Finished and working - 12 valve all DHT PP 26-71a-45

If they glow, this is the place to be
Post Reply
User avatar
pre65
Amstrad Tower of Power
Posts: 21367
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex/Suffolk border.

#61 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by pre65 »

My GK-71 monoblocks are 25 Kg EACH. :)
The only thing necessary for the triumph of evil is for good men to do nothing.

Edmund Burke

G-Popz THE easy listening connoisseur. (Philip)
User avatar
Mike H
Amstrad Tower of Power
Posts: 20156
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#62 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by Mike H »

IslandPink wrote: Mon Nov 26, 2018 9:23 pm Nice work. I like to see an amp that is so large, it's in focus at the front and out of focus at the back !
:lol:
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
User avatar
pre65
Amstrad Tower of Power
Posts: 21367
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex/Suffolk border.

#63 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by pre65 »

pre65 wrote: Tue Nov 27, 2018 11:47 am My GK-71 monoblocks are 25 Kg EACH. :)
I bet Nicks 211 monoblocks are even heavier.
The only thing necessary for the triumph of evil is for good men to do nothing.

Edmund Burke

G-Popz THE easy listening connoisseur. (Philip)
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#64 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

IslandPink wrote: Tue Nov 27, 2018 11:35 am
RhythMick wrote: Mon Nov 26, 2018 9:29 pm
It sits on industrial heavy duty runners in my cupboard, with the power supply next to it. I think the amp is about 20Kgs and the PSU about 15kgs.
Oh, hang on. Power supply only 15kgs ? Mine is over 20Kgs. Must try harder !
:P
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#65 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

IslandPink wrote: Tue Nov 27, 2018 11:35 am
RhythMick wrote: Mon Nov 26, 2018 9:29 pm
It sits on industrial heavy duty runners in my cupboard, with the power supply next to it. I think the amp is about 20Kgs and the PSU about 15kgs.
Oh, hang on. Power supply only 15kgs ? Mine is over 20Kgs. Must try harder !
Errr, I finished the wiring this weekend and 2 of us moved it and the power supply down to the cabinet ready to mount it on rollers. In process we weighed it by the trusty process of my superfit rugby-player of a son standing on scales holding it then subtracting his weight.

The amp is 35kgs ! (2 channels).

The PSU didn't get weighed but it's a touch lighter, guessing 25 to 30.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#66 HELP - SAVE ME FROM THE SCREAMING BANSHEE !

Post by RhythMick »

Having finally finished the wiring and got the amp downstairs we took over the dining room table for a thorough test. A few minor (ahem) faults were tracked down. I'm blushing to admit that in my hurry to mount all 12 of the DC boards to the amp, 8 mounted to copper sheet with heatsinks, 7 of those I'd missed the isolation collar for the D44H11. all those had to be dismounted, cleaned, fresh thermal paste applied, etc. The measurements started to make sense after that, the bias trims and valve pair balancer circuits all worked as designed. Measured throughput for all 12 valves and got them within design spec, having compensated for my mains being almost 250v.

So - the moment arrives, I swap the dummy loads for an old pair of crap Sony bookshelf speakers and fire up. JESUS CHRIST WHAT THE HELL IS THAT NOISE? TURN IT OFF, TURN IT OFF!

High frequency squealing, very loud which I'm guessing is some sort of parasitic oscillation. I think I need help from more experienced peeps. The circuit diagrams are shown in this thread but I'll reattach them here. Actually those are slightly incorrect, they more closely reflect Geoffs amp which works perfectly. Summary of the circuit follows along with changes from Geoffs.

Apologies for the lengthy post and brain dump, but I don't know what part of the circuit might just trip someone into saying "ah yes, there's the problem". Wishful thinking ...

