Valve (tube) amplifier schematic help

[CDRIVE]

Here, it's perfect.

http://www.mpja.com/9V-111A-Desktop-Supply-Fairway/productinfo/17722+PS/

The mains power cord is sold separately.

Chris

 

[Solidus]

Earlier when I was suggesting building my own supply I was just indicating that I wanted all the different supply components (amp transformer, filament / tube supply, LED driving transformer) to be routed to the same plug, I wasn't meaning that I didn't like the idea of the wallwart. I just don't want to have 3 different plug-ins routing into the back.

It won't be too hard to get some wire connectors or strip the power cord for the wallwart you linked to and route to the master plug inside the unit though, so I like the idea.

And yes, I just didn't feel working with mains directly (building a complete DC filtered and rectified circuit) suits my skill level. That's all.

 

[Solidus]

and the power cord is not an issue, I have a habit of taking files and rasps to various power cord assemblies to make them cross-compatible with one another

 

[CDRIVE]

and the power cord is not an issue, I have a habit of taking files and rasps to various power cord assemblies to make them cross-compatible with one another

You're scaring me. The darn thing cost only $1.25!!

http://www.mpja.com/prodinfo.asp?number=15447+CB

 

[Solidus]

I was saying that to lighten the atmosphere from "electronics newbie working on mains" to something more friendly. Haha.

Although I did on one instance. It was that same cable for my laptop inverter, except one side of the only cable I had was square-faced. So I grabbed a nail file and shaved the side of the plastic down to make it fit the inverter's input jack.

It's not something I intend to do again, I assure you. I want this project to work, sound good, and be safe most of all!

Speaking of the progress, I've got done machining one of the acrylic panels and I'm working on mounting the pots to it. Want a photo of the progress?

 

[KrisBlueNZ]

We would love to see your progress.

There is a "project logs" forum on electronics point you could be using to record your progress, but I don't think it gets much action. davenn or Steve, what are the advantages of using that forum? I'm planning to use it myself soon.

BTW I don't think anyone has summarised the situation re powering the filaments. The ECC8x (12Ax7) series have centre-tapped 12.6V filaments. You can power them from 12.6V connected across the ends, in which case the current is 150 mA, or 6.3V if you connect the ends together and power it between the centre tap and the ends, in which case it draws 300 mA. The filaments can be powered from AC or DC; the valve doesn't care, because the filament is isolated from the other electrodes. Powering the filament from AC won't cause (significant) hum.

Someone suggested that if you're using a 12VAC transformer with rectifier and smoothing for the main supply, you could power the filaments directly off the transformer, to reduce load on the DC rail, which means less smoothing capacitance needed and less ripple (hum). The filaments draw FAR more current than the rest of the circuit. I think that's a good idea if you're using an AC power source or building the transformer into the box.

That design runs the filaments at 9V instead of 12.6V. This will affect the valve's behaviour to some extent. I don't know whether running them at 12.6V (series) or 6.3V (parallel) will be a problem. My gut feeling is that it wouldn't make much difference.

I would avoid using a switching converter to supply audio circuits. They create common-mode and differential-mode noise over a wide frequency range, which is difficult to filter. In some cases, just connecting the negative side of a switching power supply to the earth connection of an amplifier can cause audible noise in the amplifier. It would be better to use a mains-frequency transformer or an adapter that has a mains-frequency transformer, rather than one of the modern, compact, lightweight adapters, which use switching converters.

 

[CDRIVE]

Speaking of the progress, I've got done machining one of the acrylic panels and I'm working on mounting the pots to it. Want a photo of the progress?

Do Cows fart? Of course I'd like to see it!

 

[Solidus]

I just took some photos, now I'll host them on photobucket and should have them up in the next post!

The filaments can be powered from AC or DC; the valve doesn't care, because the filament is isolated from the other electrodes. Powering the filament from AC won't cause (significant) hum.

Now does that make an issue in this schematic, as the plates at pins 1 and 6 are coupled to the filament (9VDC intended) line?

That design runs the filaments at 9V instead of 12.6V. This will affect the valve's behaviour to some extent. I don't know whether running them at 12.6V (series) or 6.3V (parallel) will be a problem. My gut feeling is that it wouldn't make much difference.

The original posts from which I got the idea quoted 9VDC as appropriate input. Of course, without even needing datasheet references you can pretty much tell that the valve intends 12V.

Something tells me there's a non-linear curve between filament voltage and valve behavior, that is, dropping the voltage by 25% doesn't correlate to a 25% drop in activity or performance. So far, as of yet it's not a worry for me.

I would avoid using a switching converter to supply audio circuits. They create common-mode and differential-mode noise over a wide frequency range, which is difficult to filter. In some cases, just connecting the negative side of a switching power supply to the earth connection of an amplifier can cause audible noise in the amplifier. It would be better to use a mains-frequency transformer or an adapter that has a mains-frequency transformer, rather than one of the modern, compact, lightweight adapters, which use switching converters.

