learning, building and testing tube amp

[fragman56]

Gentlemen/gentlewomen:

Would it possible to build from the ground up (tube out) a simple SE amp by taking directions from you (the learned ones) online--here?

For my benefit as well as other readers, could I test the components based on your suggestions/directions, and most importantly, learn the principles of tube amplification while building it? It is not important for me to have the amp at the end. Rather, the learning and experience is my goal. Perhaps there are others like me.

I have built from a kit a PP, PCB-type tube amp, but unfortunately, did not really learn a lot about the principles.

I have no doubt that I could assemble a tube amp from a schematic if given the values for the components i.e. resistance, voltage, capacitance, etc. That is not the point. I would like to eventually design/ select components.

Is it possible/advisable? Or should I stick with the textbooks (time consuming and boring).

Your comments appreciated.

Aspiring diyer,
Rick

 

[Colt45]

pretty tired, but here are my suggestions:

Look around at lots of different schemes for amps.. of different classes. see what changes between them.. and so forth.

read datasheets for the tubes involved. lots of info in there (even if you cant read the graphs.. which i'm still not too good at.)
see which tubes are similar, and what you'd have to change to make them work in the design.

read some of the NEET stuff.. (forget where it is now). its on the web somewhere. it's the military's electronics info i guess.. lots of tube theory, etc..

a real basic 2 tube SE mono deal should be very easy to do from scratch. ( I mean, youve got two tubes.. ~5 resistors.. a couple caps. two transformers. some filter caps and a tube or SS rectifier.)

I'll attatch a pic of the BASIC design of most SE amps (bare minimum deal.)

 

[phil]

Have a look here...http://home.t-online.de/home/MHuber/
I'm a big fan of Manfred's work...

 

[fragman56]

output from gain tube

Thanks for the site, Phil; maybe still over my head.

Thanks for schematic, Colt45. Pretty basic it appears! Just what I need.

Question: (Apologies for such a basic question, first.)

If the output from the gain tube is positive and is used as the control grid input to the output tube, how does it become negative thereby resisting the electron flow from the (negative) cathode?

I am certainly missing something here. I know it works, but how?\

Aspiring diyer,
Rick

 

[Colt45]

ok.. I'm not super good at this, but I'll try to explain it.

ooops.. forgot a coupling cap between the two tubes in the scheme.

anyways, in theory the grid has 0v DC on it. (and of course an AC signal [the audio]).

when the tube is in operation, with cathode/self bias (this configuration [the cathode resistor]) the cathode becomes positive! (for example we'll use say it's +10v. it really depends on the tube and resistor though.)

so if the cathode is +10v, and the grid is 0v, the grid IS negative in respect to the cathode!

neat huh?

 

[jeff mai]

If you want to design your own valve amps go here now:

http://www.audioxpress.com/resource/audioclass/index.htm

If you're familiar with the basics, these articles should tell you everything you need to know for a starting point.

 

[Brett]

Rick,

Colt is correct with his explanation of the cathode biassing arrangement for the output tube, and also that you need a cap between the plate of the driver (gain) tube and the grid of the o/p tube.

Jeff has also given a truly excellent resource in the Classroom series by Norman Crowhurst. NC was a very knowledgable engineer, and there is so much condensed wisdom in his work you will find new things in it for many years to come.

Can I make a suggestion? At the moment what is happenning in this thead is a lot of general info for tube amps. Tell us what your requirements are for an amp in terms of power etc, and we can help you design an amp from scratch, or take another schematic and pull it apart for you to learn from.

With a poweramp, start at the output and work back in. So we need to know
- power output / load impedance
- preferred tube types: DHT, IDHT, triode, trioded pentode etc
- budget
- what parts you have in your partsbox.

Note, for a SET amp 8 - 10W is about the limit before you start getting into high voltage power supplies and transmitting tubes.

Cheers
Brett

 

[fragman56]

parts and questions

I have a lot of parts--transformers, tubes, resistors, caps, just about everything!

Tubes, resistors, caps, wire are all going to work. Since I know even less about transformers than tubes, I don't know if the transformers will be useful. Perhaps I can test?

I have a complete electronics shop and lots of old electronics laboratory equipment (decommissioning and demolishing it all soon!) but I have never worked in electronics--just facility (research and development) management and environmental, health and safety.

