arnoldc said:
Yeah, pretty big. I wanted it to look like my power amp, and I had been stuck with too small a chassis before, so this time I played slightly more safe! Plus the output caps for the headphone amp section is pretty large, 470u 450V! 😀
Pengboon,
What transformer sis you finally decide on using... I remember that you have alot of trouble finding a good transformer for the project..... Also what is that piece of styrofoam under the fader knob?
Thanks!
Mark
What transformer sis you finally decide on using... I remember that you have alot of trouble finding a good transformer for the project..... Also what is that piece of styrofoam under the fader knob?
Thanks!
Mark
Mark A. Gulbrandsen said:
I used a 350-0-350 one salvaged from some's 300B project. It had 3 5V taps and a 14V tap. Tied 2 5V together for the 12B4 filaments and used the 14V for the 5687 filaments.
The "styrofoam" is actually this:
http://www.tnt-audio.com/clinica/sblocks1e.html
🙂
Brian Beck,
reading all of these threat took me whole days and still not clear for my beginner knowledge of Tube.
I just want to build 12B4 Preamp and it's very interesting reading all of this.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=68267&perpage=10&pagenumber=5
question:
1. Which component need to be change if I use CD player as a source?
2. any recommendation PSU for your schematic ?
regards
cal
reading all of these threat took me whole days and still not clear for my beginner knowledge of Tube.
I just want to build 12B4 Preamp and it's very interesting reading all of this.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=68267&perpage=10&pagenumber=5
question:
1. Which component need to be change if I use CD player as a source?
2. any recommendation PSU for your schematic ?
regards
cal
Cal,
I redrew the schematic (below), since there was some confusion over adapting my original posting into a line stage. The original design was for the filter output stage WITHIN a CD player, and it had an LC low-pass filter before the grid that you don't want for general line stage duty. The following schematic shows how make a line stage. Yes, it will work for any line level input, including a CD player. I omitted any input selector switches and balance control. I will leave that up to you. The 2.21M resistor after the volume pot is a safety resistor in case the volume pot’s wiper encounters dirt or becomes worn and goes open. This resistor keeps the grid grounded and prevents loud pops. I didn’t use a grid stopper resistor and I don’t like to use them unless they are necessary. The 12B4 doesn’t seem particularly prone to oscillation in the circuits in which I’ve used it, but as always, keep the wiring short in the grid circuit. If you think you may have oscillation problems, experiment with stopper resistors values in the range of 100 to 1K ohms placed very close to the grid. Or, to be safe, just throw 470 ohms or so into the circuit.
As to the PS, I don’t have anything but lab book sketches of what I built. I employed parts that I already had in my parts bins, and you would build it differently, and probably better. I used an L-input supply with fast SS rectifiers, flowed by an R-C filter, then a final L-C. The final C is large, 390uF at 450V. This is almost a power amp in terms of power needed. You’ll need a sizeable power transformer. Don’t skimp on it. I used a nice power transformer lifted out of an old mono Harmon-Kardon tube receiver. I used another LM317 to make DC for the heater and this is lifted to just above the cathode voltage level (say 20 volts) by a resistor divider off the B+, with a capacitor noise filter.
Reminder: This is intentionally a very high current stage, with almost 30 mA passing through the plate load resistor, the tube and the CCS. The resistor will become very hot. Place it away from other components and make sure it gets adequate ventilation, but in such a way that its leads are not exposed as a shock hazard.
You can reduce the plate current to about 22mA by changing the 43 ohm current-setting resistor to 56.2 ohms, and then increasing the plate load to 8.2K. I tried this and it sounds very nearly as nice, but the reduced current and power makes everything else a bit easier to deal with. In fact, my present line stage version is using the 8.2K option. Your choice of course.
The output impedance will be just under 1K. This should be able to drive most power amps, even solid-state (sorry) brutes. Gain will be about 14.5 dB (inverting), which is nearly ideal for a general purpose line stage.
