gk71 Matrix amp

What people are working on at the moment
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#46 Re: gk71 Matrix amp

Post by Paul Barker »

To catch you up, I was wandering which value resistor Stephie used for the interstage secondary, no resistor sounded best to Stephie.

The feedback resistor 813 plate to interstage primary listed as 150k in the schemic Stephie suggested try values around 100k.

I’m so glad our so valuable guide over these many years is still available for questions.

I feel I have the makings of an amp.

Theres a plan for 800v B+ sat on its huge plank of wood, think Ill have to fit that into a tower of power hostess table on wheels, or maybe an old hospital square topped stainless steel trolley we would take into bedside with sterile stuff on it we required. Might find a used one on ebay or somewhere?

I can pick the 400v screen from the vr’s which will cope with screen current, and still manage their main task of voltage referencing and supplying current to the grid of the 572b’s. Well that covers output stage b+’s. I just wanted to do this Stephies older way, as ive built two amps with the exact same power supply but set up for different B+, and for fun I added a few caps, even the Black Gate wkz’s. Capacitorless tranced it.

For the rest of the regulated power supplies, I’ll figure out later or just build Stephies.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#47 Re: gk71 Matrix amp

Post by Paul Barker »

Paul Barker wrote: Mon Jun 22, 2020 9:46 am



you see an unexpected link between white led and vr tube which defines the master voltage for evry voltage level. Sort of indicated on schem as optocoupled. ( Paul thinks ‘erm Stephie have you gone mad?’) so Paul dismisses the led as he doesn’t yet understand it. Later, reading the discussion; “sometimes VR tube firing voltage increases with no or low ambient light. I used a white LED optically coupled to the VR tube to prevent this.”
Image
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#48 Re: gk71 Matrix amp

Post by Paul Barker »

Just an anecdotal mention of the results of looking closer at capacitorless power supplies for this amp.

I played mind games for a while. I’d bought som cheap I sh Russian 811a’s until I noticed dissipation a paltry 40 watts. The Western world 811a’s we’re very conservatively rated at 60 watts but Ham’s I knew said they were fine at 80W. I’d need 5 Russian 811a @ 40 watt to shunt Regulate same current as output stage demands. Fortunately I have two Svetlana 572b (the b with anode out top not the 572-16) which I used on the last project. So I quickly scanned other valve options. TY4-400 which I shunted the B+ of the 833a joint effort hook up Nick and I did at Owston. Very nice I thought. But though I have five and a single one would kill the current pull demanded and then some. But I like to keep my affordability hat on. A quick Skegness on eBay shows £250 each not unusual. So mine are staying safely in the chest of drawers that’s got gm100’s 212’s 833a’s as stable mates. Also becoming untouchables due to their rarity value and prices demanded on eBay. So searching Eimac valves listed and gu82 going cheap. Excellent wattage, kills the current I need but it doesn’t bias up Where I’d want to operate it for this project. So sadly forsook it. Not satisfied to let the search stop at 572b where it’s stalled. So see where ultra cheap GK71 biases. It’s good, like 572b a pair does the job. Output stage also gk71 so serendipity. Like the Sound Practices shunt regulated 211 amp that had pair 211 as final reg. The heavyweights who designed that amp claimed goodness for the same valve element.

But 572b better at job.

Fun to explore but you end up where Stephie pointed you in the capacitorless b+ designs.

I like to explore such issues.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#49 Re: gk71 Matrix amp

Post by JamesD »

Interesting short report - thanks Paul...

GK-71 is a a quite capable beastie. Its amazing how consistently good Steffie's analysis and advice is...
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#50 Re: gk71 Matrix amp

Post by Paul Barker »

So true, Stephie is humble and says little but the more you do that he’s taught the more you like his stuff. If Stephie says I like it it’s deep,
Gk71 is the bargain of this time imho.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#51 Re: gk71 Matrix amp

Post by Paul Barker »

Found that I’d prepared for more capacitorless power supplies, which I’d forgotten about. Before the price of SV572’s took a hike I’d stashed 3 pieces SV572-160 (no top cap 125watt) added to the existing pair of SV572b (160 watt). Good job, as these days Western 572b’s are rare and expensive. Proliferation of Chinese production prevails. Not for me thanks?
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#52 Re: gk71 Matrix amp

Post by Paul Barker »

Still planning stage only of the chokes for the numero zero capacitors power supply. I sort of have enough chokes, which Ive used for the previous versions but this version demands more, os it’s a messy business. Theres the vintage potted double C core I originally bought to build Flesh and Blood which is the only choke that is 0.5 map dc current in choke input capable. But all the others are 250ma in pairs so past the first choke the rest arranged in parallel pairs. Messy.

