New KT120 PP In The Style of the Leben CS1000P

What people are working on at the moment
Post Reply
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8877
Joined: Mon May 21, 2007 9:42 pm

#91

Post by Paul Barker »

That's good to know.

It makes me wander if there is some kind of materials technology.affecting the sound, because the curves for that valve are not as good as the one's which sound same as your modern production valves.

The important message is tube rolling is worth it.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#92

Post by Cressy Snr »

Paul Barker wrote:That's good to know.

It makes me wander if there is some kind of materials technology.affecting the sound, because the curves for that valve are not as good as the one's which sound same as your modern production valves.

The important message is tube rolling is worth it.
Aye..tube rolling is a good thing to do within reason, but eventually one has to make a decision about when to stop and leave it alone (if you see what I mean) otherwise you chase your tail forever, something I have been very good at doing in the past.

The 6201 and the A2900 give a great sound, but it is not an ideal situation, as there are different valves on each channel. The A2900 is slightly higher gain than the 6201 and this skews the image towards one speaker. I do however know the potential of this setup, if I acquire two of the same of one or the other of these fine small signal valves.

For the moment then, I have gone back to the Electro-Harmonix, which are a bit more neutral sounding and less sumptuous velvet than the 6201/A2900. I'm keeping the older stock for best.

As far as the EH 12AT7s are concerned, I have two types, one is a short plate, low microphonic version with a WC suffix and the other is a long plate EH version that looks like an ECC83 inside. The short plate version is becoming my favourite of the 12AT7s I have tried.

I bought four of these to go in the Rozenblit Son of beast. At Steve S place, where it was the very last time it worked properly, the sound was a bit good.

These unassuming little modern production valves do the same for the KT120 output valves as they did for the PL519s I had in the OTL. They seem to get out of the way of the music, more than anything else bar the GEs you sent, which pull off a similar disappearing act.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8877
Joined: Mon May 21, 2007 9:42 pm

#93

Post by Paul Barker »

that's a relief.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#94

Post by Cressy Snr »

Thought I'd draw up a class A pentode version of this amp, with plate to grid balanced feedback.

Image

It'll put out 25Wpc in class A. I can't get any more without more HT.
Each valve operates at 42W dissipation and 130mA quiescent current, with 225V on the screens.
The valves are rated for 60W dissipation, so they are just cruising.
With the 6K a-a OPTs, you'd need 450V HT and 150mA to reach the 0V grid curve with 225V on g2
I can only reach the knee of the 5V grid line with this mains transformer and the BLS OPTs so if peaks go over 25W, clipping would be sudden, nasty and very tweeter frying.
But screen voltage can be lowered if necessary to bring the 0V line down, as I would want to run these in pentode with a screen voltage roughly half the HT voltage, so a better amp is doable, for maybe 20Wpc Class A with less screen volts. Extrapolating, I reckon 175Vg2 would do nicely and get the loadline nicely away from the knee at 0V, with the HT. I have available.
A pair of monoblocks with the right HT and -15v of bias would make 40Wpc in class A1. Now where's me pair of Danbury 450V 500mA mains transformers.

What it'd sound like, who knows?
Would it save the soul of rock'n roll like Gizmo sez?
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#95

Post by Cressy Snr »

Had an interesting time today with the KT120 amp.
I had felt that the sound had been going off recently so a bit of investigation was called for.
I decided therefore this morning, to check underneath the chassis, to make sure everything was OK and made a discovery.