The amp signal path is as follows;

Balanced XLR inputs -> LL1544A input transformer primary
LL1544A secondary (bias applied to centre-tap) - I've attached the wiring diagram for the LL1544A.
LL1544A -> outputs to 01A grids (Geoffs amp uses 26 valves for input, I wanted to give 01A a try)
NOTE - no grid stoppers used initially and Geoffs doesn't have them - I've tried 820R carbon comp resistors as grid stops but the Banshee still rules
01A B+ applied to LL1621 PP interstage primary centre tap, then via 10R resistors to 01A anodes.
01A are heated by Andrews DC boards - 5v, 0.25A, across 2 x 200R Virtual Cathode resistors
Virtual Cathodes join at either end of a 2W 1k pot which is used to balance the DC operating points of the valves (this design is not in Geoffs), with the wiper forming the top of the long-tail for the pair.
The wiper connects to a top resistor, then a high inductance Lundahl choke (Tail Choke in my nomenclature), then a bottom resistor to B-. The B- point is also connected to mains earth for 0v reference. Description of the PSU below but in short each valve pair PSU is completely isolated from the other valve pairs - separate transformers, rectifier valves, smoothing chokes, caps.
The top resistor is used to set a minimum bias. A 100k pot is connected from the bottom of that resistor to ground, with the wiper providing the bias. (This separate 100k pot is different from Geoffs, where the bias pot forms part of the long tail itself - but in doing so conducts the full tail current and therefore is limited in the range of bias it can be allowed to provide, because of power dissipation. The 100k pot is separate and much lower current. This seems to work as per design and the circuit allows me both to set the bias and balance the 2 valves. I'll draw the tail circuit up properly later.)

The PSU provides a separate transformer secondary winding for each valve pair, so there are 6 mains transformers in total (not including the toroidals used for the DC heater boards). Every circuit is the same design, with components chosen to provide the smoothing required. Each circuit is...

Centre-tapped Mains Sec - slo-blo Fuses x2 - rectifier valve (double-diode) - Lundahl choke (CMR mode) - Mundorf Tubecap : UMBILICAL : Lundahl choke (CMR mode) - Mundorf Tubecap transformer centre-tap - 2 x 10R resistors - 2 x Anode - 2 x 2 x Virtual Cathode Resistors - Balancing pot - Top R - Tail Choke - Bottom R - Return to Final Choke return pin (0v connected at this point).

THINGS I'VE TRIED

I tried a different pair of 01A valves on the left channel, which made no difference.

I tried connecting an 820R carbon comp resistor as grid stop on the grids of the 01A. Geoffs amp (which uses 26 valves) has the grids connected directly to the output from the LL1544A. Adding the 820R made no difference to the Banshee.

I swapped back to the dummy loads, played a stereo 1khz sine wav through the XLRs and used my 2-channel picoscope to look at the grids of the 01A valves. One phase shows noise. I used persistence mode on the scope and it's clear that one phase is pretty steady while the other is broadly speaking a sine wave but with a much wider band and much more noise. I'm not really very experienced with scopes (the pico is my first, I may go for a proper one at some point). I'll try and get a screen-shot of this to show what I mean.

Next I connected the scopes to the input to the LL1544A. Same problem, one phase is all over the place. At this point I'm thinking what the key, but I guess if one of the valves is oscillating that could be fed back through the 1544A and appear on the scope ?

What to try next

I suspect I'm on the right lines with grid stoppers, but need to increase the value ? Miller capacitance from the datasheet indicates around 67pF, so to create a cutoff of 20KHz I'd need maybe 80K ? That feels high so I thought I'd check back here before going on.

I could change the design to the known config of 26 input valves. It requires only one resistor change per channel, plus of course adjust the trimmer on the DC boards to set to 1.5v instead of 5v. Easy to do and if I get nowehere with the 01As I'll try this. I don't like giving up though.

Get help from those fantastic people on the forum. Over to you guys - suggestions please ?
Attachments
image001.png
received_10213799926571105.jpeg
received_10213799925891088.jpeg
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#67 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by izzy wizzy »

Can you start with just the output stage? Say pull the rectifier valves and audio valves from stages 1 and 2. And then work your way back through the amp by reinstating stage 2, then 1. As each previous stage is pulled, you can use the IT as an input transformer.

