Let’s Build a Low Voltage Vacuum Tube Audio Amplifier
Vložit
- čas přidán 29. 11. 2020
- In this episode, we step out of our comfort zone and see if we can build a functioning vacuum tube audio amplifier running at low and safe voltages. Spoiler alert, we do!
Also, we’re now on Discord, Reddit and Twitter, so join us!
Discord: / discord
Reddit: / usagielectric
Twitter: / usagielectric
Thanks for watching! - Věda a technologie
The capacitors between each amplification stage are called interstage coupling caps.
The easiest way to conceptualize these caps, for novices, is that they block DC but allow AC to pass through.
The DC voltage is your tube power, and the AC is your audio waveform being amplified.
Basically these block the high voltage DC from passing to the next stage, but allow the AC audio waveform to pass through unopposed for further amplification.
nah
the easiest way to conceptualize these caps for novices is that they get dc biased
u have to mention that the signal after the input is actually an ac signal with dc offset
and the capacitor when exposed to this signal slowly gets charged to a voltage because this signal is centered around a dc voltage and not 0
they get charged to the dc voltage, therefore only the ac passes since the ac is swinging around that dc bias, and the capacitor cant react/charge instantly, therefore passing the ac
this is basically a high pass filter
basically imagine the capacitor gets charged to an average value of the signal it gets (both sides), say if the capacitor has 5v charged because ur signal is centered around 5v, the resistor after capacitor has a voltage of 0 or the divider voltage, imagine the resistor is connected to ground therefore 0v when capacitor is charged, now the signal is 6v, the capacitor cant charge instantly, but since the capacitor already has 5v across it, a 6v on the capacitor positive terminal means 1v on the other terminal, the signal is forcing 6v on the positive terminal because it is a voltage source, since the capacitor has 5v, the other terminal must be 1v, and this 1v is also on the resistor since the resistor is connected after the capacitor, this is how the changing signal gets passed, now novices will ask, whos to say the capacitor wont charge to 6v and wont pass the signal, that topic is covered within filters and frequency response, in the same way, if the signal suddenly becomes lower than 5v, the resistor actually has a negative voltage, which many novices dont understand how u can get a negative signal from dc voltage, this is because capacitor discharges the other way through ground, and this is also why the voltage signal needs to be push pull, say if u have an emitter follower without a resistor to ground, u could measure a changing voltage, but a connected capacitor wouldnt discharge because when the signal becomes lower, there is no pathway to ground to discharge
if u understand these, u can further understand dc bias in capacitors, diodes, varactors, extending into rf circuits
saying they block dc but allow ac is like memorization, u cant conceptualize and u cant understand
when i was a novice, i sure didnt get or conceptualize wtf u mean block dc but allow ac, i mean i heard wat u tryna say(but i would just be a yes man, it be like "it is wat it is" for me), and theory also says so, but there was no conceptualization
but what i mentioned is the actual conceptualization
now if u learnt about superposition in ac circuit, then u can also calculate it and model it, u can both see how it passes the ac and see the frequency response/transfer function in a simple rc circuit
when u a novice, u also dont understand why everyone says only ac make sound when clearly dc also makes sound
because their definition is ill defined, though dc is defined as current in one direction, ac as current in both direction
as long as a signal is changing, no matter ac or dc, it will make sound, might be better to call it an ac signal with dc offset, in circuit model, u would put dc source series with ac source
and another thing novices dont get is how a transistor is able to work with ac when it is a dc device, the reason is also the capacitor, u cant just apply a negative signal, but if u apply that negative signal to capacitor first which has a dc bias, u can offset that signal into the positive as long as its amplitude isnt bigger than the dc bias, thus the transistor is actually still working with dc (formal definition)
@@urnoob5528 You dive into the fine details, while correct, novices will just stop listening because it is much too much information at once. A novice starts with basics, not fine details. If novices dont get it, its because the explanation is too detailed.
Thank you👍
I really appreciate the work youve done with tubes. Not many (at all) work with tubes at low voltages. Been waiting on an audio setup!
Thank you so much! I think high voltages usually associated with tubes scare a lot of people away from them, but they actually work surprisingly well even at low voltages. We actually expand this audio setup a bit more in the next episode. Also, one of these days, I would like to try to build a proper OTL amplifier, although that one will probably require a genuine HV power supply.
@@UsagiElectric i agree the plates should only need about 36 volts to charge them
The great thing about your content is that I feel like i'm not only finally learning how vacuum tubes work, but also that urge i usually get, where I'd want to get some of the stuff and try it out for myself, is fully satiated... :D Still love it all! Thank you!
Looks like a great endeavor, the LM4250 Op-Amp was my favorite.
I did it 40 years ago using 12ax7 or au7. With 12v on the plate let the bias float. It was remarkably clean.
Thankyou for demonstrating every step of the way. I'd like to build something like this but didn't know how. You have a knack for this & it's inspired me
That was cool - seeing how simple it can be! Nice one.
Thanks for the great video. I joined the Royal Australian Navy in early 1979 and entered into an electrical engineering course. My class was the last to learn about valves and I worked on the 4.5" turret which was driven, as well as an A and B motor, by a big bank of valves. One of my biggest regrets is deliberately blowing up valves in the RCA testing case!
Thank you so much!
That's awesome that you got to learn and work on some proper tube equipment. I can't imagine the amount of power needed to drive a 4.5" turret, the bank of tubes must have been truly a sight (and a great heater in the winter)!
@@UsagiElectric In general, the motors in heavy duty drive trains such as gun turrets use DC torque motors, which have their current controlled by DC motor generators (MG's). The output of the motor generator provides the current to one of the torque motor windings. The input to the motor generator controls its output current. At this point the MG input could be produced by vacuum tubes or solid state circuits in more recent equipment. Using the output power of a large group of tubes to provide the current through the torque motor winding is possible, but not very practical.
