Hey guys, new here and to power amps so please bare with me...
I'm currently working on an all-tube low-ish voltage and power guitar amp.
I've already built the preamp and it works beautifully! Basically a 2204 with a few mods I thought of to get a bit more gain from the 12U7's I'm currently using. These may become 12AT7/12AX7's depending on how brave I get with the voltage... 25.5V currently, may get up to 40-50V.
So... My question is, I want to use an SE EL84 as the output, I know I'll be looking at mW here but in order to design it properly I need to be able to calculate what voltage I'm getting off of the tonestack... How would I do this?
I've considered using 12K5, 12AL8 and other low-voltage tubes but they're all in the 35-40mW range and I'd like to try for 1/4W to 1/2W... (May not be possible with a good sound but I'd like to try.) And at least they still make EL84's!
Any help would be greatly appreciated and I'm more than happy to share successful schematics if the build can be done.
Thanks in advance,
Mike.
I'm currently working on an all-tube low-ish voltage and power guitar amp.
I've already built the preamp and it works beautifully! Basically a 2204 with a few mods I thought of to get a bit more gain from the 12U7's I'm currently using. These may become 12AT7/12AX7's depending on how brave I get with the voltage... 25.5V currently, may get up to 40-50V.
So... My question is, I want to use an SE EL84 as the output, I know I'll be looking at mW here but in order to design it properly I need to be able to calculate what voltage I'm getting off of the tonestack... How would I do this?
I've considered using 12K5, 12AL8 and other low-voltage tubes but they're all in the 35-40mW range and I'd like to try for 1/4W to 1/2W... (May not be possible with a good sound but I'd like to try.) And at least they still make EL84's!
Any help would be greatly appreciated and I'm more than happy to share successful schematics if the build can be done.
Thanks in advance,
Mike.
That's what I meant but admittedly did not word it at all that way lol.
Maybe I should try it this way, how would I calculate the voltage from the cathode follower and then after the tone stack?
I'd need to see a schematic of what you're doing in order to comment. Usually, the front end and tone stack that's ready to drive an output tube might be in the vicinity of 1Vrms with the volume turned up all the way, and have a gain of roughly 5-20 thousand. Then the output stage adds a little more voltage gain (maybe 10 - 25) as well as providing a high current capability low Z output to drive the speaker. With your very low voltages, I'm not sure the usual tube graphs will show you what you need to know. When I looked into this a while back, I found it hard to get much good info on using very low voltages. the Butler fuzz pedal is one example of a 12AX7 being run on something like 20 volts. Early car radios ran tubes on relatively low voltage too. It will take some experimenting to get each tube stage to be its cleanest, and it may never get real clean at those voltages. Tube gains may be way below normal too. It might have a real nice distortion sound for guitar.
> in order to design it properly I need to be able to calculate what voltage I'm getting off of the tonestack...
YOU will design the voltages throughout.
You start with the ENDS. You want 20mV in to saturate the power stage. You want the power stage to deliver 100 Watts, 1/2 Watt, whatever the job demands.
You have also made a "preference" decision to use tubes.
> at least they still make EL84's!
How large is your lifetime need? 100,000 tubes, you better pick one that China/etc have in production. A single personal bedroom amp, you need one now and a couple spares, maybe a dozen in a lifetime. MANY tube types are readily available at good price from 1960s warehouse stock.
> low-ish voltage .... I'd like to try for 1/4W to 1/2W...
For pure resistances, power falls as Square of Voltage. Half voltage is One-Quarter power. Power tubes are "bent resistors" and output falls faster than Square. If
EL84 is aimed at the 250V-300V end. Say 5W out at 250V. _IF_ square law applied, then at 50V supply you expect 0.2W, below your goal. If Child's Law applied, 0.2 voltage makes 0.09 current, 0.018 power, 0.09 Watts.
