Mono 5W-17W Power Amp using EL84 push-pull

[nbraz]

Hi everyone,

I'm trying to build a single channel power amp using two EL84 in a push-pull setup, with a switch to select between 5W (triode) and 17W (pentode). The idea started in a discussion with a friend, as he needed a small and portable amp to run his Line6 HD500 (modelling unit). I've always wanted to build a tube amp and so I thought this could be a nice and "simple" project to start with.

As a baseline I'm referring to two designs:
1 - 18W stereo power amp http://www.el34world.com/projects/images/stereopowerampschematic.gif
2 - Orange Tiny Terror http://www.prowessamplifiers.com/schematics/images/tiny_terror.pdf

The Line6 HD500 has balanced XLR outputs (+4dBm 1.23V), which we intend to use with this amp (instead of the usual 1/4'' unbalanced connections). The first question I would like to start with is:

Is there any need for a pre-amp at the input stage in this setup? My friend tried his HD500 with a Hot Rod Deluxe connected to the power amp in (so, no equalization stage) and he complained about the volume being to high (possible clipping). I found some schematics for a similar amp (http://www.blueguitar.org/new/schem/fender/blues_deville.gif) and I've noticed that they use a TL072 op-amp (V1B) at the power amp in stage, which I believe is to bring the voltage up before the phase splitter and also, isolate the internal circuit of the external loop effects. This makes me wonder about the usefulness of a pre-amp stage at the input.

If you think the pre-amp is necessary in this project, what gain would you recommend for the pre-amp and what voltages should I consider at the input of the phase splitter?

Any help with this would be very much appreciated!

Cheers,

Nuno

 

[walter807]

hi, I built this amp and it sounds great, just add the switch to change mode pentode ultra lneal to triode, in ul mode thd is very low at 10W, aprox 0.5%, I hope to serve you, regards
Walter

 

[nbraz]

Hi Walter,

Thank you for your quick reply! The circuit you shown me looks great!

So, do you reckon I don't need an input pre-amp stage on this circuit?

Thanks,

Nuno

 

[Fredenando]

Maybe this can help you too:

Mesa Boogie 20/20.pdf

 

[Wavebourn]

hi, I built this amp and it sounds great, just add the switch to change mode pentode ultra lneal to triode, in ul mode thd is very low at 10W, aprox 0.5%, I hope to serve you, regards
Walter

View attachment 228684

What is the point in bootstrap there?

 

[nbraz]

Hi Fredenando,

Thanks for the circuit! The details on the voltage values at each stage is quite useful!

Nuno

 

[walter807]

Nuno, the amp built for me no need preamplifier for use with cd player standar, this work well but i dont now output specification of the hd 500 line6, for this reazon looks is another circuit, maybe this is what you want, is more likely to adapt the input signal to your liking, regars

Walter

 

[VictoriaGuy]

Nuno-
With all the amp modelling in the Line6Pod, my guess is that a clean-sounding amplifier would work better than adding 'more tube' sound. That said, I'm not sure something very 'hi-fi' is necessary.

Yes, you are correct that you don't need the preamp gain stages of a regular guitar amp if you have >1volt at the input.

I'm not sure about the advantages in using the XLR outs from the Pod (unless going direct into a mixer/board). Also are you planning on tying the R&L channels together at the input?

This amp is for gigging? (18watt amps are really LOUD..)

What sort of speaker are you going to use? (Important).

Cheers
John

 

[dcgillespie]

What is the point in bootstrap there?

Ultimately, it raises OLG to lower the input drive requirements. Without the PFB included between the PI and input stages, OLG will be reduced, so that maintaining a given level of global NFB will then require a greater input level to produce a given output level. Since this appears to be a commercial design, the PFB was likely included to help meet an input sensitivity specification design point.

Dave

 

[nbraz]

Hi everyone,

Thank you all for your input into this project. Is very much appreciated!

Walter:
Thanks for the schematic. I'm trying to guide myself through specific design rules now, so I can actually understand the differences between amps... I have a rough idea of how the amp is going to look like, and I've been finding the Mullard Valve Amplifier Book quite useful for this. Do you know of any good resources it could be also useful?

John:

With all the amp modelling in the Line6Pod, my guess is that a clean-sounding amplifier would work better than adding 'more tube' sound. That said, I'm not sure something very 'hi-fi' is necessary.

I agree with you. My motivation for this project is more as an exercise to learn about tube technology, which I've always wanted. On the end is just a simple power-amp (no equalization stage, etc...), so I'm hoping to get something that sounds reasonably nice and with good sound projection and body, which I found difficult to find in a solid state amp :/

Yes, you are correct that you don't need the preamp gain stages of a regular guitar amp if you have >1volt at the input.

