All about electronic valves and hi-fi
Browsing my schematic archive I found this early implementation from some years ago on the 300B:
The amp design is straight forward. Let’s start from the output stage. The 300B is run hot at 33W (376V/90mA) with a fixed bias of about -78V. I used a pair of LL1623/90mA OPTs which I had wired on 3K:8 mode.
I devoted the majority of my audio life to single-ended DHT amplifiers. I don’t regret the path, however, I do not want a polarised perspective on amplifiers. I do like very much a good push-pull amp, though. Over time, I stumble across poor implementations or excessive use of gNFB. Local feedback (lFB) designs well implemented are music to my ears.
I played around with SPUD amps over the years, with happy results. One of them I loved over the breadboard is the 6e5p in push-pull. I will definitely build a definitive version of it. Here is the concept overall:
The input transformers is step up. It will need 1:4+4 ideally. My friend Dorin from DVB Transformers has a great device which I reviewed before and would be ideal for this amplifier. To simplify the power supplies, I used only one, which provides a bit of a strain into the source followers SF1 and SF2. If you have a good driver, you can omit them, I wouldn’t. They take care of the miller capacitance which is worsened by the step up transformer. It’s just a pair of PCBs, not big deal. RK1 is an arrangement with a pot to balance the bias. C1 is in ultrapath. I used an Oil 40uF one. T1 was a Lundahl 8K8:8Ω PP OPT.
The CCS provides a stable bias current into RBIAS1 which sets the bias voltage.
Here is the operating point I used:
The 6e5p is biased at about 40-43mA and 240V. It needs 5V bias so a cathode resistor of 58Ω should do. For 4W output in A1, you need 4Vrms, which only with a 1:2+2 step up should do. You can get about 5W in A2 or a tad more.
I loved the sound of the 6e5p as output stage (as much as I do as driver).
There are multiple variances to explore indeed
I have a pair of Toroidy PP OPTs with CFB, so would be interested in running them in tetrode instead.
With a quick mod on one of my active preamps, I managed to implement very quickly the 6Э6П-ДР (6e6p-dr) driver in my system. Here is how the circuit looks like:
I’m a firm believer that if you don’t share, you don’t get back and learn. What the point of not sharing what you’ve learned? I asked myself the question again yesterday, just to push me a bit further. Joys of Easter break is that I have the time to sit down and write. At least for a little bit.
Here is the result of my quest of the years to find the best drivers for a SE amplifier. I’d been looking and experimenting with them in terms of best linearity at large volt swings (I mean large when I say 200Vpp), harmonic profile and most importantly the sound contribution.
Why should you bother? Well if you are in the DHT space (otherwise don’t bother reading further) and, unless you are building a 4P1L amplifier, the majority of the output valves require large volt swing. You also need good headroom. Therefore if the driver is clumsy, it will ruin your expensive project. Again, one of the reasons why people claim that their 300B sound bad. Achieving a driver which can perform 200Vpp effectively with minimal distortion and a decaying harmonic content isn’t a simple task.
In one of my recent post, I blogged an example of the GM70 amplifier. Look at the curves below and the demand to get all of the juice:
Yes, you can load it with a steeper load and use a 6KΩ instead of 10KΩ to get more power, but you still need the same volts to get the full swing.
I tried it (mostly) all over the years. Transformer coupled, choke loaded, resistor loaded. However, in my experience the best is the gyrator load. You may have a different view, and so you may: well, it’s a free world and I’m not expecting you to agree with me. If you are prepared to accept my point for view, then you can continue reading this post 🙂
The hybrid mu-follower (aka gyrator load) is a very effective topology for a driver. You need sufficient volts at the supply, but that’s not generally a problem. You will need at least 25-50V more than the largest voltage swing. Most of the valves I will review below have a good compromise operating point at about 200V. For a 200Vpp or 250Vpp headroom, this means you need 200V+250V/2+50V = 375V. MOSFETs can work at this level and providing you put them the right heatsink size we’re on business!
Initially some years ago I explored the use of LEDs, diodes and particularly SiC diodes to bias the valves. However, I found later that a bit of cathode degeneration by placing and (unbypassed) resistor was a good choice. This linearise the valve a bit and won’t impact the output impedance of the driver. However, if the resistor is within a reasonable value (smaller than 300R I found in practice), the impact on the Frequency Response (FR) is manageable and also the reduction in gain of the stage.
I will present in this post my favourite contenders for the best drivers. These are:
In all cases I found the sweet spot with fixed bias which allows me to dial-in the right operating point in conjunction with the gyrator setting point. Once the best performance was measured (and listened) it was replaced with an equivalent resistor and re-tested. A tedious job, but worth the efforts.
These valves have mostly high gm and gain. You’ve been warned. Don’t even attempt to build with them without special attention on the building aspects. It will oscillate, believe me. You should add grid, anode and screen stoppers. I prefer nice ferrite beads added straight to the socket pins. Continue reading “Driving hard (Part I)”
It was mentioned here, so I decided to chip in with my previous experience from back in 2013. Here is what I tried myself:
The main challenge when implementing valve amplifiers using transmitting valves or valves which require a significant voltage swing (e.g. 300B, 45, etc.) is the driver. Getting the driver right is not easy. You’re asking for a single stage to swing 150 to 200Vpp at minimum distortion. There are some ways you can achieve this:
You should start by reading this extensive blog post. That will provide you with a lot of information around the shunt cascode and how it works. Back in 2013 I started playing with the 6Э5П in this topology. It was quite promising. Now, I have revisited and built this driver to see how it really performed.