If you are in 117V land there are many many fine tubes made to deliver 1W with ~~100V supply. Any of these will approach 0.18W at 50V, 2X the power of a low-V EL84.
How much drive is needed INto the power tube? The power tube designer has a trade-off. A high-Mu tube needs less drive voltage (saves pennies in small tubes) but a low-Mu tube makes more power out of an available supply voltage. Most "audio power pentodes" _are_ designed to make good power while using a *driver tube* fed about the same supply as the power tube Screen. If you see a 50L6 or 6V6 with 250V on G2, you can expect to feed it straight from a driver tube with a ~~250V supply.
Or: 50L6 with 110V supply for P and G2, G1 can be -7.5V bias, and similar peak voltage at full power. A triode voltage amplifier eating ~~90V supply can approach 18V peak output. Ample. At 50V we may expect like -3V or -4V bias/drive and 5V-8V driver output, works.
If you put tonestack between driver and power tube, you are likely to overload in the driver, which is unable to smack your full 1/4W out of the "power" tube with the tonestack in the way. If driver can't make over 9V peak, tonestack loss is as small as 1/4, then we only have 2V of drive for a tube which needs 3V-4V to make the full 0.2W goodness.
At the other end: you want 20mV off guitar to saturate the amp when full-up, and be able to turn-down to play 500mV clean. 12U7 will have gain near 12-15 per stage. At 50V we won't expect more than 9V peak output. 20mV-500mV input times gain of 15 is 300mV-7,500mV at the plate of the first stage. This is OK, but we can not put another gain stage here, because the next stage WILL clip most guitar and no way to avoid that. This is usually where the Volume control goes. Though you could have a second gain-knob further along; then they are traditionally Gain and Master, and the relative settings force distortion in between (or not).
The K.I.S.S. path is first stage, tonestack and volume, then gain to the power stage. 0.3V at first stage plate, tonestack gain of 0.2, is 0.06V with Vol full-up; 7.5V makes more but for Clean we turn-down. We need to gain-up 0.06V to the 4V needed to smack the power bottle. Gain of 66, which will not be one low-voltage gain stage, but can easily be two stages. Then you may re-consider putting the volume and tone networks different places instead of together, or adding another gain-knob for more to fiddle with.
YOU will design the voltages throughout.
You start with the ENDS. You want 20mV in to saturate the power stage. You want the power stage to deliver 100 Watts, 1/2 Watt, whatever the job demands.
You have also made a "preference" decision to use tubes.
> at least they still make EL84's!
How large is your lifetime need? 100,000 tubes, you better pick one that China/etc have in production. A single personal bedroom amp, you need one now and a couple spares, maybe a dozen in a lifetime. MANY tube types are readily available at good price from 1960s warehouse stock.
> low-ish voltage .... I'd like to try for 1/4W to 1/2W...
For pure resistances, power falls as Square of Voltage. Half voltage is One-Quarter power. Power tubes are "bent resistors" and output falls faster than Square. If
EL84 is aimed at the 250V-300V end. Say 5W out at 250V. _IF_ square law applied, then at 50V supply you expect 0.2W, below your goal. If Child's Law applied, 0.2 voltage makes 0.09 current, 0.018 power, 0.09 Watts.
If you are in 117V land there are many many fine tubes made to deliver 1W with ~~100V supply. Any of these will approach 0.18W at 50V, 2X the power of a low-V EL84.
How much drive is needed INto the power tube? The power tube designer has a trade-off. A high-Mu tube needs less drive voltage (saves pennies in small tubes) but a low-Mu tube makes more power out of an available supply voltage. Most "audio power pentodes" _are_ designed to make good power while using a *driver tube* fed about the same supply as the power tube Screen. If you see a 50L6 or 6V6 with 250V on G2, you can expect to feed it straight from a driver tube with a ~~250V supply.