Thank you for this One question then, why power-amps like Mesa 20/20 or Marshall EL84-20/20 have pre-amp stages at the input, given that they are used with similar modelling units (e.g. Marshall JMP-1 or Mesa triaxis)? Is it because of the regular 1/4 connections?

I'm not sure about the advantages in using the XLR outs from the Pod (unless going direct into a mixer/board).

We have the option for balanced XLR connections, which are mechanically more robust and offer the possibility of reducing noise (created in the wires). We could use a normal connection but since we have the option, why not XLR?

Also are you planning on tying the R&L channels together at the input?

My friend (which the amp is for) gives priority to portability, so we're aiming to build just a mono amp.

This amp is for gigging? (18watt amps are really LOUD..)

Good Hopefully, if the project goes forward, the amp will be used in gigs and studio.

What sort of speaker are you going to use?

Is going to be a 16ohm Celestion Greenback or Vintage 30.

Thanks,

Nuno

 

[costis_n]

XLR connection

Actually, this XLR connection is a godsend, you don't even need a phase splitter! Pins 2 and 3 (they are marked on every cable) provide you with anti-phase outputs of the same signal. So it has already been split for you. All you need to do is pick a design that has a Schmitt-type "differential" stage before the EL84's and connect pin2 and pin3 to the grid of the left and the right triode, respectively (through coupling caps). You could also drive directly the power tubes, they are sensitive enough for that, but you won't get powerstage overdrive this way.

 

[VictoriaGuy]

Actually, this XLR connection is a godsend, you don't even need a phase splitter! Pins 2 and 3 (they are marked on every cable) provide you with anti-phase outputs of the same signal. So it has already been split for you. All you need to do is pick a design that has a Schmitt-type "differential" stage before the EL84's and connect pin2 and pin3 to the grid of the left and the right triode, respectively (through coupling caps). You could also drive directly the power tubes, they are sensitive enough for that, but you won't get powerstage overdrive this way.

True.
The Line6 must have a 'phase splitter' in its circuitry to 'create' the balanced signal from the guitar input (unbalanced in 99.9% of electric guitars). Might as well use that if it makes getting the amp design easier.....though I'm not sure it will, compared to an 'off-the-shelf- circuit.
With a modelling set-up, all the overdrive you need is simulated digitally.
I'm pretty 'old school' on this- my inclination is to either do the 'real thing' with tube amp (or a collection of tube amps) or use a modelling setup with solid state gear.

It does seem to be a lot of work to build a 'single purpose' tube amp for this application, when a SS amp/mixer solution combined with a low-power-rated Celestion would probably get a similar result, at a much lower cost. But, it's a hobby.....
;-)

BTW, if the 'low power' 5W option is for practicing alone, the 'phones jack on the Line6 should be kept in mind. (5W is way too loud for not disturbing others in the house/flat, though fine for rehearsal, IMO).

EDIT: I just had a quick look at the PodHD manual, and I notice that there's a switch (for the unbalanced output) that sets the output level to match up with a 'standard' guitar amp. Perhaps the Pod owner who complained about the 'too loud' situation may have missed this, or is not using the 'Master' volume control?

 

[VictoriaGuy]

:/


Thank you for this One question then, why power-amps like Mesa 20/20 or Marshall EL84-20/20 have pre-amp stages at the input, given that they are used with similar modelling units (e.g. Marshall JMP-1 or Mesa triaxis)? Is it because of the regular 1/4 connections?


We have the option for balanced XLR connections, which are mechanically more robust and offer the possibility of reducing noise (created in the wires). We could use a normal connection but since we have the option, why not XLR?

I don't see a line level input on the Marshall 20/20; perhaps I'm missing something. Again, adjusting the modelling unit output to suit the amp may be the key.

Since the PodHD is going to be close to the amp (you didn't mention the floor 'stomp box' version) I don't think you will have cable noise and mechanical problems with either XLR or 1/4" connections, as long as you avoid poor-quality connectors. The Pod has a noise gate (??) so that can take care of low level noise when you aren't playing, I'd think.

Of course, in some studio environments, you may be plugging straight from the Pod into the board, without using an amp at all, so XLR and line level will be the choice there.

 

[nbraz]

Hi everyone,

First of all I would like to apologise for my late reply. You guys have been giving me some really good feedback and I needed sometime on my own just to get things a bit more clear.