Well, I just purchased the converter Chris linked me to.

Chris - do you think that converter would cause a noise / hum issue?

If not, I can eat the $7.20 I spent on it if a better option presents itself or if it causes a problem in terms of noise.

 

[CDRIVE]

BTW I don't think anyone has summarised the situation re powering the filaments.

Actually, I did but probably not sufficiently and clearly enough for a nube.

https://www.electronicspoint.com/valve-tube-amplifier-schematic-help-t251475p5.html#post1488128

Chris

 

[Solidus]

This is the stage where I was drawing the forms to the acrylic panels. I casted the forms on cardstock using engineering rulers and compass / protractor.

I then translated the forms to the acrylic panels with a sharpie - real high-grade technique

This is the finished acrylic front panel after drilling and cutting. This is the second run of the panel after the first one snapped through during the drilling stage - the problem with acrylic is while it is dirt cheap, it is the most brittle of the clear constructing polymers and the carbide bits that I use to drill polycarb, ABS, and PVC torqued the plate too much and there goes one panel!

Total drill time was probably about an hour, once I got bits selected and clamping adjusted to how I wanted it. The bits had to be eased in - it wasn't just "git 'er done" status.

The photo still has the protective, semi-opaque shrink-wrap mess as well as drilling burrs on it.

This is the same plate after I deburred it and removed the protective cover wraps. My hands were greasy and laden with sawdust, so I used the best deburring method known to man - the side of my finger.

And yes, that's a very crude drawing of a bridge rectifier used as a contrasty background.

What would a DIY homebrew guitar amp be without some of the most obnoxious-looking toggles known to man?

Has all the values a college student needs - colorful, unnecessary, and it'll look pretty cool with the circuitry inside.

The same panel with input and f/x out jacks mounted, as well as the tone pot mounted and the volume pot sitting in the foreground. Still waiting to hacksaw off the unnecessary part of the pot lead before mounting.

As you can see, the ground leads are being wired to a set of terminal blocks. The tone pot's ground lead, however, is still unwired. I am waiting to braid it to the volume ground when mounting.

Mountain Dew was very important for the conception of this project. Very important, indeed.

 

[CocaCola]

the problem with acrylic is while it is dirt cheap, it is the most brittle of the clear constructing polymers and the carbide bits that I use to drill polycarb, ABS, and PVC torqued the plate too much and there goes one panel!

Hobbyist access to low cost laser cutting is out there now to help remove some of those pains, if you do the layouts in a vector or CAD format...

I have used these guys before and I have been presently surprised at the price they have quoted me for cutting and/or etching...

http://www.pololu.com/catalog/product/749

 

[Solidus]

Hobbyist access to low cost laser cutting is out there now to help remove some of those pains, if you do the layouts in a vector or CAD format...

I have used these guys before and I have been presently surprised at the price they have quoted me for cutting and/or etching...

http://www.pololu.com/catalog/product/749

Hmmm...very interesting, I'll give it serious thought on the next project.

Better than going through a bunch of drill blanks!

 

[KrisBlueNZ]

Actually, I did but probably not sufficiently and clearly enough for a nube.

You're right, you did summarise it well. I hadn't been following the thread closely. Apologies.

Re powering the filament from 12VAC or something else:

Now does that make an issue in this schematic, as the plates at pins 1 and 6 are coupled to the filament (9VDC intended) line?

No, that's just how the schematic happened to be designed. Because there was a 9V DC rail, the designer decided to power the filaments from it, rather than power them separately. But you don't need to follow that design. The filaments aren't connected to anything else inside the valve. You can power them from whatever you want, including 12.6 VAC (between 4 and 5) or 6.3 VAC (between pins 4/5 and 9).

I think this was originally Steve's suggestion, and it's very sensible. The filament current is huge compared to the current used by the rest of the circuit. For a given amount of ripple on your DC supply rail, your required smoothing capacitance is proportional to the current drawn from that rail. So if you can reduce the load on that rail from say 155 mA to only 5 mA, you can reduce the required smoothing capacitance by a factor of 31 as well.

The original posts from which I got the idea quoted 9VDC as appropriate input. Of course, without even needing datasheet references you can pretty much tell that the valve intends 12V.

Yes, we know that's what the valve is specified for. I think there was some text accompanying that article where the designer tried to defend the use of 9V..

Well, I just purchased the converter Chris linked me to.

I think whether there's a problem or not depends on the quality of the converter - whether it complies with EMC requirements and has good filtering and layout.

 

[Solidus]

You're right, you did summarise it well. I hadn't been following the thread closely. Apologies.

Re powering the filament from 12VAC or something else:

No, that's just how the schematic happened to be designed. Because there was a 9V DC rail, the designer decided to power the filaments from it, rather than power them separately. But you don't need to follow that design. The filaments aren't connected to anything else inside the valve. You can power them from whatever you want, including 12.6 VAC (between 4 and 5) or 6.3 VAC (between pins 4/5 and 9).

I think this was originally Steve's suggestion, and it's very sensible. The filament current is huge compared to the current used by the rest of the circuit. For a given amount of ripple on your DC supply rail, your required smoothing capacitance is proportional to the current drawn from that rail. So if you can reduce the load on that rail from say 155 mA to only 5 mA, you can reduce the required smoothing capacitance by a factor of 31 as well.

Yes, we know that's what the valve is specified for. I think there was some text accompanying that article where the designer tried to defend the use of 9V..

I think whether there's a problem or not depends on the quality of the converter - whether it complies with EMC requirements and has good filtering and layout.

So the coupling of the DC and filament rails was merely condensing and simplification on the part of the schematic author?

Is the issue concerning audible hum the result of unfiltered switching supplies across the plate lines or the filament lines?

Would I have better luck powering the plate lines off the converter and running the filament off unmodified 12VAC/450mA?

Regarding the converter, is there anything I can do filter-cap-wise as a "just-in-case" measure? Is there anyway to measure noise directly?

 

[(*steve*)]

davenn or Steve, what are the advantages of using that forum? I'm planning to use it myself soon.

Only you can post there, so you won't have discussion popping up.

I accidentally started discussing something in one thread in there, and I think it has been tightened up so that even moderators can't butt in. (But we can still do other stuff)

 

[Solidus]

I contacted the original author on the schematic and did some digging around some links he sent me as well as cross-references to the datasheets.

Kris - You may find this to be interesting:

If I remember correctly, the relative gain of the 12A*7 tubes goes something like this:
12AU7 20
12AV7 40
12AY7 50
12AT7 60
12AX7 100

Also what I'm finding is that 9V seems to be too low voltage for anything higher than the AV7s. I have a 12AT7 in my possession and that will most likely need a 12VDC capable supply.

It seems that the consensus on later pages of that forum thread is that noise due to supply current is only an issue if a really hummy DC converter or badly rectified AC is used.

The forum thread I found this on.

Of course, until this device is built I'm taking everything it says with only casual note. I'm ordering varying valves in the 12A-7 series. The gain notion was what I thought - but otherwise tube characteristics probably don't change much.

12AT7 and 12AX7 are probably more useful as overdriving pre-amp valves than the 12AU7 would be. Seeing as the AU7 is low on the gain list (if we assume it to be true) and the unit is running below intended supply ranges, this is more designed to be a tone distorter than an overdriver. Then again, the original idea for this was a effect pedal, and it was my idea to turn it into an overdrive and distortion pre-amp.

Will check the 12A-7 valve family characteristics later.

Just thought it'd be an interesting post for some food for thought.

 

[KrisBlueNZ]

So the coupling of the DC and filament rails was merely condensing and simplification on the part of the schematic author?

Yes. The circuit has a single 9V supply so he used that for the filaments as well as the actual circuitry.

Is the issue concerning audible hum the result of unfiltered switching supplies across the plate lines or the filament lines?

Hum comes from the mains frequency and is caused by insufficient smoothing when working from a mains power supply that uses a mains transformer. The noise from a switching power supply is all sorts of frequencies, not particularly mains frequency.

Powering the filaments directly from a 12 VAC transformer output will help avoid hum in the signal because it reduces the load current on the DC supply rail, which means you don't need such a big capacitor to smooth out the hum (ripple).

Would I have better luck powering the plate lines off the converter and running the filament off unmodified 12VAC/450mA?

No. If you have (or are getting) a 12VDC switching supply, run the whole lot off it, as per the original design (except that the original design used 9V). You will find out whether your switching supply introduces noise or not.

Regarding the converter, is there anything I can do filter-cap-wise as a "just-in-case" measure? Is there anyway to measure noise directly?

That depends on the characteristics of the noise. Probably. Try it first and see if there's a noise problem. And I mean noise, not hum. You can measure the noise with your ear, by putting it next to the speaker ;-)

Also what I'm finding is that 9V seems to be too low voltage for anything higher than the AV7s. I have a 12AT7 in my possession and that will most likely need a 12VDC capable supply.

Right, well, if you're getting a 12VDC switching supply, try it out. It can power the circuit AND the filament, and when you test it, you will soon discover whether there is a noise problem.

Steve, thanks for your comments.

 

[CDRIVE]

12AT7 and 12AX7 are probably more useful as overdriving pre-amp valves than the 12AU7 would be. Seeing as the AU7 is low on the gain list (if we assume it to be true) and the unit is running below intended supply ranges, this is more designed to be a tone distorter than an overdriver. Then again, the original idea for this was a effect pedal, and it was my idea to turn it into an overdrive and distortion pre-amp.