You inspired me to go back to tubes experimentations : since your two videos on heptodes, I built a low-voltage guitar tremolo with a beautiful, small 6AJ8 pentagrid converter, and was amazed by how good the sound is, given the circuit simplicity. Thanks for your enthusiasm, keep up the good work ! By the way, if you plan to make a bigger amplifier (push-pull?) and start to get distorsion, you can disassemble the E and I laminations of the output transformer, and put them back with all E together and all I together, separated by a piece of paper, all maintened together with lacquer or duct tape : the air gap between E and I pieces will prevent core saturation due to DC voltage. This is what is done on most output transformers as far as I know
Thank you so much! That's awesome to hear that you've gotten back into tube experiments! I absolutely love those tubes that have that mesh on the inside, like the 6AJ8. I have a few 6EJ7 pentodes that look like that and they're just such cool looking tubes.
I would someday like to build a proper high voltage amplifier that can put out some serious sound, but I've actually gotten really interested in OTL amplifiers. I have something like 150 6CB6 pentodes that are decently linear when triode strapped, so something like a 10 or 20 tubes output stage using those little guys might be pretty interesting and give some wild sound!
@@UsagiElectric I love my 6BX6 for you can see the cathode glow, but man, those 6CB6 are on a whole new level ! Can't wait to see an OTL with these
The great thing about only just seeing this video now, I can see your next video, and the 48v amp video right away. 😁👍
A good amplifier is the one that makes you happy, that plays the songs you like the way you like.
Cool project and great Datsun shirt! My dsd has a restored 510 rally car - -he used to race back in the late 60's/early 70's.
Very interested in the low voltage side of tube behavior -thanks so much for sharing
Thanks for checking the videos out!
I have some more low voltage audio ideas I definitely want to try out in the future too, so more to come!
Very interesting... I did the same when diving into schemas w/ tubes, even though when starting with electronics, both ground and power 'rails' were at the bottom of the schema. For mental clarity it is just much easier to have them separated: ground on the bottom and power on the top. With that, already the orientation of a component implies what's going on. Same is then true for horizontally oriented components: they show what is going on between, for example, stages. The 'look' of a schema is so important to match the topography of the circuits operation.
I'm a fan of starved plate design . I seen many people get thousands of dollars amps and then drive everything with a tiny transistor or opamp in a pedal . Wonderful work. You sir are a genius 😎
Cool Video! One thing that would be cool is for every component/jumper you laid down, you could flash back to your drawing and point to the reference.
I remember building a small 'triode-pentode' amp from the junk box, way back about 1972. It worked amazingly well. Although designed as an output for a radio, it also served very well as "gram" amplifier. As far as I can remember, it was intended to run at around 200V HT, but to my astonishment it worked all the way down to 12V HT. The only real discernable difference was in the reduction in volume as the HT was reduced. Once below about 50V HT, some degree of distortion began to creep in as the volume pot was advanced.
Don't be fooled into thinking that at low HT you can only drive a small speaker - Not so! Taking the amp in this video, I'd happily connect a 6"x4" speaker, and you would notice a distinct increase increase in output volume.
My amp was based around a salvaged TV audio valve (tube, if you prefer) and was a humble Mullard PCL82 (it could equally have been an ECL82) whatever they might equate to in other nomenclature. I added a basic tone control, and in its native form it had no NFB. Had it done so, it may have faired better at lower HT. I drove a single 'Richard Allen' 8 inch, 3 Ohm round speaker from that amp, and at around 110V HT it had more than adequate volume for an average sized living room, unless you really wanted to drown out the pneumatic drill in the street outside the front window!
The next step should have been a push-pull version, but for some reason it never happened.
yep those valves give useful output at 100v ish ht or even less, theres a variant of it called 30PL14/PCL88 which may give more current and output for a given ht voltage, and is identical pinout to a PCL82 and near enough same heater voltage 15, PCL82 is 16 .. another alternative is a PCL85/805 , normally a TV frame output valve
What you're calling a DC blocking capacitor is generally called a coupling capacitor amongst people who build tube amplifiers and radios etc. And it is to block DC but not for the reasons you were saying. On a typical vacuum tube amplifier setup your putting 150 to 500 v say on the anode of the gain stage. The high-voltage is to attract the electrons from the cathode and the higher the voltage the more amplification you get. The amplified signal is actually a/c which will pass through a capacitor. That coupling capacitor is to block that 150 volts from hitting the grid of the next gain stage while allowing the audio signal that has been amplified to pass through to the next gain stage. I hope this makes some sense
Just a note, this would be a lot easier at 48 V, which is still UL class 2, so you can safely touch it. Amplifier output power rises as the square of the voltage, so you get four times the power at 48 V vs 24V. Also, biasing the grid of most tubes between cutoff and saturation is very difficult with only 24 V on the plate.
Back in the 50s, portable radios had "B" batteries. It was a 90v battery that looked like a long 9v battery. It fell out of popularity because the batteries were carbon-zinc (non-rechargeable) and the radios were very inefficient because of the high power usage for filaments.
Who needs a volume pot when you can change the voltage!🤘🏿
It would be fun to build a high voltage tube amp with a variac as the volume control! Want more volume, literally crank up the power!
Awesome vid! Love the toilet paper roll for the speaker surround hahah
Thank you! Nothing like a little CAD (Cardboard Aided Design) to really round out a project, haha.
Dear Usagi, excellent project!
This extremely low voltage amp is useful as a phono preamp or a similiar application. Since most tube amps need a plate voltage of 300v or or higher, most folks use transistors amps. But, I think tube amps sound better.
The quality of the sound you get is tremendous.
There are low voltage tubes made for torpedoes during WW2. A friend of mine builds APR battery powered transmitters with them.
