/or R22 increase to 680k.
That's the problem, there is a thousand ways to do these tube amps.
Last edited:
Actually, I was kind of feeling that the 12AX7 in the second stage was more than I wanted before the tone stack. I'm thinking that putting the paralleled 12AU7 there and then moving the 12AX7 after the tone stack to recover some of the loss may be the right move. If the 12AU7 isn't enough in the second stage I could try something else (12AT7 or 12AV7 maybe).
Of course, your solution is simpler. I'll think on it some.
Sold the guitar last night that's going to finance this, but I'm not in a huge rush to start building (I mean, the emotional side of me, of course, wants to build it yesterday, but the rational side of me wants to make sure I design and build it right!). I'm also doing a little amp surgery on my current amp (replacing 3 caps in the tremolo circuit, replacing the reverb tank, output transformer and speaker) so that's taking up a bit of my time as well. Still waiting on the output transformer to arrive.
The 6U8 has been used as a 1st stage preamp in a handful of amp designs I've seen and I don't believe it's much of a problem.
Valve Wizard mentions the Vox pentode preamp on his small signal pentode preamp page and said that the noise was due to a 200+ gain, which is why he recommends keeping the gain 100 or less for pentode preamps. I think as long as you're doing a head instead of a combo and keeping the gain reasonable, it'll be okay.
Also, part of the reason I did a 2 channel amp is that I wanted one of the channels to be a pentode preamp and the other to be a standard 12AX7 preamp so that I could have those two distinct tones.
That's right, I forgot you were building a head. You should be fine. And you're correct about keeping the gain down as well. The biggest problem was all the rattling going on with the speakers in the Vox and of course it being right at the input. That will be interesting to see what difference the pentode channel has. You'll have to keep us informed! It's also interesting seeing your design evolve as problems arise and knowing when to stop changing things when it isn't a problem.
A quick note on the insertion loss for the tone stack. I have a Hot Rod DeVille and according to the schemo, I have 324mV going in to the stack and only 42mV left after it. That's just 13% of the signal left.
A quick note on the insertion loss for the tone stack. I have a Hot Rod DeVille and according to the schemo, I have 324mV going in to the stack and only 42mV left after it. That's just 13% of the signal left.
Last edited:
Wow! Very interesting on the TS insertion loss in the Hot Rod DeVille... I had no idea it would be such a huge loss until I played with the Tone Stack Calculator. It's got me thinking about a tone stack bypass option...
I'm really excited to build the amp. My wife's uncle is a ham radio guy and used to own a Radio Shack in my family's home town. He doesn't know a great deal about guitar amp design, but he's got experience fixing them (and building ham radios and such and lots of tube experience). So he can help me with the build and trouble-shooting. Nice to have a resource like that relatively local.
I really appreciate all the help I've gotten here with the design as well. It's been an amazing resource!
I'm really excited to build the amp. My wife's uncle is a ham radio guy and used to own a Radio Shack in my family's home town. He doesn't know a great deal about guitar amp design, but he's got experience fixing them (and building ham radios and such and lots of tube experience). So he can help me with the build and trouble-shooting. Nice to have a resource like that relatively local.
I really appreciate all the help I've gotten here with the design as well. It's been an amazing resource!
The T/S plays a very important role in any amplifier, our ears are already "pre-conditioned" by thousands of recording done using Fender, Marshall, Vox, etc. tonestacks, so despite their lossy nature, you can't really do without them, except for some very low-power amps that sound good fully crank'd...
Last edited:
Sorry for my absence. This has by no means been put on the back burner. We went out of town last week and this week has been busy with my wife's birthday, mother's day, and other family related stuff. I also got a bunch of parts for fixing up my current amp (I've replaced some caps, the speaker, and reverb tank and I'm waiting on some parts to hook up the new output transformer. Practically a brand new amp.) So I've been busy.
Latest Schematic
I got a little time to redo the layout. I now have the stages as follows: 2 channel 1st stage pre-amp, 2nd stage paralleled 12AU7s with 6U8 triode cathode follower, followed by tone stack, then 3rd stage 12AX7, then power amp.
I've messed with some of the biases... So here's the step I'm stuck on (and we've briefly discussed some aspects of this, but I'm still trying to get it straight in my mind). We'll start with the 1st stage 12AX7 channel (easy & standard).
The way I have it biased currently, the voltage gain should be about 40. We'll call the input signal 1v P-P for the sake of values to work with... So, if I understand that correctly, that means we're looking at a roughly 40V P-P voltage swing at the output.
