|
|
Post by paulrbarker on Sept 30, 2024 6:09:52 GMT
This is the present circuit diagram how I listen to it today , except that the cathode bipass pio cap and the small one to b+ arrived from China in shocking condition bashed and bruised dented and dare not connect them in. So the bypass is only 50uF.  The future direction is to use Stephie Bench’s valve diode bias method. Wherein there is more gain for the E88CC but the cathode feedback single stage is forsaken.. but the consept Stephie was aiming for was better high order distortion with using valve diodes rather than bypassed resistors. So to stay true to Stephie’s consept I have to change both stages. If there is a audibly better sound it would probably occur in the output stage. I fear the first stage may sound better with the unbypassed resistor. But I have to try the diode bias there aswell. Stephie shows the distortion profile and to me it isnt clearly better. But I like experimenting and judging with my own wars. Even if I am deluded. If I am enjoying the sound, accepting that may not be truth I am happy to enjoy the experience even if deluded.  200 ohm grid stoppers all grid’s |
|
|
|
Post by paulrbarker on Sept 30, 2024 6:22:54 GMT
“If decouple required it will be added” stems from Jeremie Epstein’s experiemnet in the Darling Fest he found having the same amount of smoothing in both stages had qualities of cancellation. I tried this with a previous version and I found something different here was night and day better than Jeremies statement. That is why my consept was put in the first circuit above. Jack Eliano gave this concept a name but I wont use the name. We experiemnters used the technique before him before naming it. The concept is inject the power supply noise in the ratio of mu +1 out of phase into the cathode. That method only works if you use cathode bypass. So it it irrelevant on circuit two. Therefire the stage one may require greater decoupling for circuit two. I dont know the unswer until I audition it. But note if not required Jeremy is right.
The further decouple would probably be in the form of a choke and VR shunt regulation. I am not interested in using mosfet constant current anywhere, if I want ccs in the cathode I will use a triode, but this design is about using diodes in the cathode, which means no ccs there as well. What is the diode bis compared to audo bias? It is a hybrid of fixed bias. People in the old days that used fixed bias praised it, and many praised a mix of both. So maybe this diode bias in as much as it is fixed and has no bypass caps could it be the best option? That is what we are to find in hope of improvement.
but I never liked solid state diode bias, neither did I like battery bias. Before now the best bias I have heard is unbypassed resistor. But I havent valve diode bias and that is my next learning experience.
Old Amos a man who lives in the Yorkshire Dales said “experience is summat you get the minute after you need it. “ that is why we have to try things we havent yet experienced.
|
|
|
|
Post by paulrbarker on Oct 24, 2024 6:47:30 GMT
I sacrificed my enjoyment upstairs to make the changes to this EL38 amplifier. The first circuit diagram wasn’t correct. Rather than edit it I’ll just give a talk on it. Ive learned that what culminated in a bleed stroke “ Cerebral amyloid angiopathy (CAA) is a type of cerebrovascular disorder characterized by the accumulation of amyloid within the leptomeninges and small/medium-sized cerebral blood vessels.[1] The amyloid deposition results in fragile vessels that may manifest in brain bleeds. It may also present with cognitive impairments, incidental microbleeds, hemosiderosis, inflammatory leukoencephalopathy, Alzheimer's disease, or transient neurological symptoms.[2] It can occur in certain familial syndromes or sporadically.” had affected me for a few years prior to the stroke.
The Occupational therapist at the stroke unit asked me if I had any hobbies. I said “I like designing and building valves amplifiers” She said “thats great, keep doing that!”. This is primarily a tool to keep my brain occupied and counter the loss of brain function associated with CAA. The physio at stroke unit also suggested building my amps to get motor control on left side back.
So Im doing quite well at improving motor control. At the moment I am using the left hand where I would normally use my right. This remaps the brain cells to replace the ones which are bruised at present.
So I found a fault. I had put two more 2uF caps one per channel from cathode to b+, which would have been OK if they weren’t then over voltage. One bled its oil and would have pulled down the voltage on that channel. So I’ve removed both and binned them.
Having stripped out the error, Im now rebuilding that section, using EZ80 diode bias on the EL38 cathode. So no bypass caps, and no place for cap from cathode to b+ consequently. This may entail better smoothing of b+ to counter the lack of opportunity to cancel b+ noise through cathode noise which is opposite phase.
I have decided not to use vakve diode bias on stage one. Firstly the gain is 22 on that stage which is sufficient, secondly this cathode feedback is the only feedback in the amplifier. I have always liked the sound of this type of feedback. So it stays.
So the actual building is a challenge to my left hand, but by the process the brain is learning to employ that hand more. I am pleased with the progress, and with the success in dexterity so far, and with the challenge to the mind. Im not in a race, but I should like to listen to vinyl on the Quads asap.. but due to the likelyhood of mental errors, Im at least tripal checking each action.
I’ll redraw the cirquit diagram when its completed and working.
Just a note though, the bass should improve because when you bypass cathode resistors you limit bass and that all adds up to join other stages limit of bass to matter as a whole.