You must have used some stout wire on those tube pins to fit them into a breadboard. I remember the old Carvin tube amps I used to play through - everything was turret board construction, and point-to-point wire. Those things were just beastly stout and had the warmest sound of any bass amp anywhere.
I solder component lead clippings onto components that i want to stick into a breadboard.
super dope, nice job!
One of the advantages with Valve Amps I've found is when using a preamplifier as a headphone amp.
I learnt that valve amps are less 'tiring' to listen to with headphones.
I had one at work and could listen to music for the whole 8 hours, whereas with a transistor class D amplifier I had, my hearing got 'tired' after two or three hours.
This appears to be a less subjective attribute that you can test for
Tube front-ends in front of a solid state output stage is the ultimate setup. No distortion from tubes that are driving a high impedance load such as a final output stage. Sounds incredibly good
hey great videos. I've been watching a bunch of them. I truly hope you get this and can answer my question. Why are you making this entire bread board if you are just hooking up external power (24V + 12V)? What is the purpose of the vac tubes??
There are some tubes that are designed for audio power at low voltage. For example, the 28D7 tube is designed to drive a small speaker from a 28V input. Some tabletop radios used 2 of them. Other ones that were used for low voltage audio included 35L6, 25L6, 48, and others. Many radios for low voltage audio back in the day that ran on 32 volts DC that did not use a vibrator used Push-Pull (Two tubes) or even Push-Pull Parallel (four tubes) to drive the speaker with a low B+. Karola, Setchell-Carlson, Delco, and other companies made tube radios that ran on 32 volts back in the day.
I hadn't ever heard of the 28D7, but that's awesome! Even in the datasheet for it they list a plate voltage of 28V, it was designed from the ground up to be a low-voltage pentode, which is super cool. I may have to hunt one down some day and build something out of it. It seems a lot of the old battery operated radios used between 30V and 60V and on the plates as that was what the batteries could put out. I actually have an old Airmaster battery powered radio hanging around that I've been meaning to restore, although it uses a 1A5GT, 1H5GT and 1A7GT (and one more mystery tube I can't place, but I think is a 1N5GT).
There's a lot of really awesome old tubes out there and it's kind of mind-boggling to think of all the differences between them!
@@UsagiElectric Believe it or not, "B" batteries (high voltage) did exist back in the day. 22.5V, 45V, 67.5V, 90V, and even 120V did exist back then. I think you can still buy some B batteries today. A company called Exell battery makes 22.5V, 45V, 225V, and 300V batteries today. Some tools use 60 volt lithium ion batteries that are rechargeable today. I think the biggest maker of high current B batteries today is Tesla. 375 volt batteries are used in electric cars. Porsche makes a battery that is over 800 volts today for its electric cars.
The 35L6 and others in this catergory are NOT designed for low voltage... the heater is at 35 volts because it is designed to have its heater in series with several other tubes and then have them all add up to 110 volts and then have the heaters directly connected to the mains in order to avoid the need for a seperate filiment supply transformer.
The plates in those tubes require 200 volts, just like any other tube.
Just because you see a 35 volt heater in a tube, does NOT mean that the rest of it runs at 35 volts.
@@davelowets if you look some of the 32 volt radios that ran from wind chargers did use 35L6 tubes in push-pull parallel. While not specifically designed for this use, there were radio models that did use them in their 32 volt models.
Do you have any videos where a simple low volt AC current is sent to a board of tubes and the output shows a higher AC volt?
You might be able to ditch the output transformer all together by reconfiguring the output tube to be a cathode follower. You'll lose the voltage gain of the power tube, but adding cathode bypass caps on the preamp section can help with that. Alternatively, setting the output tube to be a phase splitter could work as well.
I would love to build an Output Transformer Less (OTL) amplifier one of these days, but doing so with vacuum tubes is notoriously difficult! Tubes are really supposed to be high voltage low current devices, but I'm using them as low voltage even lower current devices, which really puts me at a disadvantage. Even with higher voltages, OTL amplifiers tend to use six to eight tubes as push-pull driver tubes to generate enough current to drive a speaker. It's super difficult, but in my opinion at least, one of the coolest style of tube amplifiers out there!
Now, if I could get ahold of some high impedance speakers, that'd make life a whole lot easier... hmmm, time to trawl eBay some more!
Piezoelectric speakers scream with modest voltage. 😮
You are better than all star trek you are my new hero. This is a phenomenal video. 😊❤👍
Fun hair split fact: vacuum tubes are primarily used for heating food. Smmmmmmmm ding!
If you put enough of them together, you can use them to heat the house too!
Magnetrons..... 🧲
❤
12AX7,
6V6GT,
I,
I,
Sob,
I,
❤
you!.
Cit, print, that's a wrap.
What a story for a prime time movie of the....
Cut, print, that's a wrap.
What a story of a story made into a movie that I talk about in my new book....
Cut, print, that's a wrap of rehearsals for today? of our new broadway play about a story about a story made into a movie that a book author describes in his new book.
@@waynemasters8673 No! No! No! He was NOT referring to those types of tubes. He was talking about magnetrons !
U can open teleport portals with them too.. use 5 and align them and confine the electron
Did know about the 12vdc plate voltage tubes used in Car radios in the late 1950s and early 1960s? The tubes were rated for 16 or 30 vdc max. A typical tube like the 12K5, used to drive an output (transistor), is capable of 40mw.
I think you would get more gain if you used cathode bypass caps. The cathode resistor is there to put a small dc voltage on the cathode to bias the grids slightly negative so it is operating on the linear part of the tube's power curve. Without the cap to bypass it, it is not pure DC but varies with the signal in the plate current. This effectively will cancel out some or all of the input.
You won't really get the tube running in its "linear" region as you put it with such low voltage. Cathode bypass caps really won't make much difference at such low voltage, with such a small output transformer
@@EsotericArctosRight, the tubes are basically running in "space charge" mode here
Great video. Regarding the capacitors, they do remove dc for the anode signal in order to feed pure ac signal to the next stage. The value of the cap and the value of the resistor right after are forming a high pass which dictates how much bass is feeded to the next stage. There is no room for misunderstanding here.