So Ive ordered 23 kg of gapped (3 mm) scrapless laminations to make EI150’s, aswell as the bobbins and the frames to make four; that weight of iron is enough with margin, to neatly, though heavily, live up to the rediculous plans I have for B+, screen, and b+ and screen of the driver.

The configuration of attachment for the pair of power transformers 2,920VCT/0.25 amps each should land main b+ at the signal section at 900v, at least I know it’ll not lag beneath the target of 800v master b+. As previously stated, gk71 would barely flinch at 900 volts.

The safest approach is wind two, hold two in reserve should they be called upon. run 500mA dc through them, to see where that puts the b+, so my famout blue variac is back in favour to fine tune, from which I,ll AOT. Further to journey, but plans better completed now such that non of that daft shenannagins with power chokes that end up costing hundreds of £ each.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#53 Re: gk71 Matrix amp

Post by Paul Barker »

Paul Barker wrote: Sat Aug 08, 2020 8:42 pm Still planning stage only of the chokes for the numero zero capacitors power supply. I sort of have enough chokes, which Ive used for the previous versions but this version demands more, os it’s a messy business. Theres the vintage potted double C core I originally bought to build Flesh and Blood which is the only choke that is 0.5 map dc current in choke input capable. But all the others are 250ma in pairs so past the first choke the rest arranged in parallel pairs. Messy.

So Ive ordered 23 kg of gapped (3 mm) scrapless laminations to make EI150’s, aswell as the bobbins and the frames to make four; that weight of iron is enough with margin, to neatly, though heavily, live up to the rediculous plans I have for B+, screen, and b+ and screen of the driver.

The configuration of attachment for the pair of power transformers 2,920VCT/0.25 amps each should land main b+ at the signal section at 900v, at least I know it’ll not lag beneath the target of 800v master b+. As previously stated, gk71 would barely flinch at 900 volts.

The safest approach is wind two, hold two in reserve should they be called upon. run 500mA dc through them, to see where that puts the b+, so my famout blue variac is back in favour to fine tune, from which I,ll AOT. Further to journey, but plans better completed now such that non of that daft shenannagins with power chokes that end up costing hundreds of £ each.
I realise that my sentense construction is confusing and losing the meaning. I’ll simplify it. At the end of phase one the intention is to have three chokes in series through which the entirety of the positive voltages required by the signal circuit are sprovided, the negative bias voltages are yet to plan and not part of the b+ supply. 800v to 900v b+ here discussed is required to provide 0.5 amps for the shunt regulation of all potentials and signal current to a 50:50 target, but not critical, give or take should do. Strictly speaking I don’t fully comprehend the Matrix, so I expect the signal current to vary, which means the shunt valves shall also mirror inversely the variation. Hence why the situation shall require AOT (Adjust On Test). Dont be frightened, I ve built scenario in every serious SE build Ive made since the first Eggborough where I demonstrated it on the PX25 amplifier, my first public capacitorless power supply, which had taken a fair amount of tweaking at home before public audition. We’re talking here about the tweaking at home.

History has taught me that this phase Im discussing, the planning, is the key to not destroying valves. Though the real problem vis a vis destroying valves is the signal circuit. Admittedly I biy off more than I could chew with the last Se amp built this way which I never brought to public audition, which is a shame because after I destroyed all the expensive valves and replaced them learning from my mistakes, it was the best sound I have made in my life. A member who doesn’t post did hear it and would confirm it was the mountain peak of my efforts so far. But it occupied the centre of a large room on our top floor which had to be let out due to personal circumstances and it is no more.

It is my hope and expectation that the GK 71 Matrix hsould recover that huge leap in sound quality I achieved to date. I trust Stephie like noone else to have designed s great sounding amplifier in the 813 Matrix, so am devoting this intended exceptional build to Stephies say so that it sounds better than the se 845 amplifier.

I am very simple minded, and I prefer to continue my power supply efforts with the proven formula provided to the world by Stephie in the resistorless/capacitorless amplifier. But I am not stating categorically there shant appear a small capacitor or two in response to unwanted noises, but the 813 Matrix signal section has no coupling capacitors or bypass capacitors, so its not far from a capacitorless design.