Hope this helped. Enjoy.
I redrew the schematic (below), since there was some confusion over adapting my original posting into a line stage. The original design was for the filter output stage WITHIN a CD player, and it had an LC low-pass filter before the grid that you don't want for general line stage duty. The following schematic shows how make a line stage. Yes, it will work for any line level input, including a CD player. I omitted any input selector switches and balance control. I will leave that up to you. The 2.21M resistor after the volume pot is a safety resistor in case the volume pot’s wiper encounters dirt or becomes worn and goes open. This resistor keeps the grid grounded and prevents loud pops. I didn’t use a grid stopper resistor and I don’t like to use them unless they are necessary. The 12B4 doesn’t seem particularly prone to oscillation in the circuits in which I’ve used it, but as always, keep the wiring short in the grid circuit. If you think you may have oscillation problems, experiment with stopper resistors values in the range of 100 to 1K ohms placed very close to the grid. Or, to be safe, just throw 470 ohms or so into the circuit.
As to the PS, I don’t have anything but lab book sketches of what I built. I employed parts that I already had in my parts bins, and you would build it differently, and probably better. I used an L-input supply with fast SS rectifiers, flowed by an R-C filter, then a final L-C. The final C is large, 390uF at 450V. This is almost a power amp in terms of power needed. You’ll need a sizeable power transformer. Don’t skimp on it. I used a nice power transformer lifted out of an old mono Harmon-Kardon tube receiver. I used another LM317 to make DC for the heater and this is lifted to just above the cathode voltage level (say 20 volts) by a resistor divider off the B+, with a capacitor noise filter.
Reminder: This is intentionally a very high current stage, with almost 30 mA passing through the plate load resistor, the tube and the CCS. The resistor will become very hot. Place it away from other components and make sure it gets adequate ventilation, but in such a way that its leads are not exposed as a shock hazard.
You can reduce the plate current to about 22mA by changing the 43 ohm current-setting resistor to 56.2 ohms, and then increasing the plate load to 8.2K. I tried this and it sounds very nearly as nice, but the reduced current and power makes everything else a bit easier to deal with. In fact, my present line stage version is using the 8.2K option. Your choice of course.
The output impedance will be just under 1K. This should be able to drive most power amps, even solid-state (sorry) brutes. Gain will be about 14.5 dB (inverting), which is nearly ideal for a general purpose line stage.
Hope this helped. Enjoy.
Attachments
Brian,
wow... what an explanation, for me it's more than enough, I hope I don't have any question and I try to learn as much as I can before I ask you.
I appriciate your help and redrawn your schematic.
again
thanks alot man.
cal
wow... what an explanation, for me it's more than enough, I hope I don't have any question and I try to learn as much as I can before I ask you.
I appriciate your help and redrawn your schematic.
again
thanks alot man.
cal
Cal,
Ask any questions that you wish. Either I or someone else will surely answer you. That's what this forum is all about.
Ask any questions that you wish. Either I or someone else will surely answer you. That's what this forum is all about.
OK i'm ready to go with this linestage myself!. As this will be my 2nd scratch build, i'm still a little green. OK, alot green😀 Any of you experts willing to put this whole thing to paper? The definitive linestage seems to be brians. The schematic for the tube regulator I believe was based on what mark had for VR tubes. Would a different VR tube be a better choice? Perhaps one per channel? How about a definitve PS / transformer choice?
Brian's driver stage was easy to pick, but the rest still seems kind wishy washy if you know what I mean. Who can wrap this puppy up for us newbies (using VR tubes!)?
Brian's driver stage was easy to pick, but the rest still seems kind wishy washy if you know what I mean. Who can wrap this puppy up for us newbies (using VR tubes!)?
ok, thanks,
what's the difference between using LM317 and using cap with Resistor paralled ? in term of sound quality ?
LM317 give more what ?
cal
what's the difference between using LM317 and using cap with Resistor paralled ? in term of sound quality ?