Thank you!
And thank you for the information the capacitors. I really just tested a bunch of different values in a trial and error way until I found a sound I liked. There was very little actual science done here, I just threw parts at a board until it sounded good, haha.
Yep. That's exactly what the cap and resistors do. Block any D.C. from adding an offset to the next stage and then it also create an R.C. time constant that affects the bandwidth of the incoming audio signal.
The output impedance of the previous stage, AND the input impedance of the next stage that it is feeding, MUST be accounted for in order to come up with the correct value cap and resistor that will end up doing it's job correctly and also pass the correct bandwidth of audio through to the next stage. The math can be rather complicated in order to get this correct. Get the values wrong, and you end up with really crappy sounding audio passing through, and you can really end up wasting your gain also.
@@UsagiElectricAh, experimentation.... Yep, that's what helps a person learn what does what, as long as you learn why. 👍
To avoid the issue of having the phone think it's placing a call, add a decoupling capacitor between it and the amp. Phones detect commands on the audio jack thru changes in resistance on it.
Ooh, that's actually really smart, I never really thought much about what the phone is detecting!
Thanks for the video!
Nostalgic that never gets old
Thank you, very informative, I just uploaded a few tube Amps that I regularly use.
Thank you very much!
I wondered if you were a radio ham? It would be nice to see some low voltage ham radio projects. I have some low voltage 28d7 power output valves (4) I would love to find a use for. G7VFY.
Thanks this is awesome. I knew the answer to your question, what are tubes really good at. I think I’ll build one. 👍
Thank you very much!
Let me know how your project goes!
Mate, your explanations are the best, most understandable I've found in the past few months of research - I'm making low voltage sound makers from a box of old tubes I've borrowed form a friend, from simple motorboating oscillators, to now starting to develop driven filters... this one will help me a lot I think - I wasn't sure on the source or real use of the transformer though - does it convert the DC to AC or something else?
cheers,
Chris
Thank you so much!
That's awesome! You're essentially building low voltage tube Eurorack components, that's something I've been wanting to do for a while now. Do you have some links to your projects, I'd love to see them!
The transformer is essentially to convert voltage levels. So, if we remember P=VI, our output power from the tube and the power going into the speaker are the same, but the tube is very low amperage and (relatively) very high voltage. However, the speaker wants very high amperage and very low voltage. Power stays the same, we're just dropping the voltage level to very low and bumping the amperage up to very high.
So, in this version at 36V I think we get around 0.6W of output power (maybe, I could be totally off in calculating that), but on the tube side, that's actually 36V and 18mA. The transformer changes that to 2.2V and 285mA for the speaker. Still the same 0.6W of power, just at a different voltage with different amperage behind it!
@@UsagiElectric ah, that could be (one of) my missing link; it’s less eurorack, more art project at the moment - Convert this box of tubes into a little city of sound to install in a dead factory where they were originally made... looks like that old telephone handset I tore up to get my first capacitors will be getting a revisit, as I recall one in there somewhere - might need a hundred more though :)
Thanks 🙏
ps I am sampling the heck out of them as I build/rebuild, not very scientific but learning a lot
@@UsagiElectric btw, I forgot to send you a link, so here's two :)
soundcloud.com/christian-krupa-1/a-short-improvisation-on-dual
soundcloud.com/christian-krupa-1/inherent-tension
they're both being amplified through an iRig interface into my ipad, and processed / enveloped in AUM with other auv3 effects - I found my wee transformer but haven't yet figured out how to get it to drive a speaker directly though...
The main reason output transformers are big is not voltage related but bandwidth related. To get the frequency response, they need a lot of iron in the transformers so the magnetic coupling at the wide range of frequencies can happen. That's why a 25Hz transformer is so much bigger than a 50/60Hz transformer.
That speaker "driver" you got might also be a high frequency/tweeter driver as well. They are often called drivers when sold outside of a cabinet. :)
The 33 and 100 nf blocking capacitors block the DC component coming from the 24VDC power supply but allow the AC audio component only from the amplification to pass. Basic capacitor theory - Passes AC , Blocks DC. The Dc has no usable audio on it but would bias the tube. As mentioned elsewhere the caps and resistors play a role in how much bass you get out. Out of curiosity, did you measure the audio output distortion compared to the input? What makes or breaks the quality of the audio sound is the distortion level caused by the amplification process. Higher distortion levels for non critical audio apps can be acceptable, not so for good audio for music! The lower the better quality music.
I've been accumulating tubes over the years and always wanted to build a preamp or any audio circuit with a few, but the high voltages, expensive parts and inexperience with valves kept me away. But this example , with low V , easy to find parts and your explanations of the circuit has got me ready to give it a try..thanks man!"
Thank you for checking the video out!
The high voltages of tubes are definitely the scariest part of them, but you can still build quite a lot of really cool stuff at lower voltages that are still safe. I'm using in total, 36V here, but you can even push that up to 60V-ish and still be pretty safe!
@@UsagiElectric I think the voltage multiplier would be the key to get B+ voltages. You can use 24vac and multiply from there. I want to try building an amplifier using just ac power to start from. The output transformer will be the challenge, someone told me to try a Pa transformer probably the 70 volt type. You can only use it on single end audio amplifiers won't work on push pull amplifiers. 73
@@ronb6182A voltage multiplier wouldn't have enough current to power a tube circuit. Today, one can purchase an adjustable boost converter, that will provide 100+ volts out of 12-24 volts, online for a few dollars. Sure, they have some high-frequency noise to them, but they're fine for playing with tube circuits. The best thing about those boost converters is that they're isolated from the mains. No issue with accidentally becoming a ground and a fatal shock from the mains.