So, early suggestions to me were: SET el84 (Planet 10). Got the tubes. I have very efficient speakers: altec lansing 605a (102 db/1w/m; 16R) or jbls control 28 (93 db; 8R) or others...low amp power is fine; 8 watts is plenty, thought el 84 put out 1-2 watts. 12ax7's I have.

Budget? I think I can handle it--not looking for fancy--education and experience the goal.

I did print out and read NC's part 1 and still not getting it...

Here's what level I am:

The AC signal (audio) is going + and -, thereby attracting electrons (increasing current to the anode) and repelling (decreasing current to the anode)?

AC signal can be + as long as it is more negative than cathode?

Is the AC signal always repelling the cathode to anode flow?

Then at the anode, it has a fluctuating DC that has to pass the coupling cap to the output tube? How? Caps stop DC? There must be an AC signal out of the gain tube?

Thank you for your response and patience,
Rick

 

[jeff mai]

OK, I'll take a shot at this, though I think others may be better able to explain it than me. I've only got a rudimentary understanding of things. Others please correct any mistakes!

>>I did print out and read NC's part 1 and still not getting it...<<

What specifically about it are you not "getting"?

>>The AC signal (audio) is going + and -, thereby attracting electrons (increasing current to the anode) and repelling (decreasing current to the anode)?<<

In a sense this is correct. Reality is that the grid is always repelling electrons in "normal" class A operation. If the grid goes more positive than the cathode you have class A2 operation. There are valves designed for this, but typical receiving valves (the sort most frequently used) are not. There are a few commonly used exceptions. The 211 is one.

>>AC signal can be + as long as it is more negative than cathode?<<

Yes, unless you want class A2 operation.

>>Is the AC signal always repelling the cathode to anode flow?<<

Yes, sort of. The AC signal on the grid does the repelling. See above.

>>Then at the anode, it has a fluctuating DC that has to pass the coupling cap to the output tube? How? Caps stop DC?<<

Yes, the cap blocks DC and allows the AC signal to pass. The cap must be of a certain minimum size to pass all audio frequencies. This depends on other circuit values.

>>There must be an AC signal out of the gain tube?<<

The fluctuating DC voltage is across the load resistor of the gain tube. The way I look at it is this: The tube is a variable resistor controlled by the grid - cathode voltage. This variable resistor forms a voltage divider with the load resistor. The power supply supplies a fixed (theoretically) voltage to the top of the voltage divider. When the grid is made more negative less current flows making the tube seem like a big value resistor. When the grid is made less negative more current flows making the tube seem like a small value resistor. If you think about the way a voltage divider works (the voltage drop across each resistor is proportional to the size of the resistor) you can see that the voltage across the fixed load resistor will vary with the input signal on the grid of the tube.

Hope that made at least a little sense!

Jeff

 

[fragman56]

Thank you for the reply, Jeff.

So, the audio signal (AC) that is input to the gain tube is amplified by the fluctuating DC passing from the cathode to the anode. Audio "rides" on top of the DC to the cap, where the DC is dropped. This amplified AC (audio) as measured in milliamps is input to the output tube as audio signal.

The process begins again, except this time the audio signal (expressed as amplified milliamps, along with some voltage value) is large enough to be transformed into a usable form to drive the speakers. Is this about right?

Generally, can you give me some example values of voltage and current at different points in the circuit? As below:

# Location Current (AC or DC) Voltage

1. audio signal at gain tube
2. gain tube cathode pin
3. gain tube anode pin
4. before coupling cap (between gain
and output tube)
5. after coupling cap (between gain
and output tube)
6. audion signal at output tube
7. output tube cathode pin
8 output tube anode pin
9. at the output transf. (amp side)
10. at the output trans. speakers side

As I look at this, if it's too hard, please don't. I'll try to measure on one of my PP tube amps. I'm sure it won't be the same as SET but may prove useful.

Quick question: the literature says that most transformers are single application constructed, that is, probably can't be used in another circuit. I have many transf. and wanted to know how you can tell (measure) which will be useful (PP or SE). SE's are big and expensive, I understand. What do you look for?

Thank you for your interest.

Aspiring diyer,
Rick

 

[PassFan]

Easy Understanding

Check out the EZIndex at gabevees site. He has a lot of design info in laymans terms with simple circuits for examples. It gets easier to look at smaller the picture and then put them together for the big circuit. Just do a search in this section for Gabevee and you should be able to get his site off one of his posts.