I had originally spec'd the output pair cathode resistors for 25W, but at the last minute during construction, had fitted 10W ones,
as the calculated dissipation for the current flowing through them was 6.125W (35V at 175mA) and as they were made up of 2X100R in series, I figured, I'd get away with it. Not so it would seem.
Well these Arcol, aluminium clad 10W wirewound resistors didn't look too healthy to me. The writing had disappeared off them, and
the busbars leading to and from them were blued with heat. The solder joints were burned, as if the resistors had heated up
to such an extent that the solder had melted and hardened quite a few times as the power had been cycled on an off over the 8 weeks
it had been in operation. This was a problem that needed sorting out sharpish.
The upshot of it all, is that the 200R 10W cathode resistors were changed for 150R 50W wirewounds, as these were the nearest value I had
that were man enough for the job. I also removed the 2500uF cathode bypass caps as they had been subjected to serious abuse and I could not be sure they weren't completely shagged. The 10W resistors fell to bits in my hands on removal and were consigned to the dustbin.
Subbing in the 150R resistors has boosted the current through each valve to 103mA each, giving 206mA per output pair; around 68W plate dissipation per channel.

The sound is back to its gorgeous self, thank goodness. Treble tone is all shimmery in its loveliness once more, soundstage width is back to its former widescreen presentation and the famous KT120, ass-slappin' stomp is back with bells on :)

Here is the revised schematic with voltages as measured this afternoon.
Image

Good job I found it before Owston
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#96

Post by izzy wizzy »

Looking at the Welwyn datasheet for them WH50 jobbies, without a heatsink, they should be derated to about 12W so your 6W should be dandy :)

The 10W ones can only dissapate about 4/5W in free air.

cheers,

Stephen
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#97

Post by Cressy Snr »

izzy wizzy wrote:
The 10W ones can only dissapate about 4/5W in free air.

cheers,

Stephen
Oh dear :oops:
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#98

Post by Cressy Snr »

OK, so I have reconfigured the outputs as true beam pentodes with-plate to-grid feedback.
I have 376V across each valve, and am running regulated screens at 188V, exactly half the plate-to-cathode voltage,
which fell quite nicely, with a 75C1 and 0B2 in series.

The amp is now running class A/B and as a result, the heat output, which was frankly concerning, is much lower.
Class A triode was way too hot for my liking, and the mains transformer was also getting a bit hot for comfort. I don't like hot, only the valves should be hot.
Given the fate of the first set of cathode resistors, I now feel a little easier with the amp running class A/B.

As to the sound of KT120s as true tetrodes, the bandwidth is a lot wider, bass is less plummy, but still as well controlled,
which suggests I have got the feedback about right.

Tonally, I'm not sure it's quite as refined as triode, but it's early days.
The upside is that it rocks like crazy, with a wonderfully 'live' sound.
As the late Dr Gizmo was wont to say, you can have refinement or you can have power. Adults have to make choices.
I'm not sure which one I want just yet, but pentode mode with regulated screens is seductive.

It's putting out about 35Wpc with the first ten watts in class A.
But of course that's not the same as ten watts class A from an amp running in proper class A
as the class A part of my power output is at the crowded end of the pentode curves. So the first watt is the worst watt doncha know :roll:
Still, it doesn't sound bad to me.
Owston will be the arbiter unless I chicken out and go back to the oppressive heat of KT120s running class A, drawing half an amp at idle from the mains transformer. :angryfire:
Sgt. Baker started talkin’ with a Bullhorn in his hand.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#99

Post by JamesD »

Steve,

Umm - this isn't the sound of true tetrodes. With plate to grid feedback the characteristic curves are triode like not tetrode like :D

With the local feedback the distortion should be quite low and, like Paul found with the Aikido output, it takes a bit of time to get used too.

Not surprised the bottom end is better as the output impedance is quite reduced by this but also the reflected speaker load 'funnies' are subject to feedback and reduced too.

If you really want to play you might try the plate to grid feedback in triode mode too - it should do the same improvements there as well and you can still run AB1 or AB2.

ciao

James
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#100

Post by Cressy Snr »

JamesD wrote:Steve,

Umm - this isn't the sound of true tetrodes. With plate to grid feedback the characteristic curves are triode like not tetrode like :D

With the local feedback the distortion should be quite low and, like Paul found with the Aikido output, it takes a bit of time to get used too.