As far as expensive output iron goes, a car ignition coil is cheap, and works well in many lower power tube amp circuits.
So, for less than $30, you could have an isolated adjustable high voltage power supply that runs off 12-24 volts, and a large output transformer....
Now, order it up, and get to experimenting. 👍
Ha ha, that transformer looks like the type you get in 1960's/70's transistor radios.
It's so tiny! It does put out a decent sound though, despite how small it actually is!
i hope you have a polarized line cord because i see a potential death cap.
a death cap is a capacitor that connects the hot side of the line cord to the chassis and can be deadly as they usually short and leak the whole line current through the device.
a polarized plug prevents the dangers by making sure the grounding cap connects to the neutral side.
One of my electronic textbooks was all solid-state, except for an appendix about vacuum tubes. One comment has stuck with me: "The vacuum-tube triode works much like an FET transistor."
That's actually really true! Though, I have very little experience with FETS, haha.
Cool project! The negative feedback circuit is there to limit distortion, Fender was trying for clean and loud and cheap. I cringed at the transformer and driver selection, these are usually selected with speaker impedance and the reflected plate load in mind for maximum efficiency. I guess a wide range will work, but when working with 100s of volts, heat and magic smoke are a concern. You could probably get a more usable volume with a push pull setup, but transformer selection might be a bit harder.
Thank you!
Yeah, the majority of this was built with random scraps from other projects I had laying around. I just wanted to see if I could get a decent sound out of random junk, and I was pretty happy with the results!
He could just series two of those transformers together, use the common input terminals as the center tap, and then parallel the outputs.......
😂
Imagine the math to try and figure THAT one out... 😳
2:58 "But I'm not above stealing another design". Said almost every single amp and pedal company ever.
Haha, that's like when I tell people I know how to program, what I really mean is I know how to copy and paste from stack exchange!
Much appreciated this video, it's also good to know that tube circuits can be low voltage too, it's more beginner friendly this way... If you want to try Pentodes, go with something like 6AK5(EF95) or 6AU6(EF94) or even 6DS8(ECH83). Your output transformer looks like a telephone line transformer. Is that correct?
Thank you!
I've actually got an entire series dedicated to building a processor using the 6AU6 at just 24V (with -12V for grid biasing). The 6AU6 is a brilliant little tube and works surprisingly well at low voltages!
I actually have no idea where the output transformers are from. I ordered them years ago on eBay and they were listed as just "Audio output transformers". I didn't look closely enough at the pictures because they were about 1/4 the size I thought they were going to be!
Always a good idea to assemble everything at 20v. Everything works ok.... Replace the caps and tubes with ones rated for 300v.
Good project, nice too see this. I hope to realise this for my self. In the seventies I have build amplifier but they were high voltages units. But which vacuum tubes did you use. Thank you very much.
Thank you very much!
For this one, I used a 12BH7 dual triode for the amplifier and a 12B4A single triode for the driver tube. They're actually not specifically low voltage tubes, I'm just running them at low voltages, which means almost any tube can work! All it takes is some tweaking of resistor/capacitor values and proper expectations that it won't be all that loud of an amplifier.
@@UsagiElectric Thanks for the information.
On your vacuum tube computer, I was just wondering did you consider sub miniature tubes? I’m just starting out with vacuum tubes and my mind is flooded with ideas. I’ve been scared to use anything other than sub miniature tubes because I’m worried about working with higher voltages. Can you also maybe recommend one of your videos that would be a good introduction on tubes? Thank you 🙏 sorry for the long comment.
It would be helpful to measure the power in and power out to show that it actually amplifies. My IPhone is fairly loud by itself. You could measure the AC input voltage, then the output of each stage, and we could look at amplification that way.
I love your videos on tubes. Is it hard to do and stereo FM radio receiver?
Hi
Great setup, I like to make something similar, with more sound output, may be from a 32V ex HP printer supply.
May be you can make a series heater setup, 2x 12V = 24V, with an extra power resistor, if the voltage is over 24V. Then you can get away with only one supply voltage of 24VDC or a bit higher if you have the power resistor. (You Might use a triode pentode combo for the last tube like ECL86 or PCL86 and PCF80 for the first tube, the P types are from old TV's with a current of 300 mA and different heater voltages)
Hiya!
Thank you! You're definitely on the right track to where I went in my next episode. More voltage means more power, so even a little more voltage above 24V makes a big difference. There's also a few other things playing against the overall sound output. The speaker and output transformer being the biggest bottlenecks in this one. Having said that, I do manage to get some decent sound out of it in the next episode, so definitely stick around for that!
I have a few triode pentode combo tubes, so I may have to try tinkering with those for future audio projects! I don't have any ECL86s, but I do have quite a few 6GH8s, which are pretty decent little combo tubes, and the pentode can be triode strapped for even more linear response. I would love to build a proper high voltage amplifier someday!
Use a car ignition coil for your output xfrmr. They're cheap, and they work fairly well for experimentation
I have some good information here if you gonna read it:
the decoupling capacitors are meant not to change the signal but to pass it through without getting the B+ voltage on the next stage as that would burn the tube in a fully turned on state, so generally these caps pass the audio AC signal through and block the DC which is on the plate. I am not sure if lowering the plate resistors this much on the pre-stages is that good since tubes are high impedance devices it may not be able to pull the voltage low if the resistor value is too small leaving the signal clipping like crazy. A properly balanced resistor and tube bias should leave the plate voltage floating around half of the supply or so. this makes the voltage potential for signals work in both directions without hitting upper or lower saturation points.
so if you have 24 volts B+ your plate voltage should be like 12-15 volts without signal to be good. and to increase gain you can put parallell capacitors across the cathode resistors on the pre-amp stages as well with a value of 100uf or even more and or less depending how the sound gets. you can also use a way bigger cap on the output stage cathode such as a 1000uf or even bigger. remember you wanted to get amps at low voltage, this means you need to have huge caps to bypass the resistance of each step as much as possible. the last thing I can mention is that the bias or cathode resistors could instead been smaller, since the tubes require more negative grid voltage the higher the voltage gets you don´t really need that much when you go that low. the bias is again the thing that puts the tube into just the right amount enough into to a middle point between on and off state to pass a signal without reaching the upper or lower saturation points. so depending on the actual voltage the values may need to be tweaked for peak performance and audio quality
That is some excellent information, thank you for letting me know!
