Last time I wrote about the 4P1L in screen mode. It was great to see some DIYAudio member (Blitz) to post about his work on the 4P1L with screen as anode. I call it screen mode but probably is incorrect.
His post about G3 structure remind me to post this, I have tested it but never blogged about it. Yes having G3 as part of the anode structure will increase conductance and will form a nice “mesh” anode. Here it is how I implemented:
The pin 4 (G3) is now connected to 3 (G2) to form the anode. I reduced the anode voltage down to 110V to get 10mA. It could be increased, sure within the Pd limits.
The response is very good:
Here you have the distortion at 1kHz:
How does it sound?
Well, I wrote about it before. The 4P1L is one of my favourite valves. In this mode it sounds great, with a particular clear detail in the treble. I like this valve and will play it for some time to get further impressions.
Great article!
Now a curiosity: which software you use to produce these superb circuit diagrams?
Cheers,
Giorgio
I have modded my 4P1L version to run in with is & am finding that my anode voltage is sitting at 118V(not sure this would make much difference anyway) with Anode current sitting at 10mA, did you change MU-R, also is the 300R resistor still needed on input to stop oscillation? This sounds quite a bit more detailed with more depth, it’s sweeter with no harshness & goes higher, I truly have found it has improved in every respect, I’m sticking with this I think, it’s definitely got a much blacker background also, I haven’t really noticed a drop in Dynamics either (Blitz mentioned this on DIY Audio). I also feel it’s a bit warmer sounding which if anything this valve can only benefit from. This is definitely a positive move Ale. Excellent work from Yourself & Blitz!
Good to hear your feedback Barry, very interesting. I like the valve in “screen mode” which is obviously not the right name but hey, no one has coined a term for it. It may be a bit misleading as is close to screen drive.
I like the sound of this and agree with your comments. I would experience driving it hard to 20mA and close to the operating limit, am sure it won’t be worse, at least the bass response will be slightly improved. At least in theory.
I haven’t changed the gyrator PCB and the diagram is up to date. R-MU stays the same
Currently playing the 2P29L which is a fantastic stage, wow!
I am running it exactly the way your diagram suggests, do You think that changing to BF862 (in screen mode) might make a further improvement? I have some 2P29L’s, do you prefer this valve over 4P1L, if so in what way? I have read that Andy Evans couldn’t get on with some harshness in the treble, I like the idea that the filament current is low for it also, I do intend to build this when time allows, I have a pair Gyrators setup for it, incidentally for those who prefer using output transformers, should not be too biased with regards to the Gyrator boards, I certainly believe they are superior in sound alone, as I’ve built 3 of these Amplifiers now, 2 of them in both modes (26, 01A & 4P1L).
Hi Barry
I don’t think you will notice a big difference with the BF862. I’d probably try first increasing the current to 20mA.
The 2P29L is very close to the 4P1L in terms of overall sound. I like the clarity and detail it provides. The winning aspect is that it’s not microphonic at all and is easier to run in filament bias. What else do you want here? I think it’s great valve.
I’m not able to do a proper ranking these days as I’m changing constantly too many things on my system at the same time. Also due to travel, I can’t spend enough time for an A-B listening test. I have currently 5 preamps active on my system which I can swap easily: VT25/10Y, 4P1L Screen Mode, 2P29L, UV201a and 01a preamps. I was planning to do an Audio meeting here in London including Andy before end of year but I’m struggling to find the time unfortunately.
The tricky thing of the hybrid mu-follower (aka gyrator load) design is that as it’s very high bandwidth (above 200kHz and close to 2MHz for a 4P1L) it can reveal unwanted distortions (due to HF noise, oscillation, intermodulation) which you won’t get in a restricted bandwidth of a Tx.
What Andy complains is likely to be a result of the gyrator unveiling some other by-products in his system.
I can actually hear now some HF very low noise from my DAC which I couldn’t before when there is no music playing and if I place my ear close to the tweeter.
Try the 2P29L for yourself and listen when you get the time!
Cheers
Ale
Hi Ale, would running 4P1L not put screen dissipation around 2.2W, when recommended max is of 1.5W, or as G3 & G4 are connected together would this be closer to 3W?
I meant running at 20mA
OK Ale, I did 3 things today, I decided to run 4P1L at 20mA, I also modified Valve mounts so they were sitting on Maple Blocks & I removed the 300R Grid Stoppers, my finding are that running at 20mA gives more Dynamics & cleaner midrange with a more solid Bass & better detail, this change made a subtle but worthwhile improvement, not as bigger difference as the initial change to screen Mode, I must say that screen mode gives you the music, there is no smearing to the sound, but I am thinking there is a little more audible hum at the higher current setting, the Maple blocks improved Microphony to a degree also, but I soon put the Grid Blockers back, it was merrily singing & howling away, so that’s definitely a no no! There is no doubt in my mind that coupling G2 & G3 & running somewhere between 15mA & 20mA is the way to go with this Valve( bear in mind though that Anode Voltage will be around 150V at 20mA though, if You run it with Anode, G2 & G3 connected together, You will be missing out on what the Valve can really do, especially to the Musicality & lack of Smear!
Blitz posted on the construction of the screen grid. I’m sure it could take the 20mA between G2 and G3.
Glad to hear you like it, it’s quite encouraging for others to follow this path
Cheers
All that being said, I have found that after listening for a couple of hours, I actually went back to 10mA & noticed no negativity in the sound, (although when using in plate mode it was being run at 30mA without such negative issues) but without the noticeable hum issues & less heat being generated, now this might be down to my P/S, I built this in one chassis with a separation plate & B+Trafo as far away from the main Amplifier as possible. I would say try it but like with most changes you make, there might be a downside. I do think screen mode is better overall! Blitz is running His at 22mA, His build is Transformer Coupled I think, this might be a completely different ball game.
Hi Ale, Could this circuit be adapted for Headphone use either with or without source follower? My current phones are 600 ohms.
Hi Barry,
Yes you can, but I’d recommend a step down transformer and cap coupling, famous parafeed. With a 600R load you will need a higher gain stage.
Have you seen this?
http://www.bartola.co.uk/valves/2016/12/04/headphone-parafeed-amp-part-i/
It sounds beautiful. Still need to do final build, I’ve been playing with different valves
Loading the mu follower directly with 600R is a challenge and distortion may be too high. Try it and let me know.
What about lower impedance headphones, but not using a tranny, such as your Grado’s?
With 32R the distortion would be way too much and slew rate distortion will come into play as the valve won’t be able to provide the current sink needed given such a low impedance load. A follower will improve this, but distortion will be too high anyway.
As an example, simulating a 2P29L stage cap-coupled with a 600R load, yields over 2% THD (mainly H2) for 80mW output. Same stage with a 100K load produces 0.005% at 10Vpp output.
That’s what I was concerned about, so for 600R, Sowter Tranny wired 4:1 then with 2/3 uF Coupling Cap. There must be an awful lot of Headphone Amps out there with more than 2%THD(some of them commercially built). Thanks Ale.
I do have a pair of LL2745’s built for Thomas Mayer for using with 26/01a. It is suggested in a couple of forums that the standard version would be suitable, not sure about the TM version though, as I believe that the standard version is rated at 18mA. Seems a shame to have them sitting around.
Hi Ale,
I’m thinking that if my 8mA version of 2745’s were configured with Primary’s connected in parallel, the current rating should become around 16mA, allowing me to run 4P1L’s (screen mode) at around 15mA.This would allow for a step down of 2.8/1. Should be ok with 32R Headphones, but not with 600R, for that I would need 5.6/1, so it would be a no no. What are your thoughts?
There are threads suggesting that 18mA version of the 2745 is a good match for Headphone amp, using a suitable Valve.
Why not in parafeed with a gyrator board in 5.6:1? That way you will get away with the limitation of the primary current gap.
With the 600R headphones you will be looking at a gain stage of about 20 to get 150mW output with a 5.6:1 step down though
I will try it Ale.
Hi Ale,
I do have a slight problem with my 4P1L amplifier in screen mode & possibly before this as well, this being that when the volume is increased to about half way (approx resistance as in on your suggested circuit of R9 is about 50K) using 100k pot (I am intending on changing this for an attenuator), the Amp is absolutely dead silent, but when volume is reduced & obviously resistance increased, hum is also increased, I would leave it this way, but listening levels are too high at 50K on input, would it be worth removing fitting a 47k resistor in the position or R9 & connecting a suitable value shunt attenuator on the output of the Amplifier? I only noticed it during listening this late last night, as I thought this not too significant hum was being introduced via a Transformers, I now appreciate that it’s vary likely due to changing impedance on the input due to the pot, obviously with a decrease in volume there is an increase in input impedance, the input source used last night was from my Battery Phono Stage, I could reduce the gain of the Phono Stage, but this would then be a problem for my Dac. I don’t really want to go down the TVC route, which would be costly.
I suspect this why my 01a Amp sounds better(quieter), as gain is lower & listening would be with input at 50K or less!
Barry, I think you have some constant noise I suggest you need to track and trace. This isn’t the valve, is the implementation. In fact the 4P1L doesn’t pick up any hum at all, but the 01a does. Both have same gain, so if gain is too much for your system, then you should add an attenuator, at the output may be better for you given the hum already present.
careful track of the hum is my recommended way. If you achieved that with the 01a, can’t see why you won’t be able to with the 4P1L. Can you spot out any differences in terms of grounding and layout between the 2 preamps?
Hi Ale, I didn’t think it was the Valve. The 4P1L Amp is in a single chassis, but with a divider between PS & main Amp, as I said hum is gone with Pot set at half way, the PS, is a CRCLC with hybrid rectifier using AZ31 & a pair of shottky diodes, all grounding goes to the same point & can be lifted, which makes no difference, so grounding is as the 01a, but with 01a has separate PS, which is a lot smaller obviously. So there is no hum with Volume at half way & above, but as it’s reduced it increases to a noticeable level!
I am still waiting for a 23 way switch to com for the attenuator, so will have to wait for a while, the 100K pot is on the input, where R9 would be. Do You think that connecting the pot on the output temporarily would be worth trying as an experiment?
What you are describing is simply a better SNR due to the fact that the hum is constant and when the output is at a level that your ear cannot notice it, you think the hum isn’t there. The hum is introduced by your preamp clearly.
A pot at the output may help but what you need to check is that driving the 4P1L preamp stage at full tilt won’t create any distortion.
Hi Ale,
I have tried various operating points now & have found that 15mA is about right (10 has a little confusion in the bass & less depth & 20mA the sound becomes a little thick a smeared, I also found changing the PS from CLCRC to CLC made a surprising difference (removed 1 x 45uF & a 120R 12W Mills Resistor) to the openness of the Preamp, this mod was also suggested by Blitz on diyaudio. We all strive for excellence & sometimes less is more, I would say that 20mA definitely sounds better than 10mA, but 15mA to me seems to be the sweet spot between the two! There is less noise also, which would make sense!
Hi Barry,
Happy New Year! Thanks for sharing your experience. I haven’t modified mine for some time. I’ve been working on a new DHT raw “adjustable” supply to experiment with other DHTs and output stages and damaged some Rod Coleman regs during my careless tests. Luckily I have a plethora of DHT preamps now that I can switch and play other preamp whilst fixing the damaged!