Will check the 12A-7 valve family characteristics later.

Just thought it'd be an interesting post for some food for thought.

Just for the record, it's replies like this that gave me no clue that you may not understand the information and the schematics I've given you. It simply doesn't sound like an electronics nube at all. ... Really!

Now, moving on to the switch mode supply you just ordered. They don't produce hum, as you might think. It's high frequency spike noise that can easily radiate and be picked up by AM radios. Filtering this out on the supply lines (if need be) can usually be handled with small bypass caps. What's floating around in the air is another matter. This amplifier has a 1Meg input impedance. Hi Z inputs are very sensitive to airborne crapola. Uh,... that's a technical term.

The best I kept for last.. your shop and shop work. Nice work! How large are those large holes and what are the for? Did you cut them with a center bit hole saw? Did you know that they make drills with an included angle specifically for acrylic? The angle is close to what you'd see on a masonry bit.

Oops! I better get this posted before another feeder band blows though here. We've been loosing power intermittently and the battery on my laptop has been dead for well over a year. I run it off the mains.

Edit: I forgot to mention that you must be an engineering student or a home shop machinist and possibly wood worker too!
Chris

 

[Solidus]

Just for the record, it's replies like this that gave me no clue that you may not understand the information and the schematics I've given you. It simply doesn't sound like an electronics nube at all. ... Really!

Now, moving on to the switch mode supply you just ordered. They don't produce hum, as you might think. It's high frequency spike noise that can easily radiate and be picked up by AM radios. Filtering this out on the supply lines (if need be) can usually be handled with small bypass caps. What's floating around in the air is another matter. This amplifier has a 1Meg input impedance. Hi Z inputs are very sensitive to airborne crapola. Uh,... that's a technical term.

The best I kept for last.. your shop and shop work. Nice work! How large are those large holes and what are the for? Did you cut them with a center bit hole saw? Did you know that they make drills with an included angle specifically for acrylic? The angle is close to what you'd see on a masonry bit.

Oops! I better get this posted before another feeder band blows though here. We've been loosing power intermittently and the battery on my laptop has been dead for well over a year. I run it off the mains.

Edit: I forgot to mention that you must be an engineering student or a home shop machinist and possibly wood worker too!
Chris

Haha I'm a chemistry / chemical engineering student. I've been a science wacko my entire life. Got my first science encyclopedia at 5 and read it religiously. I've consistently placed 99th percentile since kindergarten in science and math, so you're seeing my reasoning capability shine!

Chemistry I excel in to a degree you don't want me to elaborate on; physics I grasp very well.

I understand a lot of the underlying notions behind electric theory and electronics and what not, it's the component-level experience that I lack. Ask me to figure some relationships and I can do that with a bit of struggling. Ask me what component goes best where in what situation, and you're going to be getting crickets.

The reason why I have so many questions and seemingly don't get these circuits well is because I have no experience in working with them, and that's most of it.

High-frequency AM noise - cue the tinfoil hats! Although I one time read an MIT study that claimed that tin-foil hats serve no protection against government mind control - apparently they actually focus and sink the frequency bands that the government uses (the "X band"). I'll try and dredge up the article later for some humor.

Well, in regards to the high frequency noise, it probably wouldn't be too inconceivable to devise a Faraday cage shield around the switching converter. Wouldn't be terribly effective, because of the ever-frustrating omnipresent notion that the effectiveness of a device is usually proportional to its mass, and I don't want to be lugging around big metal bricks.

Problem solved. Add some duct tape and I'm on my way to amplification glory!

And in regards to the panel cutting, I was pleasantly surprised with how it turned out as well. I used standard general-purpose bits in 5/16", 3/8" and 7/16" sizes. The large speaker holes I did with a 2" carb hole saw. They are for the main driver cones of the speakers. I did not know that about the acrylic bits though! If I had, it would have saved me much frustration and a half-panel of cutting.

Now that we're getting into high-frequency noise and mess, I pulled this out of my closet:

Eli

 

[KrisBlueNZ]

Well, in regards to the high frequency noise, it probably wouldn't be too inconceivable to devise a Faraday cage shield around the switching converter. Wouldn't be terribly effective, because of the ever-frustrating omnipresent notion that the effectiveness of a device is usually proportional to its mass, and I don't want to be lugging around big metal bricks.

It ought to be possible to eliminate common-mode and differential-mode noise using inductance in series, and capacitance in parallel. I think it will depend on how _much_ noise is present. I'm interested to know whether you have a noise problem initially, and if so, what it takes to overcome it. I'm trying to use a cheap-and-nasty after-market laptop power supply with an audio amp in a project, and the interference is awful!

Ha! That oscilloscope is the same model as my first one!