The resistor and capacitor values I ended up with here were actually just from about a week straight of trying different values until I found a combination that gave a sound that I liked and that didn't clip. Having said that, there's definitely a better way to go about building this, and I definitely think you're on point here!
@@UsagiElectric No problem, I am always glad to share knowledge with others about stuff like this. In fact when running tubes on very low voltage it may even be necessary to bias the grid positive to even increase the flow more simply by making a voltage divider from B+ and grid 1 ground preferably with a pot so that you can tweak the voltage to be just right for best sound. If this method is used and working you can omit the cathode resistor and bypass capacitor since you won’t need them anymore to bias the tube. The good thing to do is to measure the plate voltage or even better use an oscilloscope to check which part of the signal that starts to clip first. Then you know where to bias the tube to gain most signal before clipping :)
A bit late to the party watching this video, but absolutely loved it! Usagi demonstrates that despite there being a science to designing "good" tube amplifiers, a lot can be done by playing around with rudimentary designs and parts you have laying around. As always experimentation is the joy of electronics as a hobby! One little tip for anyone building a very low power audio amplifier is that if you have a large speaker lying around the place, use that. The common idea that small amplifiers can only drive small speakers could not be further from the truth - generally speaking, the larger the cone area, the greater the output volume will be for a given input power. Older speakers from the '60s and '70's are particularly good for sensitivity.
Pretty cool amp ! ....Swap out your Interstage-- Coupling Caps to 1uF for much Better sound...Your Undersized 35nF caps make great Hi pass Filters with Little sound content
Is that a bad hair day or did you become ground one too many times in your life ahahaha jk yo I'm a brand new subscriber to your fine channel, very comprehensive presentations my man keep up the good work pal!
Interesting low voltage vacuum tube amplifier
Sound great for such a small guy!!!
Thank you! I was pleasantly surprised at the quality of sound it could produce with such a low part count and such low voltage!
He said, "...how to make this cooler..." That's easy. Replace the tubes with some MOSFETs. 2N7000s for the small signal amps, and a power MOSFET for the power output. It will run a lot cooler. 😊
Bypass caps, try reducing your cathode resistance too
I'm only a few minutes into the video..... But when you said let's switch the 100K to a 10K on your first tube stage...
I think you'd do better to switch it to a 200k or even 300K. You will want all the gain you can get.
I believe it just works that way. More technical than that. And of course after that, bias it properly as you really want to find the correct knee for your voltage of that tube.
Without a cabinet, the back side of a speaker will generator the opposite phase of the front, thus cancelling out much of the sound when they mix together.
@Usagi . . . BIG Question: how did you come up with your voltage and component selection and values? Do you do your modeling with LT Spice or some other circuit-modeling software?
That's actually kind of a convoluted story, haha.
In this video I use +24V, then +24V and -12V together to essentially get a B+ of 36V and then finally in the next video I use 48V.
So, to start with, my choice of +24V. This is actually completely centered around my primary tube project - building the vacuum tube computer. I really wanted to see if I could build some good tube circuits at voltages low enough that anyone could get into it, even children! I tested tubes at 6V, 12V, 18V, 24V and 30V (pretty much the entire range of my bench power supply). I found that anything under 18V was starting to push it, and I was having a hard getting reliable results. The next problem was figuring out heater voltages. Since I was planning on using pretty much the 6AU6 tube exclusively, I figured I could get away with running the heaters in series. So, 24V meant I could build everything around 4-tube modules.
Next, the -12V. To get the logic circuits working reliably and properly, I need a negative voltage strong enough to push the tubes fully into cutoff. Using a -12V supply gave me the perfect negative bias, and it meant if I needed to, I could run two tube heaters in series off the -12V. This gave me flexibility to build 2-tube, 4-tube or 6-tube modules.
So, I built my power supply around those voltages, and since that's what I had on hand, that was what I used for this video. However, ultimately, I wanted to run four tubes, and since I was using tubes with 12V heaters, I needed to think about how to power them. I did a bit of searching on Amazon, and it turned out that 48V power supplies were pretty common and super cheap, and that meant I could run four 12V tube heaters in series and use that 48V for the B+ for a little more volume over the 36V version.
A bit of a convoluted story, but that's how I ended up with the voltage you see here!
As for a modelling software, I do tend to check everything in software initially. I use TINA TI with some imported SPICE tube models to get me in the ballpark, then I trial and error different resistor and cap values on the breadboard until I find the sweet spot for the tubes I'm using.
Instead of thinking of the caps as DC blocking, think of them as AC coupling. They let AC (the audio) through, but gets rid of the DC bias
I always had a hard time with wrapping my head around capacitors letting AC through. There’s no direct connection inside a capacitor and we’re not inducing any voltage with magnetic fields like in a transformer, so how can AC pass through?
What ended up clicking in my brain was that AC isn’t passing through so much as the capacitor is just charging and discharging in sync with the frequency of the AC. When the AC is on the positive phase, it charges up the capacitor, making the other leg of the cap act like a ground pin. When the AC is on the negative wave, the capacitor discharges, making the other leg of the cap act like the positive pin. (At least, that’s how I’m seeing it in my brain, that’s probably not the best way to describe it though).
@@UsagiElectric And isn't the electricity flowing one way on the charging and then the opposite way on the discharging? So that electricity never actually flows through a capacitor?
