Hi all! First post here, been lurking for a while. Hoping ya'll might have some advice for a newb.
I have a Valve Jr.
http://www.muzique.com/news/images/Epi1.gif
that I'm converting to turret board build. Looking for best clean sound, with lots of headroom. Have a great Pedal for distortion, so not looking for lots of gain. Anyway, I've always felt that the B+ PS is the weakest point on most guitar amps. Valve Jr. uses simple CRC network.
Glassware webpage has a janus shunt regulator kit
http://www.tubecad.com/2008/09/13/janus regulator.png
that might be a candidate for this amp. If I use solid state rectificiers, current PS transformer (filament supply) could likely power the extra 12dw7 tube.
My question is this, can I remove any of the CRC B+ network to the other tubes? If I'm getting this right, the problem of having a lot of capacitance after a shunt regulator is that when the amp is powered up, cap network acts like a short for a short time, and could actually stress the shunt circuit, so this might not be a direct drop in. The amp will have a standby switch wired in, so filaments should be well heated before B+ is powered up. If I'm understanding this right, CRC network is meant for ripple filtration and as sort of a power/voltage/current reserve for the different gain stages. Shunt regulator should take care of ripple, but not sure if there is enough 'reserve' to not starve the down-line gain stages. Should I decrease capacitance downline, relying on large filter capacitors in shunt circuit to take up some of the burden? Or am I picking nits?
I'm reading a lot, trying to understand tube circuits, and this seems a simple build for first timer, but I'm clueless as to whether this is feasible. Can anyone help?
I have a Valve Jr.
http://www.muzique.com/news/images/Epi1.gif
that I'm converting to turret board build. Looking for best clean sound, with lots of headroom. Have a great Pedal for distortion, so not looking for lots of gain. Anyway, I've always felt that the B+ PS is the weakest point on most guitar amps. Valve Jr. uses simple CRC network.
Glassware webpage has a janus shunt regulator kit
http://www.tubecad.com/2008/09/13/janus regulator.png
that might be a candidate for this amp. If I use solid state rectificiers, current PS transformer (filament supply) could likely power the extra 12dw7 tube.
My question is this, can I remove any of the CRC B+ network to the other tubes? If I'm getting this right, the problem of having a lot of capacitance after a shunt regulator is that when the amp is powered up, cap network acts like a short for a short time, and could actually stress the shunt circuit, so this might not be a direct drop in. The amp will have a standby switch wired in, so filaments should be well heated before B+ is powered up. If I'm understanding this right, CRC network is meant for ripple filtration and as sort of a power/voltage/current reserve for the different gain stages. Shunt regulator should take care of ripple, but not sure if there is enough 'reserve' to not starve the down-line gain stages. Should I decrease capacitance downline, relying on large filter capacitors in shunt circuit to take up some of the burden? Or am I picking nits?
I'm reading a lot, trying to understand tube circuits, and this seems a simple build for first timer, but I'm clueless as to whether this is feasible. Can anyone help?
Last edited:
Please note that guitar amplifier threads really belong in the Instruments & Amps forum. Please see the subheaders at the top of each forum for the exact reason why.
Ok, you want a clean sound, and "lots of headroom".
So you probably do want quite stable supply rails.
But maybe you already have them? The two triodes and the pentode are all operating in single ended class A, so the change in current consumption with output power is minimal compared to a push-pull amp.
If you really do want a regulator, why a shunt one? I can't say that the janus looks that good for this application either, though I'm open to persuasion, if anyone's actually bothered.
I'd suggest a solid state series regulator, have a look here:
Gearslutz high voltage regulated tube supplies
- you don't need a bias supply, so you might find gyraf's circuit appropriate
or if you really want to use valves, read Steve Bench's pages starting here:
Steve Bench on Tube Regulators
So you probably do want quite stable supply rails.
But maybe you already have them? The two triodes and the pentode are all operating in single ended class A, so the change in current consumption with output power is minimal compared to a push-pull amp.
If you really do want a regulator, why a shunt one? I can't say that the janus looks that good for this application either, though I'm open to persuasion, if anyone's actually bothered.
I'd suggest a solid state series regulator, have a look here:
Gearslutz high voltage regulated tube supplies
- you don't need a bias supply, so you might find gyraf's circuit appropriate
or if you really want to use valves, read Steve Bench's pages starting here:
Steve Bench on Tube Regulators
Last edited:
As soon as you hit clipping you are no longer in class A. I have seen some wilder current swings in big SE amps than in P-P amps of the same power level when driven to extreme with a multi effects box.