Cheers,
Stephen
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#68 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

izzy wizzy wrote: Sun Dec 16, 2018 8:20 pm Can you start with just the output stage? Say pull the rectifier valves and audio valves from stages 1 and 2. And then work your way back through the amp by reinstating stage 2, then 1. As each previous stage is pulled, you can use the IT as an input transformer.

Cheers,
Stephen
Great suggestion thanks.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#69 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

One other thing in case it turns out to be important. It's 50/50 that I have the phasing out. I made the cables from the output transformers too long and when connecting them to the appropriate points I cut them before making a note of the colour-coded shrink on the end of the wire.

Now if there was global feedback in the amp I can see that would matter, in fact would be pretty much guaranteed to oscillate. However there is no global feedback whatsover, no connections across transformers at all.

With it being the output transformer I can't see that it would make any difference except that the outputs would be out of phase, which I can tell by scoping or listening to the stereophile test cd and simple to correct once I know.
User avatar
Nick
Site Admin
Posts: 15694
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#70 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by Nick »

Or you could use the scope starting with no B+ and check the signal on the first two grids and work forward. I wouldn't have thought those DHT's would need grid stoppers, thats normally for RF, a squeal is in the audio band so more likely to be feedback of some sort. One thing that occurs, what option is there to balance the valves at each stage?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#71 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by izzy wizzy »

Something I make the mistake of doing with the small Lundahl transformers is looking at the pin side where Lundahl id them from the top.

On Nick's advice, I did away with grid stoppers on my amp to what seems no detrimental effect and it's all trans coupled; relatively very short wires though.

Cheers,
Stephen
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#72 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

Nick wrote: Mon Dec 17, 2018 8:52 am Or you could use the scope starting with no B+ and check the signal on the first two grids and work forward. I wouldn't have thought those DHT's would need grid stoppers, thats normally for RF, a squeal is in the audio band so more likely to be feedback of some sort. One thing that occurs, what option is there to balance the valves at each stage?
Hi Nick - revised diagram showing the 45 OPT/VALVES/BALANCER POT & LTP Stack attached. The other valves are identical except for choice of choke and voltages.
PP Balancer with bias trim.PNG
The idea of the balancer pot is to adjust the resistance in each cathode leg (along with the bias trimmer) until the target DC current is reached. Works a treat in real life (unless that's where the Banshee lives of course). The pot for the 500R is a 10 turn 2W. I checked dissipation calculations very carefully. If (say) the pot is at 10%/90% in order to achieve balance then it's like having a 50R 0.2W on one cathode with a 450R 1.8W on the other cathode, each of which has 30mA through it. Dissipation is 23% of rating.
User avatar
Nick
Site Admin
Posts: 15694
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#73 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by Nick »

But how do you measure the current in each valve?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#74 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

Nick wrote: Mon Dec 17, 2018 11:15 am But how do you measure the current in each valve?
Across the 10R resistors on each anode. Which I now realise aren't on the diagram - updated below
PP Balancer with bias trim.PNG
RhythMick
Old Hand
Posts: 747
Joined: Fri Jun 23, 2017 12:24 pm

#75 Re: Finished and working - 12 valve all DHT PP 26-71a-45

Post by RhythMick »

izzy wizzy wrote: Mon Dec 17, 2018 9:26 am Something I make the mistake of doing with the small Lundahl transformers is looking at the pin side where Lundahl id them from the top.

On Nick's advice, I did away with grid stoppers on my amp to what seems no detrimental effect and it's all trans coupled; relatively very short wires though.

Cheers,
Stephen
Thanks Steve - yes I found it confusing at first (especially as some datasheets show 1-2-3-4, while some show 1-3-4-6 etc). Definitely got it now, and whats more these chokes were used successfully in the prototype. I write the pin numbers on the chokes. Those in CMR mode are all wired identically (in+ to pin 1, out+ to pin 2, in- to pin 3, out - to pin 4 - all numbered from top). Once it's in your head it's hard to go wrong
Post Reply