LM317 give more what ?
cal
As I said in earlier posts, the only purpose of the LM317 in the cathode (bypassed with large caps) is to tightly control the current of the 12B4A, especially important as we are really running the tube pretty hard. The sound difference between the bypassed CCS and an “equivalent” power resistor in the cathode (also bypassed with the same caps) will be small. But the LM317 is cheap, maybe even cheaper than the power resistor it might replace. The quality of the bypass caps is really critical, because all the audio current must pass through them. Use good electrolytics (Black Gates, whatever) and bypass them with smaller "designer" film or PIO caps, seasoning to taste. Since the goal is to run the 12B4A at a very high plate current to achieve high transconductance, and, especially, low plate resistance, it is important that we don’t let the tube wander away too far from the desired current. We don’t want to risk over current in the tube or dropping the plate voltage too low or (too high) as the tube drifts or when different tubes are used. We also want to keep the two channels in a stereo pair in a tight parameter match. The cathode CCS locks in plate current and therefore the plate voltage, while only the cathode voltage can vary slightly as needed. There is a slightly greater low-end extension with the cathode CCS because the cathode caps see a higher impedance (longer RC time constant), and that will give a bit better low-end phase response, but I don’t make a big fuss about that point either. I think that the 12B4A, run “hot”, sounds balanced and pretty remarkable overall. The CCS sounds slightly better to my ears than a cathode resistor, but I can't really say why with any certainty. Enjoy.
calico, build it, you won't regret it. Mine is built with the suggested LM317 from Brian, bypassed with Black Gate. B+ is supplied by a MOSFET series regulator. Can't complain about the sound! (I don't want to use "you'd be blown away" he he he)
Brian,
Could you please post the details about doing this? I'm running a straight LM317 for my heaters.. so far sucessfully but I've heard about the advantages to raising them to near B+ level...
Thanks,
Mark
Mark,
I wouldn’t advise running the heater up to anywhere near B+ because that would greatly exceed the heater-cathode voltage rating. There is evidence that elevating the heater slightly more positive than the cathode will reduce noise by reverse biasing the residual emission from heater to cathode. Since the cathode in my design sits at around +15 volts, you can safely anchor the negative terminal of a floating DC heater supply at just over that – let’s say +20 volts (the positive side of the heater will be at +32.6 volts, or +26.3 if you center-tap the heater for 6.3 volts). So all you need is a resistor divider off the B+ to ground. Something like 300K (1W) to the B+, and 20K (1/4 W) to ground, the two tied in the middle. That will bring a +320 volt B+ supply down to +20V. Precision is not needed here. Then attach that center point to the negative terminal of the DC heater supply. Place a cap from this tap to ground to make sure no noise goes in either direction. Value is not critical, any small (say 10 uF) electrolytic rated at 35 VDC will do. And that’s it. I’m not near my lab notebook at the moment, so I’m conjuring all this in my head, but it’s not hard.
I wouldn’t advise running the heater up to anywhere near B+ because that would greatly exceed the heater-cathode voltage rating. There is evidence that elevating the heater slightly more positive than the cathode will reduce noise by reverse biasing the residual emission from heater to cathode. Since the cathode in my design sits at around +15 volts, you can safely anchor the negative terminal of a floating DC heater supply at just over that – let’s say +20 volts (the positive side of the heater will be at +32.6 volts, or +26.3 if you center-tap the heater for 6.3 volts). So all you need is a resistor divider off the B+ to ground. Something like 300K (1W) to the B+, and 20K (1/4 W) to ground, the two tied in the middle. That will bring a +320 volt B+ supply down to +20V. Precision is not needed here. Then attach that center point to the negative terminal of the DC heater supply. Place a cap from this tap to ground to make sure no noise goes in either direction. Value is not critical, any small (say 10 uF) electrolytic rated at 35 VDC will do. And that’s it. I’m not near my lab notebook at the moment, so I’m conjuring all this in my head, but it’s not hard.