The usual CRC filter with smallish filter caps (22uf) and a 3.3 K resistor contributes a lot of the "tone" found in these little SE amps. When driven to extreme (1 or more overdrive pedals) the output tubes screen current goes up driving down the preamp voltage and gain. As the note dies out the output tube begins to recover drawing less current allowing the preamp voltage and gain to increase. This improves the amps "sustain" and is useful for some music styles.
As with any amp, what you want to build is determined by what you want it to sound like. As a generality, a regulated power supply is good for a loud clean tone, a unregulated power supply is best for a loose distorted tone with more touch sensitivity.
A shunt regulator wastes energy by shunting some of the B+ to ground. They were not really intended for this type of application. I have never tried it in a guitar amp, but who knows, you might invent a new sound. I tend toward a mosfet series regulator for a guitar amp though.
Last edited:
But hang on...
I had to read it twice. He doesn't want it to clip.
Let's just read those beautiful phrases again:
Big SE amps ... driven to extreme
When driven to extreme ... the output tubes screen current goes up driving down the preamp voltage and gain.
.. 
Hi all! Thanks for all the reply's!!!
First off, sorry for posting in the wrong section. Thanks to the admin for bumping it!
Next, what I'm looking for is cleaning up the grunge inherent in a simple CRC network, more background noise/hiss/hum elimination, so that the cleans get cleaner. I realize a guitar amp is probably the worst environment for doing this, with all the inline equipment and long cables. It's an experiment.
In terms of the type of regulation, I could care less whether it's SS or tube, just thought tubes were generally a little more forgiving. I do have limits on what the power transformer is capable of powering, though. Worst case scenario, could get a separate transformer for the filament supply, easing demands on the transformer and maybe allowing for a little more current/voltage on the B+ secondary.
Again, the distortion pedal in front would be used for clipping at the preamp level, so I'm looking for best cleans here. Power amp clipping I do want, but relatively 'soft', if you catch my meaning. Not looking for metal. Will be using a VVR, to turn down actual volume (SPL) while still keeping output tube dist characteristics at reasonable levels. I know it's not exactly the same, but am willing to live with the tradeoffs for my many roomates sakes.
I realize routing of wires, shielded wires to pots and jacks, and grounding are all huge deals for eliminating hum, and hope to be thorough there as well. I'm actually making a separate DC filament board and mounting it far away from the B+ for the same purpose (eliminating noise). Could use some suggestions for that as well.
I'm just trying to take down all noise floor to a min to help the cleans. Maybe it's not recommended, but this is a cheap amp and worst case scenario, I can redo it. Looking at shielding sections from each other (TI shield), shielding and routing of wires, etc. anything to bring down the noise floor.
Thanks again for the reply's, hope to learn more!!!
First off, sorry for posting in the wrong section. Thanks to the admin for bumping it!
Next, what I'm looking for is cleaning up the grunge inherent in a simple CRC network, more background noise/hiss/hum elimination, so that the cleans get cleaner. I realize a guitar amp is probably the worst environment for doing this, with all the inline equipment and long cables. It's an experiment.
In terms of the type of regulation, I could care less whether it's SS or tube, just thought tubes were generally a little more forgiving. I do have limits on what the power transformer is capable of powering, though. Worst case scenario, could get a separate transformer for the filament supply, easing demands on the transformer and maybe allowing for a little more current/voltage on the B+ secondary.
Again, the distortion pedal in front would be used for clipping at the preamp level, so I'm looking for best cleans here. Power amp clipping I do want, but relatively 'soft', if you catch my meaning. Not looking for metal. Will be using a VVR, to turn down actual volume (SPL) while still keeping output tube dist characteristics at reasonable levels. I know it's not exactly the same, but am willing to live with the tradeoffs for my many roomates sakes.
I realize routing of wires, shielded wires to pots and jacks, and grounding are all huge deals for eliminating hum, and hope to be thorough there as well. I'm actually making a separate DC filament board and mounting it far away from the B+ for the same purpose (eliminating noise). Could use some suggestions for that as well.
I'm just trying to take down all noise floor to a min to help the cleans. Maybe it's not recommended, but this is a cheap amp and worst case scenario, I can redo it. Looking at shielding sections from each other (TI shield), shielding and routing of wires, etc. anything to bring down the noise floor.
Thanks again for the reply's, hope to learn more!!!
Hi all!
After reading thru the links ya'll posted, I'm hoping to ask a few questions.
1st - OK, maybe a shunt reg was a bad idea. The only thing that made me think the janus might be a candidate was this statement from his order page -
"The Janus regulator is not a DC voltage regulator, as it cannot maintain a fixed B+ voltage; instead, it works to eliminate any AC perturbations on its output voltage. In this way it is similar to a large inductor-filled power supply that bucks quick changes in output voltage, but allows slow DC voltage to vary with the wall voltage variations."