Or: 50L6 with 110V supply for P and G2, G1 can be -7.5V bias, and similar peak voltage at full power. A triode voltage amplifier eating ~~90V supply can approach 18V peak output. Ample. At 50V we may expect like -3V or -4V bias/drive and 5V-8V driver output, works.
If you put tonestack between driver and power tube, you are likely to overload in the driver, which is unable to smack your full 1/4W out of the "power" tube with the tonestack in the way. If driver can't make over 9V peak, tonestack loss is as small as 1/4, then we only have 2V of drive for a tube which needs 3V-4V to make the full 0.2W goodness.
At the other end: you want 20mV off guitar to saturate the amp when full-up, and be able to turn-down to play 500mV clean. 12U7 will have gain near 12-15 per stage. At 50V we won't expect more than 9V peak output. 20mV-500mV input times gain of 15 is 300mV-7,500mV at the plate of the first stage. This is OK, but we can not put another gain stage here, because the next stage WILL clip most guitar and no way to avoid that. This is usually where the Volume control goes. Though you could have a second gain-knob further along; then they are traditionally Gain and Master, and the relative settings force distortion in between (or not).
The K.I.S.S. path is first stage, tonestack and volume, then gain to the power stage. 0.3V at first stage plate, tonestack gain of 0.2, is 0.06V with Vol full-up; 7.5V makes more but for Clean we turn-down. We need to gain-up 0.06V to the 4V needed to smack the power bottle. Gain of 66, which will not be one low-voltage gain stage, but can easily be two stages. Then you may re-consider putting the volume and tone networks different places instead of together, or adding another gain-knob for more to fiddle with.
Thanks for the reply, It's a 2203 preamp "clone" with minor changes. The schematic is here.
The changes I made were to remove the B+ line so 25.5V is going directly to the tubes via their respective plate resistors, The 10K cathode resistor on V1A has been dropped to 1.82K and bypassed with a 1uF and the cathode resistor on V1B has been dropped to 1.8K with the bypass cap also being changed to a 1uF. Sounds good to my ear.
As you can probably tell, I don't know nearly as much as I need to... This is why I'm playing with low voltage tubes as a learning tool.
I have indeed made a preference choice to use tubes... A simple TBA820M would do the job if I were willing to go down the SS route, Indeed many circuits would get me more output with less effort. I just like the idea of all tube. I may yet still be forced down the SS route but that's not a road I want to take.
I understand your point on life-time need but finding the tubes can be difficult here in the UK. EL84's are reasonably cheap and the output transformers readily available. Plus if I learn more about them and as I become more confident and knowledgeable with respects to designing with them, I would like to start venturing into "normal" B+ ranges.
The rest of what you said is extremely helpful too and I'll have to go over it a few times tomorrow after I've has some sleep and some coffee!
I could probably have just taken the easier route of cloning an amp or building a kit but I'm not going to learn any more than I already know that way so I really do appreciate the help here!
I understand your point on life-time need but finding the tubes can be difficult here in the UK. EL84's are reasonably cheap and the output transformers readily available. Plus if I learn more about them and as I become more confident and knowledgeable with respects to designing with them, I would like to start venturing into "normal" B+ ranges.
The rest of what you said is extremely helpful too and I'll have to go over it a few times tomorrow after I've has some sleep and some coffee!
I could probably have just taken the easier route of cloning an amp or building a kit but I'm not going to learn any more than I already know that way so I really do appreciate the help here!
I wouldn't waste time trying to get tubes to work well at safe low voltages, unless you were doing that for a certain distortion effect. High voltages are dangerous, but if you know that and are very careful, it shouldn't be a problem. When you turn off an amp, the power supply electrolytic caps will have a stored high voltage charge that should dissipate over a period of about a minute. You want to monitor that with a voltmeter until it's down below about 20 volts, before sticking you fingers into the circuit. There's usually a "bleeder resistor" across these caps to make sure they dissipate their charge when the amp is turned off, roughly 500Kohms and of the appropriate wattage. You might want to check out my Deuce Jr. guitar amp on my hobby website, where I talk in detail about how and why I designed each stage. It has EL84 output tubes.