On the Quad’s there is a need for better bass, this should help that goal.
|
|
|
|
Post by paulrbarker on Oct 24, 2024 7:19:33 GMT
Oh and I had another stroke, no beds in Scarborough hospital so joined the queue in the corridore of Scarborough hospital between A and E and Resus. Just one night. When the Dr finally got to me 6 hrs later 7 hrs after the stroke, bogged down by all the malingerers.
Anyway he was an idiot, he sent me for a ct scan. Ct scan doesnt reveal the damage in CAA. He also suggested I take aspirin (the worst thing a person with CAA could ever do is take blood thinners.) I told him that, he went away and on return said “you musnt take aspirin!” “You have had a TIA, you can go home and Ive arranged for the stroke unit consultant to give you a telephone appointment later today.” I thought, good because Im getting no help here!
So the consultant rang me, told me it wasnt a tia it wast a repeat CAA stroke. He asked more about my life leading up to the stroke and explained all the symptons I had experienced in the years before now make sense. He aplauded my GPfor starting me on Ramipril additional to the Amlodipine which wasn’t controlling my BP sufficently. The GP is young and only just qualified as a GP but she did a six moth rotation on a stroke unit. So she will do for me, and the consultant was pleased to hear of her experience.
The main action I can take to improve my prognosis is get bp under control. I am delighted to tell you it is now controlled. So my journey of recovery has begun in earnest.
|
|
|
|
Post by paulrbarker on Oct 24, 2024 7:22:23 GMT
The GP has also referred me to neurology physio team in York.
|
|
|
|
Post by paulrbarker on Oct 25, 2024 6:43:54 GMT
Meanwhile I have made some progress. First of all I just want to give some background. I have a visual mind not a wordy one. When I build an amplifier I understand the circuit visually, i know the maths and calculate my own maths. I dont need someone else to give me a cirquit, decide for me the parts the valves and the voltages. Its a bit like in the building trade, you dont want anyone else getting the toping out ceremony.
Every choice is fitting my overall plan. For example I wanted to work with a first valve that meets all my criteria. In about 2001, before VSAC at any rate. I built an V1505 Amplifier. Chris Vryonides asked me to bring it down to London. There were two others there, one guy name forgoten he had built a push pull 813 amp but didnt bring anything and was on a flying visit. Didnt stay. The other was remaining quiet. I told him something about cathode followers, he replied “I know I told you that!” Then he told me he was Morgan Jones.
So why am I telling you this? Because Morgan Jones said to me, “since you have such a high B+ for youre output stage, why don’t you use resistor loading onthe voltage amplifier?” A simple statement but also a good principle. Geuss what a solid state ccs is not by any means faultless. Every method has a sound. If I used a ccs it would be valve achieved. I heard Gary Pimms stuff at vsac. I wasn’t sold on it. Morgan taught me a tool to harness.
Do you see how a circuit starts building in my head. The choice of ECC88 was arrived at to satisfy the simplisity of Morgans suggestion. Psychologists say we carry the past rolled up like a carpet behind us. Thats just one example. In my carpet are other nuggets from Stephie Bench; Jonathon Noble, and what others have said to Jonathon.
Something else Morgan said that day, was “the bass roll off at each stage is additive.” A stimple principle but often ignoranly missed.
That will do for now, I could talk all day on every little lesson that is in this aparently innocent amplifier.
My condition can give me the same mental result as Altsheimers. I am starting to realise why the OT’s eyes lit up when she asked if I had any hobbies, and I said “I like designing and building radio valve amplifiers.”
|
|
|
|
Post by paulrbarker on Oct 25, 2024 16:30:12 GMT
Here it is  Sounds better. Very glad I made the effort. |
|
|
|
Post by MikeMusic on Oct 25, 2024 18:29:24 GMT
Keep at it Paul
All the best with the recovery
|
|
|
|
Post by Clive on Oct 26, 2024 7:56:56 GMT
Interesting stuff Paul, both for your recovery journey and the development of the circuit. Your recovery journey gives us all insight into how to overcome the trauma and effects your type of stroke, A close friend of mine lost his ability to speak and read, he has made a lot of progress over an extended period repurposing other areas of his brain to take over these functions. Progress has been impressive though there are limits to what’s possible as his stroke was very severe.
Your design journey with this amp may even get me building it.
|
|
|
|
Post by paulrbarker on Oct 26, 2024 8:57:32 GMT
I couldn’t speak at first. I’ll give moredetail about that journey later. Staying on the amp. I liked the late Patrick Turner from Australia. He gave his knowledge freely. If he hadn’t drawn up these triode connected characteristic curves. I couldn’t have designed this amp quite as well. On his chart, my operating point is the highest one, I drew my operating point over his curves. So you see here. I put in the 5.9% 2nd harmonic at max output. Reasonable for SE triode. I’m pushing it a little with my 430 B+, take off 5v across opt, 20v for bias and you have Anode voltage 405. Rating of g3 is 400v. Not out of tolerance exactly. The sound after this change to valve diode bias has cleaned up the sound and the bass is much fuller, the highs cleaner and the singers real people in the room. I think we could call valve diode bias fixed bias. 25 years ago more knowledgeable friends said “fixed bias works best”. Turns out they were correct. Its much better. I dont think Id go back to auto bias. this then leeds on to ‘shouldnt I try diode bias on stage one. I suppose I must. But I have to recover from going this first change. My left hand needs a period of rest. It found the dextrous demands very difficult and threw me out by rebelling and going into shaking spasms. Its good to use it and get the brain remapped, but its baby steps and rests between. My green load line is 4k1. But now I am using 5k6 OPT’s, which is better sounding and I still have plenty of power for the Quad 57’s. Too loud for my wife at 20% volume pot position.  |
|
|
|
Post by paulrbarker on Oct 26, 2024 9:29:33 GMT
My take on speaking after my stroke is that its stress / psycho somatic. When I had to explain what had happened to the first medic after stroke out came some very strange sounds I couldn't mimic but it wasnt intelligable English language.
Probably now two months after that first stroke on normal subjects I talk fluidly, but within the limitations of my mind’s ability to present me with English language words to speak. So I know in my case the limitation is in the mind. The speech area of the brain isnt at fault if there is no psychological challenge. But if I have to reiterate what happened to me during the stroke my mind thriws me out. No, I will not relive that time when I was asaulted (my mind’s reasoning, my thoughts about it). So if I dont punish my mind tasking it to bring to my conscious something it has thrown out. I speak fluidly, only limited by my minds ability to recall, which right now recall is poor but that isnt a speach issue its a memory issue, which I also have.
|
|
|
|
Post by paulrbarker on Oct 26, 2024 10:18:09 GMT
|
|
|
|
Post by paulrbarker on Oct 26, 2024 10:24:32 GMT
|
|
|
|
Post by paulrbarker on Oct 26, 2024 10:41:15 GMT
Sorry cant fathom why I hadnt posted a public limk and cant fathom how to right now. So sorry, not happening.
Baby steps climbing out of a stroke
|
|
|
|
Post by paulrbarker on Oct 27, 2024 10:24:50 GMT
|
|
|
|
Post by paulrbarker on Oct 27, 2024 11:01:07 GMT
|
|
|
|
Post by paulrbarker on Oct 29, 2024 5:40:48 GMT
Strangely the second soundbight is now free’d up so you can hear it. Go figure? I gave up on it and it was fixed without me.
|
|
|
|
Post by paulrbarker on Oct 29, 2024 5:51:54 GMT
By the way I was naughty after I realised that Patrick limited the max dissipation to the anode only, so I added the 8 watts allowed via the g2, and had a real good naughty time alone knowing I was pushing it a bit, as to get the 33 watts B+ was turned to 470. The Black gates’ are OK up to 500v. It was a moment of naughtiness and sounded better still.
You see when you have fixed bias and you up the b+ the current lifts. Lifting the current draws the operating point up and away from the pinching in the characteristic curves in the bottom right corner to where they open out, which translates into less 2nd harmonic distortion. In this case enough to hear.
Next task : draw on the chart where it was. When Ive done that it shall appear below. I leave the amp on all the time for benefits you get keeping Black Gates charged. So I backed off the voltage to 430.
|
|
|
|
Post by paulrbarker on Oct 29, 2024 6:50:38 GMT
|
|
|
|
Post by paulrbarker on Oct 29, 2024 7:06:26 GMT
The above would require 450v B+ Next I’ll have a go at estimating 2nd h distortion. The answer is 3%. The advantage of raising the current on a Single Ended Triode.
The downside is I have pushed the g2 beyond its 400v limit by 5%. That is probably not a problem. Up to you, I have no fears.
There is another option to raise anode voltage without pulling up the g2 beyond it’s limit, which is ultra linear.
we don’t have the ultra linear curves, but in the interests of experimentation, if I decide to try it.
I would raise the anode voltage as high as I could. The limit is 800v, 43% of that 344v keeps g2 safe. But I don’t have a SE transformer with ultra linear tap. So this becomes a major change. But not impossible.
I have Sowter 8k ultralinea 43% Push Pull transformers, so that’s one way. Another way is to use those transformers parafeed SE. But no, I think Ultra linear PP is the next best option, if we dont mind feedback. Ultra linear requires less feedback than Pentode, and that could apply subjectively to taste at the cathode of the voltage amplifier. So the circuit would have single input valve same E88CC, 6AL5 valve diode bias = mu 33. II phase splitter II output stage. In class A 40v peak to peak PP at grids. / 33 = 1.2v input sensitivity. In class A power output unknown due to no characteristic curves. But if it were triode connected it would double, 13 watts. But we have a clue of power output Ultralinear in the Vade Mecum characteristics for the EL34 which shows class A push pull triode as 6 watt, class A and AB 16.5 watts UL as 20 watts class A (16v bias) and 37 watts Class AB 26v bias.
So we could expect 20 watts PPul class A. Anyway, its worth a try.
|
|