Sounds like sag characteristics? My hope is for a drop in board, but I'll entertain any ideas.
Simon B - Looked at the Gyraf circuit http://www.gyraf.dk/gy_pd/g9/g9_sch.gif , I wouldn't even know where to start in trying to extract the B+ portion. The circuit by Waveborn looks interesting, but could it be done without the Bias circuit? And would the inrush current adversely affect it?
Thanks for the link to Steve Bench article. The single tube, fixed biased pentode looks interesting. Any idea of a current production tube that could be used for it? Also, would there be a solid state equivalent?
For KevinKR, you seem to know your PS's, is there a circuit you have used/developed for this type application?
Lastly, would there be any advantage to inductor filled PS over standard CRC network? Sort of a CLCLC...etc.?
Thanks again for any reply's!!!!
After reading thru the links ya'll posted, I'm hoping to ask a few questions.
1st - OK, maybe a shunt reg was a bad idea. The only thing that made me think the janus might be a candidate was this statement from his order page -
"The Janus regulator is not a DC voltage regulator, as it cannot maintain a fixed B+ voltage; instead, it works to eliminate any AC perturbations on its output voltage. In this way it is similar to a large inductor-filled power supply that bucks quick changes in output voltage, but allows slow DC voltage to vary with the wall voltage variations."
Sounds like sag characteristics? My hope is for a drop in board, but I'll entertain any ideas.
Simon B - Looked at the Gyraf circuit http://www.gyraf.dk/gy_pd/g9/g9_sch.gif , I wouldn't even know where to start in trying to extract the B+ portion. The circuit by Waveborn looks interesting, but could it be done without the Bias circuit? And would the inrush current adversely affect it?
Thanks for the link to Steve Bench article. The single tube, fixed biased pentode looks interesting. Any idea of a current production tube that could be used for it? Also, would there be a solid state equivalent?
For KevinKR, you seem to know your PS's, is there a circuit you have used/developed for this type application?
Lastly, would there be any advantage to inductor filled PS over standard CRC network? Sort of a CLCLC...etc.?
Thanks again for any reply's!!!!
Then the Champ is not for you. Some people just add bigger caps, a regulated heater supply and change the 6V6 for the 6l6. More clean headroom and less noise. But I wouldn't do that. Not with what you have. Waste of time and money, just get yourself a cheap push pull amp. One that makes sense modding it if you want to - it makes no sense to mod a cheap bastard like the Epi junior.
Fender Excelsior | Sweetwater.com
There is a bit of a conflict between wanting soft clipping and max headroom......
At least a cheap amp is less to worry about experimenting with than an expensive one if you're just starting out on this stuff.
If you put an inductor in you'll have to do some maths, otherwise you'll find you've created something that rings.
If you really want to try a regulator, use a series one, with a mosfet as the pass element, as tubelab suggests, something like the one shown below. The ordinary diodes are simply to protect the mosfet in case you manage to discharge the upstream cap before the downstream one. The zener diode sets the output voltage. Sorry it's not a drop in board.....
At least a cheap amp is less to worry about experimenting with than an expensive one if you're just starting out on this stuff.
If you put an inductor in you'll have to do some maths, otherwise you'll find you've created something that rings.
If you really want to try a regulator, use a series one, with a mosfet as the pass element, as tubelab suggests, something like the one shown below. The ordinary diodes are simply to protect the mosfet in case you manage to discharge the upstream cap before the downstream one. The zener diode sets the output voltage. Sorry it's not a drop in board.....
Attachments
Hi all!
Tube Amp Mods - - QS-LV - Quiet Supply for Low Voltage Tube Amps
It just goes to show me, if I look long enough, I will find it!!! Neat site, lots of stuff I would like to try out!!! Thanks again!
Tube Amp Mods - - QS-LV - Quiet Supply for Low Voltage Tube Amps
It just goes to show me, if I look long enough, I will find it!!! Neat site, lots of stuff I would like to try out!!! Thanks again!
But hang on...
I had to read it twice. He doesn't want it to clip.
A "clean" guitar sound still clips on pick attacks, especially on cord strums. If you turn the gain down to prevent clipping on pick attacks, you will have very little volume. You would be surprised how little volume you get from a Fender Twin Reverb if you turn it down far enough to prevent any clipping. A guitar with a hot pickup can clip the input stage before it hits the volume contol where you can turn it down. You need an oscilloscope to reveal the clipping, you rarely hear it when the amp starts to clip.