The design is very similar to what we discussed back then. I shall proceed in describing the circuit, in particular the changes made. The driver is still the marvellous 6Э5П. There are few valves out there that I don’t like as much as I do with the 6Э5П. I measured the curves long time ago when I started with the curve tracer project. I also tested the 6Э5П and 6Э6П extensively. I do love the 6Э6П as well, it’s one of my favourite drivers.
The 6Э5П is biased at about 200V/30mA with a degeneration cathode resistor of 120Ω. As the gain of this stage isn’t dependent on the μ of the valve, then is good to do this to improve the linearity of the driver. M2 forms a CCS with Rmu. It provides the current to the 6Э5П as well as the current to the common base stage formed by Q1 and Q2. The gain of this stage is gm times R5. The gm is the valve’s transconductance The collector current of the MPSA92 is kept low to ensure distortion is minimised as well as its operated under SOA. D3 provides a protection to the darlington pair when is reversed biased.
The gain of this stage was measured to be x140 (or 43dB). That equals to a degenerated transconductance of 5mA/V with a cathode resistor of 120Ω and a gain resistor for 27kΩ.
Well, it was obvious I couldn’t leave last post as it was. There is an option to change the driver for a different valve. You can use a C3m (low gain in triode mode which is ideal here), a C3g, E180F/E280F. 6S45P or my loved 6E5P (or 6E6P) as the driver. Not longer a 100% DHT, but a nice option for sure. The 6E5P is extremely linear, good driver, with a nice gain (μ=30) in triode – perhaps more than enough for a 4P1L stage and would help in avoiding additional filament supplies.
The 6E5P has curves not dissimilar to the 4P1L as no further distortion cancelation can be seen. Here is the updated schematic if you’re interested in playing with:
Again, the gyrator PCB can be easily used to simplify the build of this amp. The 6E5P is not driven hard, but at a nice current of 20mA which makes the driver operate in a linear region (and with good sound) with just a pair of red LEDs. The nearly 30dB of gain will make this amp to be very sensitive. The 5W can be easily achieved with 1Vpp, so you will need to have an attenuator, no preamp needed clearly. The 6E5P will drive an 300B nicely here which needs the voltage gain, not like the 4P1L.
As you can see, there are plenty of option to try on this 4P1L PSE amplifier.
Playing this afternoon with the SiC C3D02060F, which can happily run +20mA with very low dynamic resistance. Ideal for the 6e5p/6e6p driver I had in mind for the 300B amp:
At 20mA of cathode current the forward voltage is 0.85V and dynamic resistance 1.5Ω. If cathode current is 40mA instead the resistance drops down to 1Ω:
The 6e5P/6e6P will run comfortably around 30-40mA and bias tends to be around 3.5-3.7V to swing nice volts as needed. Therefore, we will need 4 SiC in series. 4Ω resistance is good enough and not adding much when reflected to the anode…
Thanks to the great work that Derk put into his great Extract Model tool, I helped him to refine and debug the application by tracing the challenging 6e5p beam tetrode. After 4 versions we managed to optimise the model:
The model fits really well including the kink section but given the saturation of the 6e5p tetrode at currents above 50mA there is a slight divergence as the valve cannot reach the same anode current at higher voltages.
After some proper time tracing the curves, here are a set of Russian high-frequency pentodes in triode-mode for comparison. I was looking at closer ones to D3a and these are the ones I had at hand and wanted to include in the tests:
| 6Z5P | 6Z9P | 6Z11P | 6Z49P-D | 6Z51P | 6P15P | 6E5P | 6E6P-E | D3a | |
| Vf [V] | 6.3 | 6.3 | 6.3 | 6.3 | 6.3 | 6.3 | 6.3 | 6.3 | 6.3 |
| If [ma] | 450 | 300 | 440 | 300 | 300 | 760 | 600 | 610 | 315 |
| Pa [W] | 4 | 3.75 | 6 | 3.4 | 3.5 | 13.5 | 10.6 | 8.8 | 4.5 |
| Gm [ma/V] | 7 | 35 | 33 | 17.75 | 29 | 16 | 32 | 25 | 34 |
| μ | 44 | 36 | 40 | 44.3 | 79.5 | 22 | 30 | 34 | 68 |
| Ra [Ω] | 6K2 | 1k | 1K2 | 2K5 | 2K7 | 1K6 | 960 | 1K3 | 2K |
| Ia [ma] | 16 | 29 | 40 | 15.7 | 23 | 52 | 50 | 30 | 22 |
| Va [V] | 252 | 125 | 150 | 200 | 150 | 250 | 200 | 200 | 200 |
Well, as you can see in the table above, the 6e5p and 6e6p-e (both tetrodes) were included in the list. Some interesting points to highlight: