Building a Valve Amplifier Part 1: Design, Components and Layout
Building a brand new valve amplifier using the RH84 design and off-the-shelf components.
A number of recent visitors have registered surprise that we do not have any music playing in our workshop. We do have a small radio, but it seldom gets turned on. I reckoned it was about time to put that right. I have always been keen on valve or “tube” audio amplifiers, both for HiFi and musical instruments. I’ve owned various valve guitar amps in my youth and fondly remember the glow of the radiogram in my childhood home.
Valves were largely superseded by solid-state systems from the mid-1960s onwards, but have made a real comeback in the past decade or so – particularly amongst audiophiles – as people realise they produce a warmer sound, as well as having great visual appeal. These days any kind of valve amp attracts a premium price and I thought that building one for myself, daunting prospect though it was, would be a relatively economical route to take.
After doing some research it seemed the perfect beginner's project was the RH84 single-ended amplifier. It was created by Aleksandar Kitic, who has said that “The amp was really designed for the first time builder, to provide results even with salvaged parts or the cheapest possible. At the time I designed it I was practically penniless...”
In this case, we are going to use readily available “off-the-shelf” parts, rather than salvaged ones, or indeed some of the specialist parts referred to in many of the online forums discussing the building of valve amps.
The components for a single mono amplifier circuit are:
- 1 x 3W Single Ended Output Transformer (123-7242)
- 1 x ECC81 Double Triode Valve (678-4101)
- 1 x EL84 Pentode Valve (678-4120)
- 2 x Valve base (B9A0 (678-4094)
- 1 x 20V zener Diode (654-7628)
- 1 x 22k, 2W Carbon Resistor (707-8921)
- 1 x 100k, 22W Carbon Resistor (707-8940)
- 1 x 240R, 0.6W Metal Film Resistor (014-8354)
- 1 x 1M, 0.25W Carbon Resistor (707-7903)
- 1 x 470k, 2W Carbon Resistor (707-8974)
- 1 x 27R, 2W Carbon Resistor (707-8801)
- 1 x 10k, 2W Carbon Resistor (707-8906)
- 1 x 10uF, 400V, PEG124 axial A1 Electrolytic capacitor (226-7182)
- 1 x 100uF, 50V, A1 Electrolytic Capacitor (839-6261)
- 1 x 220nF, 1kV Axial Polyprop Capacitor (011-4610)
- 1 x 1.2V to 37V, 1.5A Standard Regulator (714-0792)
There are also two transformers for the power supply: one 6.3v (050-4561) for the valve filaments and a toroidal transformer 230v transformer (117-6060) for the HT (high tension/voltage). In addition there will be a bridge rectifier and smoothing components in the power supply, and this will be covered in a future post. For now we’ll just deal with the amplifier circuit.
As well as reading online blogs and forum posts about the amplifier, and other DIY tube amp projects, I referred to “Building Tube Amplifiers” by Morgan Jones, particularly with reference to the layout of the components. Morgan emphasises that the layout should be aesthetically pleasing as well as practical, and the book has lots of advice on how to arrange components to minimise the risk of hum due to induction from transformers into each other and the valves – “This hum is generated by oscillating electric currents induced in sensitive (high gain or high impedance) audio circuitry by the alternating electromagnetic fields emanating from nearby mains-powered devices like power transformers” (Wikipedia).
Where he seemed lacking in information was in the use of a toroidal transformer, which is what I had, although on reading up about them it would seem they have a very low or negligible stray field.
Placing the valves so that they keep as cool as possible and orientated so the “heater” wire takes the shortest route was another consideration.
I imported the images for the transformers and valve bases from PDF data sheets into Inkscape, drew some of the other components and experimented with different arrangements until I was happy. I then shared my design with Alex Kitic via email, and he seemed to think the layout was OK, which was good news. I was now happy that I had a viable design at least for the audio stage of the amplifier. We would be using a pair of bench power supplies for testing, so the power supply could come later.
Having used the Fostex FE83En full range drivers in my project to build a Portable 12V Sound System we knew they performed well with relatively small amplifiers and so decided to use them again. In fact, we decided to order cabinets from the same supplier and paint them with the same “Tuff Cab” coating. The one disadvantage with this coating is that it only seems to be available in black, red or white unless you are prepared to wait for weeks. We got in touch with our Goth side and went for the black.
Given our previous experience and the fact that these were passive speakers and therefore a lot simpler, these would be comparatively simple to put together.
As is often the case with these blog posts I have embarked upon something completely new to me and somewhat out of my comfort zone. It is going well so far, but there is some way to go yet.
In future posts, I will be sharing the building of the prototype single amplifier, followed by the final version with two of these and a power supply. Finally, fitting this all into an attractive enclosure, with cabling and connectors etc. and, most importantly, how it sounds.
CommentsAdd a comment
[Comment was deleted]
How do you get the 325v dc with this 2x 115v transformer? The circuit asks for a 2x 300v trtansformer.
@Jdev99 230 volts (2 x 115v) is the RMS voltage on the transformer. But the peak voltage of an AC sign wave is the RMS voltage times the square root of two. 230v x 1.4142 = 325.266v
Hi Dave, I'm a bit confused about how you would build the power transformer with the two transformers you list. If you loom at the schematics for the classic CLC power supply on the RH84rev2 page the it would need a 300V-0-300V/120mA, 5V/3A and 6.3V/2.5A. The 117-6060 seems to be a isolating transformer and can't provided you with the 300V-0-300V.
Something like this seems ideal for this purpose: https://www.edcorusa.com/xpwr172
Awesome project. Even more awesome is that these parts are available from RS.
A couple of things to note about the toroid HT transformer:
1. Mains type toroids can have a primary to secondary coupling up to 20 to 30 KHz, so are effectively full audio bandwidth for any noise coming in on the mains supply.
This is compared to a standard EI laminate transformer which has the primary and secondaries in separate sections on the bobbin - coupling there should be 1 KHz at best (worst?).
2. Toroids can also produce an audible hum if the mains sine wave is distorted, e;g; by computer switch mode power supplies. Check the mains supply with a DVM on it’s DC setting if it has a peak function.
By comparison the EI lamination layup gives a nominal gap so the transformer is less susceptible to small amounts of DC in the supply.
The HT should be applied ONLY once the valve heaters are warmed up. I would suggest a delayed relay and 555 (or other) timer for a few seconds turn on. A zero switching SCR could be used, although that might be adding too much “sand” to the project. A mechanical relay (with snubbing across the contacts) gives a nice satisfying clunk :)
Single ended amplifiers need well smoothed HT supplies, otherwise the ripple will modulate the signal.
Hope this helps.
At the risk of self-promotion, I have a 555-based circuit here http://valve-audio.blogspot.com/2018/11/a-better-time-delay-startup-circuit.html passed on for interest only. Note the 100µF cap after the rectifier - do not reduce this to 68 or 47 µF if that's all you have - circuit doesn't work properly.
Hi Dave, This looks like a great project, do you have any idea of what the finished project might cost? ( a general idea of what you mean by "an economical route to take" ) Thx in Advance. Greg