Hi All!
Thanks for replys!
Simon - thanks for your research and experience. Probably going to do a drop in board for now, and may revisit this later. Yes, after rethinking it, I probably do want clipping.
Cassiel - As I understand, Valve Jr. is already biased very hot on the EL-84, many tweakers are turning down the voltages there for tube life. Worst case scenario, it won't work and I'll have to incorporate something else. Valve Jr. is a big compromise, I understand, but right now it's what I have, and I can't afford much else. So it's tweak time with the limited resources I have.
Loudthud - Yes, I stand corrected on clipping. Ahhh, I haven't got an osciliscope nor the budget for one. Still trying to justify a variac purchase I want to make. Anyway, thanks for the tips from all of you.
Have a blessed day!!!
Thanks for replys!
Simon - thanks for your research and experience. Probably going to do a drop in board for now, and may revisit this later. Yes, after rethinking it, I probably do want clipping.
Cassiel - As I understand, Valve Jr. is already biased very hot on the EL-84, many tweakers are turning down the voltages there for tube life. Worst case scenario, it won't work and I'll have to incorporate something else. Valve Jr. is a big compromise, I understand, but right now it's what I have, and I can't afford much else. So it's tweak time with the limited resources I have.
Loudthud - Yes, I stand corrected on clipping. Ahhh, I haven't got an osciliscope nor the budget for one. Still trying to justify a variac purchase I want to make. Anyway, thanks for the tips from all of you.
Have a blessed day!!!
Good point Loudthud, dry guitar signal does show big picking transients, probably less noticeable even on a Fender Twin Reverb at low volume than through full range speakers.
Whose tweeters will of course blow much sooner than the unwary might like them to......
That circuit is pretty much what I use. You can wire a pot across the zener, connect the wiper to the mosfet gate and you have a VVR. VVR's are the hot ticket these days, but they are intended for getting dirty at low volumes by reducing the B+.
You probably already have some clipping from the initial pick attack transient. This is why guitar amps have smallish coupling caps compared to HiFi amps. The pick attack moves the guitar string away from the pickup pole and then lets it fly across the pole's face before it starts to vibrate. This generates a large low frequency transient. If the transient reaches the OPT it will saturate causing a large increase in tube current and clipping even before there is any sound.
I started making tube guitar amps in the early 60's because tubes were free out of dead TV's at the local trash dump. They worked pretty good for a teenage kid.
I made my first solid state guitar amp in high school. I used circuitry from HiFi amps since there was no schematics available then. I got a 15 inch field coil speaker from a Hammond organ. The amp made a loud pop every time you plucked a string and destroyed the speaker in a week. Frequency response down to 10 Hz is not a good thing in a guitar amp.
Hi Guys
Like all economy SE amps, the VJr is limited in what can be done in the context of the circuit. The PT is the ultimate limiter of power output and effective headroom. The OT is skimpy and restricts bandwidth when pushed hard. A larger OT - one with hifi bandwidth instead of just mids - will provide cleaner output without tone change all the way up to clipping. A replacement designed for Champ, VJr, etc will not provide any more bandwidth or cleaner tone than the original.
You have to be careful throwing in different output tubes. In a Champ, just because a 6L6 fits the 6V6 "hole" does not mean the PT will be happy supplying the extra 450mA of heater current required - that's a 63% overload on the heater winding!
In a VJr, the heater load is just over 1A, so installing an octal socket for alternate tubes present possibilities of quite different tones. Pulling the EL-84 and adding 6V6 gives a creamier/muddier tone with reduced heater draw. Plugging in 6L6 instead will provide cleaner tone but no power increase with just a 13% overload on the heater winding.
Note that tube life is easily extended by adding a proper value screen resistor - 1k-1W flameproof is the absolute minimum for safety. EL-84s have sensitive screens, so a 2k2-5W is better. You don't have to change Rk if you add a screen resistor. TUT has advocated this since 1994.
VVR is a poor copy of an early version of Power Scaling. Real Power Scaling is applicable to any tube amp of any power level and has a control range of "unusably loud down to unusably quiet". his allows the player to select his own control limits.
One point not addressed is speakers. The biggest improvement you can make will be to bypass the tone crippling combo speaker and plug the electronics into a full size detuned cabinet or two. See SPKR for construction details.
Have fun
Kevin O'Connor
London power.com
Like all economy SE amps, the VJr is limited in what can be done in the context of the circuit. The PT is the ultimate limiter of power output and effective headroom. The OT is skimpy and restricts bandwidth when pushed hard. A larger OT - one with hifi bandwidth instead of just mids - will provide cleaner output without tone change all the way up to clipping. A replacement designed for Champ, VJr, etc will not provide any more bandwidth or cleaner tone than the original.