Bob's Website
Bob's Website
In general use, you would be right, but this is guitar mplifier and that volue was chosen on purpose to reach a certain goal.
Self bias can never turn a tube OFF, for the very good reason that you need *some* current across the cathode resistor to create bias voltage.
No current=no bias.
The net effect is that tube is not only "cold" biased, but also clips very unsymmetrical: it can only go a few volts one way, almost 200V the other.
What´s desirable in Hi Fi is often not so in the Guitar World.
Self bias can never turn a tube OFF, for the very good reason that you need *some* current across the cathode resistor to create bias voltage.
No current=no bias.
The net effect is that tube is not only "cold" biased, but also clips very unsymmetrical: it can only go a few volts one way, almost 200V the other.
What´s desirable in Hi Fi is often not so in the Guitar World.
I can confirm that the cathode resistor on V1A is indeed 10K, it seems popular to change it out for something smaller... Maybe it has something to do with the sound Marshall wanted from the low input.
Thank you for the explanation... Fortunately I already know all of that.
I have built 2 high voltage preamps before, a 2203 clone and an ENGL Fireball clone, neither of which worked (I imagine because of the output buffer I had designed on the 2203 and the Fireball had an LND150 based effects send bought as a kit, but for some reason the transformer burnt out and I never did get around to investigating why.)I've had more luck with low-voltage circuits hence wanting to try a full low-voltage amp. I have no desire to make something more powerful than I need since I'm not in a band, I just love to play... And with my clear lack of knowledge about the voltages I would expect to see I can't even design a smaller power amp for a given preamp.
I would love to build a 2 channel high gain ~5 watter if I could but with my current lack of understanding of how to go about calculating what the signal in/out is it's not justifiable.
I'll check out your website, should make an interesting read!
Thanks!
> V1a shows a 10K resistor on its cathode. I believe that's a mistake.
You may be correct. Several much-loved guitar amp details almost had to have been found by mistake. The Fender 5F6a NFB connection for one. Marshall started by copying the 5F6a (with EL34 power jugs) but shifted to the "too small" 0.68uFd cathode bypass and the cold-bias 2nd stage. These persist off-and-on through many models and copies, with dart-toss value changes to builders' taste.
You may be correct. Several much-loved guitar amp details almost had to have been found by mistake. The Fender 5F6a NFB connection for one. Marshall started by copying the 5F6a (with EL34 power jugs) but shifted to the "too small" 0.68uFd cathode bypass and the cold-bias 2nd stage. These persist off-and-on through many models and copies, with dart-toss value changes to builders' taste.
The Marshall 2303 makes 50 Watts with 480V.
Thumb-math, that would be 0.5 Watts at 48V (hardly lethal).
While it seems absurd to use 6L6/EL34 for a half-Watt, they are readily available and will NOT wear-out at this level. (Mechanical and chemical failures will still happen.)
It won't be "the same" at 48V as at 480V. I'd expect to do a lot of changes to the gain structure. However you say your low-V build works for you, so maybe not. I'd leave all cathode bias resistors out where you can EASILY change them to adapt for the lower voltage. The "power" stage bias may be quite fussy. I sure would change to cathode resistor bias here-- a string of 100 Ohm 1W resistors so you can short-out to get 100 200 300 400 etc ohms to find an optimum.
At 48V gitar duty, the output transformer can be 115/230V:8V 10VA line transformer. MUCH cheaper than any SE or P-P Audio part.
If you object that 5 tubes for half-watt is a lot... I agree.
Thumb-math, that would be 0.5 Watts at 48V (hardly lethal).
While it seems absurd to use 6L6/EL34 for a half-Watt, they are readily available and will NOT wear-out at this level. (Mechanical and chemical failures will still happen.)
It won't be "the same" at 48V as at 480V. I'd expect to do a lot of changes to the gain structure. However you say your low-V build works for you, so maybe not. I'd leave all cathode bias resistors out where you can EASILY change them to adapt for the lower voltage. The "power" stage bias may be quite fussy. I sure would change to cathode resistor bias here-- a string of 100 Ohm 1W resistors so you can short-out to get 100 200 300 400 etc ohms to find an optimum.
At 48V gitar duty, the output transformer can be 115/230V:8V 10VA line transformer. MUCH cheaper than any SE or P-P Audio part.
If you object that 5 tubes for half-watt is a lot... I agree.
Don't worry, I'm not fooling myself into thinking that at 48V it's going to sound like a mini 2203. I know in all likelihood there will be many things I'll be unhappy with.
I wouldn't call 5 tubes for a half watt amp a lot, at least in my situation... I guess that's subjective though lol. If your thumb math is indeed correct, I'm very intrigued and excited! I am expecting the bias to be fussy, could I not, with the cathode resistor bias use a 1W potentiometer? Something like this: P11S1V0FLSY00501KA Vishay / Sfernice | Mouser United Kingdom
I really appreciate the help here, I can't thank you guys enough! Even if it is absurd.
I wouldn't call 5 tubes for a half watt amp a lot, at least in my situation... I guess that's subjective though lol. If your thumb math is indeed correct, I'm very intrigued and excited! I am expecting the bias to be fussy, could I not, with the cathode resistor bias use a 1W potentiometer? Something like this: P11S1V0FLSY00501KA Vishay / Sfernice | Mouser United Kingdom
I really appreciate the help here, I can't thank you guys enough! Even if it is absurd.
That could be true. Thanks for pointing that out. It could be a way of increasing even harmonic distortion, since it would cause asymmetry when over driven.
The 10K is correct, as is your analysis. It's part of the intended sound.
Rob Robinette has written a very good explanation of how the JCM800 works.
Modern hi-gain amps go much further - the Soldano Super Lead Overdrive runs a 39K cathode resistor in a similar cold clipper stage
Last edited:
It's hard to tell someone what the various levels in an amp are, because there's a bunch of variables. On most of my guitar amp schematics, I've measured and written in the gains for each stage. That should give you a rough idea of what works. If you want to build a great 5 watt guitar amp, the one called Fat Mama on my website is one of the best I've ever heard. It uses a 6SJ7 pentode 1st stage, which nobody makes any more, but places like Antique Electronics in AZ have many NOS of that tube for pretty cheap ($10ish). It is however fairly microphonic, more than a 12AX7, so it might be wise to shock mount its tube socket, if it's in the same cabinet as the speaker driver. Or just replace it with a 12AX7 stage, which will reduce the gain a bit, and may have a slightly different sound, but should in general work fine.I would love to build a 2 channel high gain ~5 watter if I could but with my current lack of understanding of how to go about calculating what the signal in/out is it's not justifiable.
I'll check out your website, should make an interesting read!
Thanks!
For home use, I run my guitar thru a compressor, a stereo chorusser (sometimes), a stereo digital delay (sometimes) and a Strymon Big Sky stereo digital reverb (always) and the stereo outputs go to two of my amps that are about 6 feet apart. It's majestic. It makes me a better guitarist because I love how it sounds. And 5 watts is plenty loud for home use.
This is all true. I'm still really into making a low voltage one for the time being though... At this point it's more of a curiosity to see if I can make it work. Plus it has the added benefit of forcing me to learn more about amps which can't be wrong!
I'll check out the Fat mama too! It's always interesting to see how people approach things in different ways.
I'll check out the Fat mama too! It's always interesting to see how people approach things in different ways.
If you want low voltage sub 1W amp, why not use tubes made just for that ?
Have a look here Battery-Powered Guitar Amp
Mona
Have a look here Battery-Powered Guitar Amp
Mona
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.