 

[Brett]

Thank you for the reply, Jeff.

Pretty good basic description of what's going on.

So, the audio signal (AC) that is input to the gain tube is amplified by the fluctuating DC passing from the cathode to the anode. Audio "rides" on top of the DC to the cap, where the DC is dropped. This amplified AC (audio) as measured in milliamps is input to the output tube as audio signal.

The audio signal is the difference in <i>voltage</i> applied between the grid and the cathode. For now, you can safely assume that in most cases zero current will flow in the grid, except in some circumstance in power tubes. Ignore this complication for the moment.

A varying voltage applied to the grid will cause the quiescent current to vary in proportion through the load impedance depending on the transconductance of the tube. Due to Ohms Law, the now varying current through the load resistor will give a varying voltage across the load, this being the amplified signal. The series cap between stages removes the DC component of the signal, and the voltage is applied to the grid of the power tube, and the process repeats.

The process begins again, except this time the audio signal (expressed as amplified milliamps, along with some voltage value) is large enough to be transformed into a usable form to drive the speakers. Is this about right?

As above, voltage in, current out across a load (and therefore a voltage). The transformer turns ratio will change the high voltage low current primary signal into a low voltage high current signal that speakers need. The turns ratio also determines the impedance seen by the power tube and therefore the voltage able to be generated across the primary.
Zp = Zs x (turns ratio)^2

Generally, can you give me some example values of voltage and current at different points in the circuit? As below:

# Location Current (AC or DC) Voltage

1. audio signal at gain tube
2. gain tube cathode pin
3. gain tube anode pin
4. before coupling cap (between gain
and output tube)
5. after coupling cap (between gain
and output tube)
6. audion signal at output tube
7. output tube cathode pin
8 output tube anode pin
9. at the output transf. (amp side)
10. at the output trans. speakers side

As I look at this, if it's too hard, please don't. I'll try to measure on one of my PP tube amps. I'm sure it won't be the same as SET but may prove useful.

I'm happy to do this, but I need a specific circuit to answer with any detail. Do you have any idea what you want to build yet? If not, I'll find one online and post it and we can work through the actual circuit.

Quick question: the literature says that most transformers are single application constructed, that is, probably can't be used in another circuit. I have many transf. and wanted to know how you can tell (measure) which will be useful (PP or SE). SE's are big and expensive, I understand. What do you look for?

A PP output trans from say a Dyna ST70 can be used in any number of circuits providing that the primary Z, power rating and current through the primaries 'works' for the new circuit.

The same goes for a SE OPT, however you generally can't use a PP OPT in a SE amp and vice versa.

A SE OPT only has one primary winding, whereas a PP amp has two (3 wires). In both types, the DC bias that flows through the transformer cases a magnetic flux in the core, that takes up some of the available flux for the signal. In a PP OPT, the two primary windings are wound opposite each other, so the flux should be opposite and cancel. You can't do this in a SE OPT, so a small airgap is included in the core, and the size of this gap will determine the standing bias for the output stage. SE OPTs are specified for a primary impedance, and quiescent current. For a 2A3 amp this might be 3000 ohms, and 60mA. At approximately double the quiescent, the transformer core will saturate, with a corresponding huge increase in distortion.

Cheers
Brett

 

[Brett]

Rick,

I had a HDD failure, and I can't follow up as promised for a few days. No data on hand, and I'm using a glacially slow borrowed laptop.

Cheers

 

[fragman56]

el 84?

Hello Brett!

Haven't forgot about you. I have been busy reading the "2A3 amp" and "my first valve amp" posts because of all the info contained there that may be applicable here.

I originally thought about the EL 84 and its famed sound, but the 2A3 and the 300B (equally famous) might be good. Objective remains: learning and building takes priority. Simple is best for me.

One person suggested that the selection of tube will probably be based on the iron (tranny) available. Question: can I choose an output tranny that would serve equally well for any of the three tubes listed above? I will probably want to use the tranny for later projects. For example, build the EL 84 first then the 300B using tranny?

Just a question,
Rick

PS: Still sorting through some schematics--would prefer to do like mig-ru, that is, complete design from blank page

 

[Brett]

Re: el 84?