@@UsagiElectric in the process of trying to build a low voltage tube amp for a pair of old 1920s head set I’ve got, and remembered this video, hoping it will work, any how seen the comment about capacitors and I figured I’d add some more info the Charge is being stored within the dielectric of the capacitor, Benjamin Franklin explained this when the layden jar was created, and somehow it changed to the metal plates!
@@UsagiElectric I've always just accepted DC blocking caps without properly understanding until you explained it.
@@edgeeffect There's also AC-blocking capasitors, called capasitor dropper. Take mains voltage it is 120 ACv 50Hz or 230-240 ACV 60Hz. When the caparcitor charges on the rising voltage it passes current trough, but when it is fully charged the current stops to flow. So based on the capacitor value and the current flow on the load it drops the voltage as it only lets part of the wave in each "hump" through.
nice do have some mic tube preamp that can distortion? non-transistor project
Marvelous
This is cool. What is power rating on the resistors?
DC Blocking (Bypass) capacitor is easy. The plate of V1 has positive voltage on it. +24 in your case, +210v in my case. A cap can pass AC (audio signal) but not DC. +210vDC on the GRID of V2 would cause it to run away and fail........ the Bypass cap Blocks the DC from the plate of the previous tube.
you should have looked up the vacuum tube sound card for pc's then, use that as a basis since it only has 12v to work with
Would be interesting to see what happens when he puts the negative feedback back in. Just how much power is being shunted back? Is it worth the increase in fidelity for the loss in power.
I bet this amp produced a couple of mW! Tubes are high voltage devices and should be used on high voltages to get a decent output power. A few years ago I have built an audio amplifier with two transmitting tubes operating at 5000Vdc and produced 1300W of audio power ! That's a real challenge.
tbf 24v can be VERY loud, 10w kinda loud
if u have ever made a solid state amplifier
there are many other factors that it is not as loud
from the coils in the speaker, to the amp design
and even tubes themselves, as the low voltage region would be unusable
and for amplifier, u really gotta get into that linear region, if u seen graphs of tubes, they only get into linear region after few tens of volts
Insist to make it in mono block. Yeah, it costs more (alot) in PSU but we'll be rewarded with tons of soundstage & air.
the 100nf and 33nf capacitors are just to isolate the dc voltage to the grid of the next stage
Yup, those are the coupling capacitors!
Collins radio like R392 used a dual pentode 26a7 in pushpull at 28v about 300mw out.
i'd like to modernize a Magnavox sterteo amplifier made in 1960 that i'm sure can sound great, and get reasonably loud, have you done any videos regarding them [small 3 tube amps]
I'm not that familiar with old radios unfortunately. Still, it shouldn't take much to get one of them up and going. The tubes are generally pretty hardy and the only real weak point would be capacitors. If there are any wax paper type capacitors in there, change them out for new ones of the same value and then fire it up and give it a go!
@@UsagiElectric i have the new capacitors and resistors that i need here already but my other question is concerning the place where the signals from cartridge come in,is there a way to use a modern cartridge with out it sounding tinny?
Nice build,
Saturation is coming from the transformer. It is not made for DC. It should be a gaped core one.
For the rest there are 12V tubes that exist that were used in car-audio conceptions
Wow ... 10,000 views on this, as I click here today ... congrats ...
Thank you! This video has really picked up some steam lately!
Nice video , I think if you can use better Output transformer the efficiency will improve even more or deliver more power. I like low voltage tube amplifier..
Thank you! I definitely think you're right! Those output transformers are a little too small to give proper good sound, but I have a handful of them, and I really wanted to use them on something, haha.
@@UsagiElectric Normally they have lower impedance , in my old book there is ST-32 which is 1K : 8 ohm they use for 9V battery , i did clone it with ferrite core in my youtube video.
if u can find a small scarp transformer maybe u can wind urself..I have several 16K to 8 ohm and 32K to 8 ohm use for my vacuum tube experiment..
@@6p1p Just checked out your channel, you have a ton of awesome videos on CRTs! I'm going to have to check those out as soon as I get some free time to binge some CZcams!
The output transformer is one of the most difficult parts of a building a tube amplifier I think, because the rest of the amplifier can be amazing, but if the output transformer is mediocre, it'll totally kill the sound! There's a definite art to getting it all right.
Your 6J6 push-pull hybrid amp actually sounds really good, that's an awesome build! I'll dig through your channel more as I get more time!
@@UsagiElectric Yes i agree with you , to make it sound good the output transformer is the most difficult one.. I tried wind my output transformer since 2004years and i wasted alot material for keep try and error.. If you just want it work and sound ok not as grade as hifi , you can use any power transformer core , the hifi grade they normally use Oriented core Z11 or M6 lamination core which the thickness is 0.35mm , typical transformer core is 0.5mm thickness. Normal core will work good as well. The only you need to wind it . The winding isnt difficult just need patience to do it. The ratio of primary turn versus secondary turn , square of it will be primary impedance versus secondary. for example 5K:8 which is ratio of 625 , and square root of 625 is 25 , so your primary need to have 25 time number of turn of your secondary by using this u can make all transformer u want.. the primary inductance impact your low frequency bass the more the turn the better the low frequency response. If u have any question is welcome to message me. good luck
@@6p1p Interestingly, I think my next proper audio project is going to be trying to build an Output Transformer Less (OTL) amplifier. Everything I've read about them makes them seem like the vintage Ferrari of amplifiers - super finnicky and difficult, but when running right, better than everything else out there. I think I understand it at a fundamental level. You've got the standard amplification part, then you've got a ton of tubes in parallel to provide enough current to actually move a speaker.
There's a few different schematics out there, though this one seems to be the most common: www.bonavolta.ch/hobby/images/audio/6as7_2a.gif
There's a lot of really interesting things going on in that schematic too, I particularly like the dual triode amplifier stage with the cathodes tied together, that's an interesting way to do it. I wonder if that was done specifically for some aspect of the OTL, or if that just so happened to be the preferred amplification method by the designer?