Not surprised the bottom end is better as the output impedance is quite reduced by this but also the reflected speaker load 'funnies' are subject to feedback and reduced too.

If you really want to play you might try the plate to grid feedback in triode mode too - it should do the same improvements there as well and you can still run AB1 or AB2.

ciao

James
Thanks for that James. Time to read OH Schade's article again to refresh the memory as to what actually happens with this sort of feedback scheme.
Thanks for pointing it out.
I'm really getting to like this presentation. As you say, it can be quite surprising when you hear it for the first time.
I heard similar with Alex Kitic's EL84 SEP, but with push-pull, the clarity and insight is quite startling.

Here's a nice glowy pic.

Image

It's crying out for a night shot :wink:
Sgt. Baker started talkin’ with a Bullhorn in his hand.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#101

Post by JamesD »

Sure is! Its a really nice looking amp - everytime I see a picture of it I'm very impressed. I can't wait to hear it!
User avatar
Mike H
Amstrad Tower of Power
Posts: 20157
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#102

Post by Mike H »

Blimey! :shock:
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#103

Post by Cressy Snr »

I re read the whole of the 45 page OH Schade paper on beam power tubes.
There is certainly some impressive engineering afoot in the production of a beam power pentode! I would recommend this paper to anyone like me who is interested in building amps using these types of power valve; no-one else on here I guess :sad:

The explanations of how the characteristic curves are the way they are and the mechanism behind the knees of the curves appearing at low anode voltages was extremely interesting, as was the mechanism by which the electron beams are formed and focussed towards the anode, and the fact of the beam pentode's suppression of objectionable higher order harmonic distortion by virtue of its design differences to those of true pentodes.

Lastly, the addition of the plate to grid feedback characteristics to the pentode curves so that they become triode like made my understanding of how ultralinear curves are formed a good deal clearer. It's a similar mechanism.

Given all of this it seems to me that operating my KT120s in class AB1 pentode mode, with plate to grid feedback, is much easier on the valve itself than operating it Ultralinear, because for the same, triode like characteristics, the screen voltage can be made much lower, thus stressing the screen less. Hell, these hefty valves will outlast me at the rate I'm running them now.

One can see that commercially, UL operation of push-pull output stages is much cheaper to implement than pentode with schade feedback as you only require a couple of transformer taps and a single feedback resistor and cap, back to the input stage cathode, as opposed to a whole regulated screen supply, plus a pair of carefully matched feedback resistors per channel. The risk that UL might bugger your valves more quickly, throws the additional cost onto the end user rather than the manufacturers, who of course want the biggest bang for the smallest buck.

Us DIYers have no such problem luckily.

The sound of the amp with Class AB1 pentode operation gives nothing away to Class A1 triode operation. The bonus is that after reading the Schade paper. I am now beginning to understand why this is so.

There's nothing like a bit of theory to back up your practical skills. How come those old guys, seemed to be able to put across such complex concepts, in such a user friendly manner?
Last edited by Cressy Snr on Sat Oct 26, 2013 4:47 pm, edited 1 time in total.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#104

Post by JamesD »

Excellent Steve,

Now when you read the Tubecad 'partial feedback' tutorials you will find an alternative explanation of the feedback mechanisim - one not at odds with Schades at all but from a different perspective of a general feedback model. One thing JRB brings out is that the power tube grid now becomes, in most ways, a virtual earth point and that this is best driven from a high impedance i.e. constant current style driver. This means a small signal pentode is in many ways a better driver than a triode... Something to think about - a pair of EF86s or 6BR7s would probably be optimum...

J
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#105

Post by Cressy Snr »

Thanks James.
Yes,
A pair of small signal pentodes was in my head when I was conceptualising this amp; sadly, I have no means of matching them up as I lack a test rig.

I've just read the Broskie article. Interesting take on the plate to grid feedback model. It makes a nice addition to the 1938 Schade paper, particularly with his explanation of the need for higher impedance drivers in the plate to grid feedback situation.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Post Reply