Hello Brett!
One person suggested that the selection of tube will probably be based on the iron (tranny) available. Question: can I choose an output tranny that would serve equally well for any of the three tubes listed above? I will probably want to use the tranny for later projects. For example, build the EL 84 first then the 300B using tranny?

PS: Still sorting through some schematics--would prefer to do like mig-ru, that is, complete design from blank page

Hi Rick,

The impedance range that the tubes like to work into is similar enough that at a quick glance they might work. but all 3 tubes run at different quiescent currents, which means a different gap in the transformer, and so different primary inductance. The 300B and 2a3 have an Rp of about 700R and the EL84 (trioded) is about 1700R, and bias currents would probably be 40/60/80mA fpr the EL84/2A3/300B. So putting an EL84 onto an OPT optimised for a 300B would give less low end due to (probably) lower primary L and higher Rp, and the reverse would have the 300B sturating the EL84 trans at idle.

I'm happy to work with you with any of these, but you need to decide how much power you want and $ you want to spend. Even though you want it largely for the experience, might as well build something you can use, no? Power outputs are likely to be 1.5-2W/4W/8W in SE. It's much more interesting to build something from scratch too, you'll learn a lot more in the process.

Because it might take me until the weekend to get up to speed again, here's some reading for you to do.
http://www.aloha-audio.com/library/index.html
Read articles 2 through 7, especially the Amity/Aurora/Raven one, and the others if you're interested.
Also grab the 300B PDF datasheet from Western Electric and take a look. It has a couple of operating points for the 300B there which are a good starting point. There's also an excellent one for the EL84/6BQ5 at www.retrovox.com.au in data section. Both of these also have some distortion vs power and/or load resistance. Even if you don't understand a lot of what's in these articles and datsheets, some of it will still sink in. All of them are excellent.

I'll refer you to some online pages about picking loadlines etc later.

Cheers
Brett

 

[fragman56]

300B or el84

I agree with you that it doesn't hurt to have something at the end.

Budget? Perhaps keep it under $500 US??? Probably must purchase SE tranny(s) since they are so unique(?) I don't have 300B tubes but do have 84s, but it really doesn't matter. Power doesn't matter either. I assume both are equally simple to design around.

I read the references--excellent stuff--maybe a little advanced for me now.

I have an article by Crowhurst on loadlines. Still interpreting it.

Rick

 

[Brett]

Rick,

If you have a suitable power tranny, the a 300B stereo SE amp could be done for around $500. Tubes are $US140/pr and up, and transformers are $US200 +/pr. Some caps, PSU chokes and passives, plus a DIY chassis would take the rest. This wouldn't make a SOTA 300B amp, but a good one, and one that if you like the flavour of the sound the 300Bs and SE produce, could be improved over time as you had cash spare to tackle it. 300Bs might also be able to be borrowed for a short time from other DIYers to try different brands and plate constructions, which all give a different tone.

For the PSU the 300B would need a B+ of 370-450V at about 110 -120mA/ch, and the EL84 approx 250V B+ at 60-80mA/ch.

The EL84's would make excellent 300B driver tubes too, either triode or pentode. If your preamp has a low output Z, decent drive abilities and can swing about 20Vp-p, you won't even need a driver for the EL84. You could make a spud amp (a 1 tuber).

The only saving over the 300B to make an EL84 would be the tubes. PSU (chokes, caps, rectifiers), OPTs and passives are still required and will cost about the same for both. F'rinstance the interstage coupling cap between driver and power tube could be $10 ea or $80 ea. The $10 part will probably give 90% of the performance.

<b>I assume both are equally simple to design around.</b>

In theory yes, but in practice the 300B circuit will require a bit more detail in execution. The 300B is a direct heated tube, so the heater and the cathode are the same thing. Any noise on the cathodes of the 300B will be amplified because the tube won't be able to discriminate between the signal and noise. The EL84 is indirectly heated, so the heater and cathode are seperate and there is a degree of isolation between them. You'll get better sound (tone), more power and probably learn more with the 300B, but it might not work as well upon first power up. Some stuffing around will be required to optimise it. EL84's are probably more ídiot-proof''

Glad you looked a Lynn's articles. He is a gifted engineering writer and has a very intuitive feel for what's happenning in circuits. Keep a note of the URL, and look back on them later as you learn more, and more will become clear. Crowhurst is the master of this: in one sentence can be a ton of wisdom. You might only get the basics now, but more becomes shown in the same articles, as your understanding grows. Crowhurst has 6 articles in GA, and on the audioxpress site, the first two are most useful here.