The second stage is the 12AU7 which are biased somewhere in the area of -1.6 and -1.9V. So does that mean I want to get my 40V voltage swing back down around 1.6V with a voltage divider? So, would I want to put say a 10K resistor in R22 (the 220K was just a value I threw in with no idea what I wanted at the time)?
With a 40V P-P input, going through the VR9 and R19 we're looking at a voltage swing of roughly 0.5-1.7V, depending on VR9.
Am I approaching this correctly?
Latest Schematic
I got a little time to redo the layout. I now have the stages as follows: 2 channel 1st stage pre-amp, 2nd stage paralleled 12AU7s with 6U8 triode cathode follower, followed by tone stack, then 3rd stage 12AX7, then power amp.
I've messed with some of the biases... So here's the step I'm stuck on (and we've briefly discussed some aspects of this, but I'm still trying to get it straight in my mind). We'll start with the 1st stage 12AX7 channel (easy & standard).
The way I have it biased currently, the voltage gain should be about 40. We'll call the input signal 1v P-P for the sake of values to work with... So, if I understand that correctly, that means we're looking at a roughly 40V P-P voltage swing at the output.
The second stage is the 12AU7 which are biased somewhere in the area of -1.6 and -1.9V. So does that mean I want to get my 40V voltage swing back down around 1.6V with a voltage divider? So, would I want to put say a 10K resistor in R22 (the 220K was just a value I threw in with no idea what I wanted at the time)?
With a 40V P-P input, going through the VR9 and R19 we're looking at a voltage swing of roughly 0.5-1.7V, depending on VR9.
Am I approaching this correctly?
Last edited:
Pretty much, but the typical input level is around 100mV peak, if you use 1V, you will end up with too much inter-stage attenuation. Also, it is best to work backwards from the output stage to the input stage for calculating the proper gain distribution.
I like the idea of running the gain stage at 2 bias levels via LED.
does this add 2nd harmonic Distortion more evenly?
does this add 2nd harmonic Distortion more evenly?
jazbo, thanks! I see what you mean by working backwards. That makes a lot of sense.
I've been very confused by the voltage coming out of the guitar. I've seen numbers ranging from 100mV to 10V (Wikipedia claims 10V P-P with a hard strum of all 6 strings on humbuckers. That seems a bit far-fetched to me).
I'll work with 100mV as you suggest.
stocktrader200, it's been mentioned before in this thread, but the dual 12AU7s is to "thicken" the tone.
So I'll work on the math and try to get this all worked out, hopefully today or tomorrow.
I've been very confused by the voltage coming out of the guitar. I've seen numbers ranging from 100mV to 10V (Wikipedia claims 10V P-P with a hard strum of all 6 strings on humbuckers. That seems a bit far-fetched to me).
I'll work with 100mV as you suggest.
stocktrader200, it's been mentioned before in this thread, but the dual 12AU7s is to "thicken" the tone.
So I'll work on the math and try to get this all worked out, hopefully today or tomorrow.
According to an author and active member of our forum, the humbucker is 100mV RMS strumming within the first second. After that, it obviously decays. You can get up to 1 volt with active pickups. So 100mV peak seems right. If you end up with too much signal, add attenuation later on in the design process.
Jazbo8 repeats almost any of the processes I have read about, work from the power section back to the preamp. You know from the power tube's data sheet what voltage swing you need to arrive at to drive them properly, so start from there.
Jazbo8 repeats almost any of the processes I have read about, work from the power section back to the preamp. You know from the power tube's data sheet what voltage swing you need to arrive at to drive them properly, so start from there.
Oh, you give me too much credit...
lol.
I haven't updated the schematic, but I believe I want a 330 ohm cathode resistor on the 6L6GC. With 320V on the plate that should have me at 84mA plate current, 12mA screen current, and a grid voltage of -32V. That should give me a plate dissipation of 27W.
So does that mean I'm looking for a 64V swing or a 32V swing? I'm thinking 64V P-P, but I want to verify before I start figuring all the math and find out I made the wrong choice.
Or maybe not (with regards to the biasing). I get confused by a lot of the calculators out there and I think it's because they're not taking into consideration the screen grid voltage.
Looking at the data sheets Average Transfer Characteristics (interpolating between the Eb=250V and Eb=400V graphs and Ec2 at about 315V) I believe I want something closer to -19V for the bias which puts the cathode resistor closer to 200 ohm.
I definitely need help...
Looking at the data sheets Average Transfer Characteristics (interpolating between the Eb=250V and Eb=400V graphs and Ec2 at about 315V) I believe I want something closer to -19V for the bias which puts the cathode resistor closer to 200 ohm.
I definitely need help...
- Status
- Not open for further replies.