Actually, this XLR connection is a godsend, you don't even need a phase splitter! Pins 2 and 3 (they are marked on every cable) provide you with anti-phase outputs of the same signal.

costis_n many thanks for this! Sometimes you get so focus on a problem that you forget the simplest solutions... On my design (attached) I've used the Texas Instruments THS4140 Fully-Differential Amplifier with Gain 20- for a 1.23Vrms signal it will give 20*0.5(differential)*1.23=12.3Vrms. The driving voltage on the EL84's will be 10Vrms and so 12.3Vrms should be enough?

I'm pretty 'old school' on this- my inclination is to either do the 'real thing' with tube amp (or a collection of tube amps) or use a modelling setup with solid state gear.
It does seem to be a lot of work to build a 'single purpose' tube amp for this application, when a SS amp/mixer solution combined with a low-power-rated Celestion would probably get a similar result, at a much lower cost. But, it's a hobby.....

John, I understand your view and for this reason I made some experiments with a digital guitar processor on a tube and solid state amplifiers. In this experiment I found a really interesting product Line 6 DT50, which basically consists of a tube pre-amp (2x12AX7) -> digital amp modulation -> tube power amp (2xEL34) in one box. I was impressed- it sounded good (it had more character than a SS) and it was really versatile (relays to switch between Class A and ClassAB; Pentode and Triode; and 4 different negative feedback loops)! You'll see that on my design I've incorporated some of these features, but instead of digital controls everything is analogue:

- Switch SW2 for Class A or Class AB;
- Switches SW1 and SW3 are used to switch between Triode, Ultra-Linear and Pentode. Ideally, I would have wanted to use a 3P3T switch, but this are difficult to find in the market for the voltage ratings I'm using...

I also wanted to add some options for negative feedback, but I'm leaving that for a later date as from what I've read this is more a audiophile exercise than a technical one.

I would be very grateful if you could give me any feedback on the circuit. I'm sure I've made some silly mistakes, so if you spot any major problems (e.g. grounding, power supply, etc...) please do let me know! I'm also open to any suggestions or contributions into the project you may have.

Many thanks,

Nuno

 

[Ian444]

I believe, from 20 odd years playing guitar, that it is a good idea to have at least one tube block in there somewhere whether it be the preamp or the poweramp. nbraz go for it!

Edit: Forgot to mention, if you are going to use an opamp to drive the tubes, I have found that a very clean DC supply is needed for the opamp, a gyrator and carefully implemented LM317 with dedicated transformer will get you there, anything less and you may as well use a tube instead of that opamp, except you don't need the gain I presume.

 

[Ian444]

Edit 2: R14,15,16,17 look way too big. Check the datsheet for EL84 for R14,15. 20K is a large but possible value for R16,17.

 

[costis_n]

Ahh, why the op-amp? Can’t you use an ECC81 in there? It is simple to do so, and then you won't need the dual 15V suply either.

 

[nbraz]

Hey guys,

Thank you for your feedback!

I believe, from 20 odd years playing guitar, that it is a good idea to have at least one tube block in there somewhere whether it be the preamp or the poweramp.

Cool! It's always good know that other people share the same opinion

if you are going to use an opamp to drive the tubes, I have found that a very clean DC supply is needed for the opamp, a gyrator and carefully implemented LM317 with dedicated transformer will get you there

That is good advice. I might have to think a bit more on this one then. I'm not totally happy if I've to have an extra transformer though... You mention using tubes; if I still wanted to use the balanced XLR connection, how would I go about to do that in a way I can still remove the common-mode noise?

Edit 2: R14,15,16,17 look way too big. Check the datsheet for EL84 for R14,15. 20K is a large but possible value for R16,17.

I made the wrong assumption that the datasheet values for the Grid 1 resistor where the internal values of the tube, but clearly it isn't. Thanks for noticing the mistake! I need to readjust this and maybe add an extra switch to have different values of R14/15 for each class...

Ahh, why the op-amp? Can’t you use an ECC81 in there? It is simple to do so, and then you won't need the dual 15V suply either.

The op-amp is a differential amplifier and it was an attempt to avoid using a phase splitter by using the inverted signal on the balanced XLR connection (based on your previous advice).

Nuno

 

[Ian444]

This circuit here from Pete Millett's site shows two differential amplifiers after a transformer that does the phase splitting. I just posted this as an example to give you an idea of what costis_n means by a "Schmitt-type "differential" stage before the EL84's". That would be a really good way to do it.

 

[nbraz]

Ok, thank you for your reply. Do you know of any guide for dummies which explains this Schmitt-type differential stage in a bit more detail? I'm struggling a bit understanding how this thing works.

Thanks,

Nuno