At any rate, I love the idea of a tube amplifier that uses just plain brute force instead of an output transformer. One of these days, I'll get around to actually giving this a go!
Capacitors do block dc, in that configuration only passes AC audio signal to the next stage, is important to operate correctly also on high voltage operation are necessary to basically avoid red plating and grid overload. Thats because if you watch Mr. Carlsons Lab or David tipton or anybody who restore old radios and purchase this type of used equipement always says dont plug it to the person who is selling, because those capacitors if they are too old papers and bumblebees turns into a resistor, those always go bad and that destroys tubes an cremates the output transformer.
I always had such a hard time understanding when people would say caps block DC and let AC pass, and you can have both on a circuit. I think it's more of a terminology problem for me though. Fundamentally I understand exactly what's happening in terms of signals, just the wording always tripped me up. For my brain, it helped to think about the output on the plate being an oscillating waveform that moves from +10V to +20V for example. The capacitor shifts that signal to -5V to +5V. Still a 10V P-P signal, it's just shifted the center point to be around ground. It does this because as the voltage at the plate increases, it charges the capacitor. And then, as the voltage on the plate decreases, the capacitor feeds that voltage back (because it's at a higher potential). The result is that the other side of the coupling cap changes potential in accordance with the charging and discharging of it, which replicates the signal but eliminates the DC bias.
That's probably not the best way to explain it, but thinking about it in that way is what made it click in my head!
Build the Worlds Second Best Speaker a match made in heaven, even if you can not hear it.
You know you can use 24 volts ac and make a voltage doubler or more using diodes and capacitors. And there are tubes that were in battery radios before we had transistor radios. They used 90 volts for the plates or known as B+ voltage the filaments in the tubes were one and three volts. To get 90 plus voltage you can make a voltage quadrupler using diodes and capacitors. I believe theses circuits would be fairly easy. As for output transformers maybe try a Pa line transformer like the 25 or 70 volt type. Power transformers is another option. 73
Yup! But the main purpose of this was to see if I could build a tube amplifier with a super common low voltage.
Granted, with limited voltage, we have very limited power, but it still produced a surprising amount of sound for such low voltage! This is definitely not the best way to build an amplifier, but it's certainly an approachable way for anyone wanting to get into tubes without being quite ready for higher voltages.
You're definitely right on the output transformer, a properly sized one would probably make a much better sound!
@@UsagiElectric yeah I know there are a lot of fake tube headphone amplifiers out there . I bought one and it had a solid state amplifier for the output. Low voltage plate current can be used for headphones and with good sound. It's a shame they don't make high impedence headphones.
I did not.comment.to take away from your idea of low plate current but to tell people we are not limited to 12 Volts or 24 volts. I just threw another option out there for higher.voltsges using only 24 vac. Power transformers are hard to find at a low price. I commented on another page how to use two transformers for isolation. There are many ways around using what you already have. 73
Vacuum amplifier is the best amplifier for audio. Audio system that cost more than 1 million dollars use pre amp, and amplifiers that use vacuum TUBES
Thanks for the interesting video, i just ran across a regenerative receiver (from France) using the same tubes and operating on 6.3 v on the filaments and only 12 volts on the plates , i havent built it yet but the author says it is very good ! Thanks ! K9UT
Don’t sound too bad. Perhaps you can design a push pull amp. You single ended triode is a good start.
Thanks!
I actually really want to design a proper OTL amp someday. I built a low voltage OTL using a boat load of 12B4As and it put out a really great sound comparatively, so I think it would be a ton of fun to expand upon that idea!
@@UsagiElectric I saw the video, it worked well.
you could try using a dual section 6080, either paralleled or in push pull, or similar for the output section,(or a 6N7, lower power and lower heater current) that should give more current and output for a given HT/B+, only problem is the very high heater current 😉 and they need a fairly high amplitude drive.. there may be alternative 12 or even 24/25 v heater versions of both those types, i dont know...you' d need a bigger output transformer obviously ..or if using the same one as in the video, maybe an ECC88/6DJ8 in push pull output, you 'might' be lucky to get 200mw or so output,
some years back at a vintage radio rally, someone displayed a radio he made that worked off a 12v lead acid battery, all valve, using special valves designed for car radio use at 12v supply, there wasnt any high enough 'power' valve in this range as wasnt possible, so car radios using that valve range was hybrid with transistor output... he managed to get 200mw or so output in that radio by having push pull output, cant remember what exact valves were used, but the RF ones were ECH83 and EBF83 , with EF98 audio driver,
So funny to watch.
What about raising the cathode voltage with a diode? That would be +.6v compared to ground. Could that work?
That's actually a really smart idea! I never thought about it, but that would indeed elevate the cathode above ground and potentially get some more gain. It's certainly worth a shot anyways!
Would a step up transformer bring any help?
Nice.
You need cathode bypass capacitors, i didnt see these in your schematic. These increase the AC gain of each stage by eliminating cathode degeneration.
I'm using a cathode bypass capacitor on the driver tube, but not on the amplifier dual triode. The reason for this is because I was trying to build a low voltage replica of the famous Champ Amp, which doesn't use cathode bypass caps on the amplifier stage dual triode either. When it comes to audio amplifiers, there's a million ways to build them to make them fit whatever goals the listener has in mind. My goal here was to just take one of the most famous and simple tube amplifier designs out there and see how it did at low voltages!
I have something I wonder if someone can help me on. I have a vacuum tube Geiger Counter... I just bought a 12v to 135v inverter for the B+ supply, but I am thinking the filament supply would be power hungry on batteries. That got me to thinking... I would like to see if it is possible to make solid state tube replacements...
Do the tubes glow with low voltage?