So, the resistorless (apart from the resistance of the chokes and valves) power supply plan is:

Two parallel 250mA HT transformers shall each have a pair of U19’s, the sum of which paralleled shall combine to single HT through which three 10h 500 mA chokes take over with a total dc resistance of 600 ohm because the U19’s require it. (I don’t have mercury valves and anyway the u19’s never let me down, plus I purchased sufficient for my lifetime when they were cheap and plentiful. In honesty exposing my fallability I am really struggling with the dc resistance needs at this point of the design. Effectively the resistance of the three chokes. one pair of U19’s rectifying for 250mA demands 600 ohms choke input. But there is another pair rectifying the other transformer the output of which combines before the three series inductors through which the combined current becomes 500 mA So actually what is the required resistance? My simple mind thinks 600 ohms. Have I got this right my better minded friends? I could ask Stephie but I honour and respect Stephie from long ago and don’t wish to awaken discomfort in Stephies present way of being.

But say I am right I know the resistance of choke one, I just have to do the calcs of winding wire to fill the window, and provide the right resistance yet remain capable of the power demand without exceeding 55 centigrade (a fairly common maximum insulation limitation of winding wire), 0f course potting in sandy pitch was the old school means of addressing tempertaure demands as the sand spreads the heat away from the core, so that shall be my style of choke design. All things old ways, nothing new under the sun. Those old pitch and sand potted military transformers remain solidly reliable to our time. How do I know this method? My transformer design handbook, which I cant place my bands on, but trust me its not a feature I would get wrong, with my type of mind. I dont need to give you author chapter and verse, its in there.

Ive drifted off topic haven’t I.

Any thoughts on the dc resistance of these combined transformers? is it 600 ohm?

TIA

Oh joy of joys, looking for the book I found two mesh plate 41 MTL’s. Those shall go in a wholly different project, a mu of 45, I feel a dinky little PX4 amp in my blood.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#54 Re: gk71 Matrix amp

Post by IslandPink »

In the first para, you mention shunt regulation. Why do you need three chokes, and choke input, if you're going to use shunt regulation at the end ?
The effect of the shunt-reg is to decrease the demand on the PS to almost nil. Therefore the fist part of the power supply can be conventional, it just needs to smooth the voltage ripple.
Mine uses cLCLC with two 10H chokes, and gets about 2mV ripple before the shunt section.
However, maybe I got lost in your description there !
"Once you find out ... the Circumstances ; then you can go out"
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#55 Re: gk71 Matrix amp

Post by JamesD »

Isn't it that Steffie's design uses shunt regulation in place of the capacitor so its a no C psu that might be described as an LShunt psu rather than LC or RC?

Then we have choke input into multiple LShunt pairs to produce the low ripple psu required with the secondary characteristic of having good psu regulation via the shunt reg rather than that being its primary characteristic - OK that is maybe splitting hairs w.r.t. the shunt reg functionality :D
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#56 Re: gk71 Matrix amp

Post by Paul Barker »

IslandPink wrote: Sun Aug 09, 2020 12:20 pm In the first para, you mention shunt regulation. Why do you need three chokes, and choke input, if you're going to use shunt regulation at the end ?
The effect of the shunt-reg is to decrease the demand on the PS to almost nil. Therefore the fist part of the power supply can be conventional, it just needs to smooth the voltage ripple.
Mine uses cLCLC with two 10H chokes, and gets about 2mV ripple before the shunt section.
However, maybe I got lost in your description there !
Shunt regulation has same effect on U19 as the first capacitor surely?
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#57 Re: gk71 Matrix amp

Post by Paul Barker »

Image
Image

Found the book, here’s the relevant paragraph on pitch and sand.

I might be remedial educationally but I never forget what those with PHD’s in this subject have writen when it’s a small element of this hobby I’ve studied for myself alone.
Last edited by Paul Barker on Sun Aug 09, 2020 2:12 pm, edited 1 time in total.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#58 Re: gk71 Matrix amp

Post by Paul Barker »

With utmost respect Mark you don’t say what is the dcr of you’re chokes and what rectifiers used or their dcr requirements, so how can youre anecdote help answer my question?

As you said James.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#59 Re: gk71 Matrix amp

Post by Paul Barker »

I’m still baffled by the the Thevian resistance as seen by each U19 is, and for each u19 it has to be 600 ohm at least. It’s got nothing to do with the quantity of chokes or their inductance, my question is solely about the dcr required. Whether it’s one choke 2, 3, 4 or 5. Other aspects of the design require a particular inductance, I don’t need that question answering for me as I’ve successfully built such supplies as afore mentioned. The difference this time is the decision to go u19 in place of the schottky bridge I’ve used formerly.

TIA

I don’t wish this question to demonstrate arrogance, right is right. That’s all. I wish to treat the u19’s right !
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8879
Joined: Mon May 21, 2007 9:42 pm

#60 Re: gk71 Matrix amp

Post by Paul Barker »

Image
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Post Reply