Cheers
Brett

www.tubesandmore.com
www.welbornelabs.com
www.triodeelectronics.com
www.hndme.com
will give you an idea of parts prices, esp the first one, though Triode might be cheaper for the tubes.

 

[fragman56]

el 84 gain with 300b output--

does that sound ok? Do you need to get a matched pair or can you simply order 2 tubes? I have read that the JJ 300b is very good and worth the money but matched pair not available until Nov 7. Singles available now.

I have a Heathkit preamp (tubed) but I have not tried to get it up and running yet. I was not counting on the need for a preamp. One tuber you say? Another opportunity to learn something?

Should I start the design at the gain tube by drawing loadlines, etc? Or, get a proven schematic and review design, check adequacy, etc? Or, start at the speaker end and work backwards?

Your input greatly appreciated,
Rick

 

[Brett]

Re: el 84 gain with 300b output--

does that sound ok? Do you need to get a matched pair or can you simply order 2 tubes? I have read that the JJ 300b is very good and worth the money but matched pair not available until Nov 7. Singles available now.

I reckon it's gonna sound very nice. I like the EL84 a lot as a tube, and the few 300Bs I've heard have been good too. Best thing about the EL84 as a driver is that if you find it doesn't have enough gain in triode, you can easily pentode it and get the gain, and <i>reputedly</i> better sonics in some applications. Because it will be a simple circuit, all the components potentially can be tweaked to give exactly the tonal palette you want, and because there are so few, it's cheap to do. Relatively. Adjusting the bias of the 300B will also change it's tone. Lots to play with.

As for matched tubes, it probably won't matter that much in a SET amp, but it's nice to have. Email Ned at Triode and ask if he can match them. Alternatively Antique Electronics (www.tubesandmore.com) sell them and will match too.

I have a Heathkit preamp (tubed) but I have not tried to get it up and running yet. I was not counting on the need for a preamp. One tuber you say? Another opportunity to learn something?

The Heath pre, if it's like all the other older circuits I've seen will have tons of gain and should swing the 4V required to drive an EL84 (t) -> 300B combo.
Is it 12__7 based? Do you have a model # or schematic? True, tweaking it will be another opportunity to learn about different types of circuits, but still apply similar techniques that you'll learn in the poweramp.

Should I start the design at the gain tube by drawing loadlines, etc? Or, get a proven schematic and review design, check adequacy, etc? Or, start at the speaker end and work backwards?

Start at the output and work back. If you don't already have it, get the WE 300B datasheet from http://www.westernelectric.com/spec_sheets/300B.pdf
At the bottom of page 5 you'll see a chart that shows the relationship between load impedance (transformer primary Z), distortion and power for a given bias. The same basic relationships hold for most tubes. At at Z=3000R, Pout=8W. 2500R - 3500R seems to be the most common range of load impedance for SE 300Bs, so we know it works.

For the excercise, lets use the WE parameters. They designed the tube and probably had the brightest collection of tube engineers and knowledge ever, so I assume they knew what they're doing.

Vp-k (Eb) = 350V
Vg (Ec) = -74V so Vp = 424V
Ip = 60mA
Pd = 21W

With a dissipation of 21W (36W rating) it's a conservative op point, and will give long life. Anecdotally the best sound/tube life compromise seems to be at 80% max, or 28.8W, so there is still some room to get a bit more power. Many good amps use a similar op point.

Plot the loadline if you like on the WE graph on the top graph on p3. Play with some different points until you get a 'feel' for what's happening.
Steve Bench's page at http://members.aol.com/sbench102/po-dis.html gives more loadline data.

I have to go to a callout. More later.

Cheers
Brett

 

[fdegrove]

300B

Brett,

Another marvelous job you did here.
You get a standing ovation and a BIG thumbs up from me!

Has anyone ever thought along these lines:
Why not build dual mono integrated SE amps?
You would end up with what I would consider to be the shortest possible signal path.
Assuming you're not using any analogue stuff, you would have say one stage for digital amplification/buffering/source selection and volume control going directly into the one tube output stage .
Longish speaker cables aren't really a problem,so?


Cheers,