You have to be careful throwing in different output tubes. In a Champ, just because a 6L6 fits the 6V6 "hole" does not mean the PT will be happy supplying the extra 450mA of heater current required - that's a 63% overload on the heater winding!
In a VJr, the heater load is just over 1A, so installing an octal socket for alternate tubes present possibilities of quite different tones. Pulling the EL-84 and adding 6V6 gives a creamier/muddier tone with reduced heater draw. Plugging in 6L6 instead will provide cleaner tone but no power increase with just a 13% overload on the heater winding.
Note that tube life is easily extended by adding a proper value screen resistor - 1k-1W flameproof is the absolute minimum for safety. EL-84s have sensitive screens, so a 2k2-5W is better. You don't have to change Rk if you add a screen resistor. TUT has advocated this since 1994.
VVR is a poor copy of an early version of Power Scaling. Real Power Scaling is applicable to any tube amp of any power level and has a control range of "unusably loud down to unusably quiet". his allows the player to select his own control limits.
One point not addressed is speakers. The biggest improvement you can make will be to bypass the tone crippling combo speaker and plug the electronics into a full size detuned cabinet or two. See SPKR for construction details.
Have fun
Kevin O'Connor
London power.com
Hi All!
Yes, I do understand the limited nature of the Valve Jr., in terms of PT and OT. I have been considering an octal instead of 9 pin for output tube. I do have a new OT for it.
Hey Kevin! Glad you had that B+ board available! Looking at getting a few books from you to understand the circuit. Also looking to get your Power scaling version. Which do you recommend?
Yes, I do understand the limited nature of the Valve Jr., in terms of PT and OT. I have been considering an octal instead of 9 pin for output tube. I do have a new OT for it.
Hey Kevin! Glad you had that B+ board available! Looking at getting a few books from you to understand the circuit. Also looking to get your Power scaling version. Which do you recommend?
Hi Guys
In a Champ, VJr et al, you can go two ways to Power Scale:
Use SV2 for a Full-PS solution, taking care with heat management, or wire it for PS-TT with less concern about heat.
Use SV84 for a single-node Full-PS solution where you have either added another RC or two to the Va filter, or added QS-LV for a much quieter amp. SV84 can also be wired for PS-TT operation with less attention to heat management.
If you are installing QS-LV anyway, then the QS-LV + SV84 solution is much more compact with only two mosfets to mount, compared to QS-LV + SV2 with a total of three mosfets and a larger foot print.
The kit notes for SV2 and SV84 show both wirings. See the articles on our site to learn a bit about Full-PS and PS-TT. TUT7 will be out early 2013 and has a chapter about PS-TT - it is the corrected power control approach everyone else tries and gets wrong.
As far as voltage drops across QS-LV: this will be essentially just the gate-to-source threshold voltage plus a small drop across the current-sense resistor.
Note that the regulator circuit shown in a post above will have a strong tendency to kill mosfets at turn-on. People afraid of solid-state circuits seem to think it is better to remove vital current clamps to make the circuit simpler and have less silicon. For the sake of one R and one Q you can remove all worry about replacing the expensive mosfet.
Have fun
Kevin O'Connor
londonpower.com
In a Champ, VJr et al, you can go two ways to Power Scale:
Use SV2 for a Full-PS solution, taking care with heat management, or wire it for PS-TT with less concern about heat.
Use SV84 for a single-node Full-PS solution where you have either added another RC or two to the Va filter, or added QS-LV for a much quieter amp. SV84 can also be wired for PS-TT operation with less attention to heat management.
If you are installing QS-LV anyway, then the QS-LV + SV84 solution is much more compact with only two mosfets to mount, compared to QS-LV + SV2 with a total of three mosfets and a larger foot print.
The kit notes for SV2 and SV84 show both wirings. See the articles on our site to learn a bit about Full-PS and PS-TT. TUT7 will be out early 2013 and has a chapter about PS-TT - it is the corrected power control approach everyone else tries and gets wrong.
As far as voltage drops across QS-LV: this will be essentially just the gate-to-source threshold voltage plus a small drop across the current-sense resistor.
Note that the regulator circuit shown in a post above will have a strong tendency to kill mosfets at turn-on. People afraid of solid-state circuits seem to think it is better to remove vital current clamps to make the circuit simpler and have less silicon. For the sake of one R and one Q you can remove all worry about replacing the expensive mosfet.
Have fun
Kevin O'Connor
londonpower.com
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.