GoodSoundClub - Romy the Cat's Audio Site

11-29-2011

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	181
Post ID:	17445
Reply to:	17444

Filament considerations

Alex, have you killed your ham finally?
What was the source?

AlexBerger wrote:

Is the filament ground connected to the common ground?

Of course! Otherwise the fixed bias would not work. Moreover it's connected
directly at the V2 to minimize the loop in the V2's grid created by the connection and the voltage dividers.
BTW, I tried shorting this grounding point to the input stage ground with no result.

Actually those voltage dividers and my mess around V2's grid were my first suspects, but apparently
shorting the grid to gnd with a 0.47u cap gives relatively small hum reduction, smaller than shorting the plate.

AlexBerger wrote:

Is the filament ripple is small?

It is small already at the first cap behind the input choke@200mVpp.
Cheers,
Nset

Cheers,
Jarek
STACORE

11-29-2011

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	182
Post ID:	17446
Reply to:	17443

Air caps?Buffer?

Romy the Cat wrote:

Did you try just for test to increase the caps value that short the plate to ground? Will it amount and the character of the noise change?

Just like one would expect: 0.1uF shorting cap gives approx 4x more ham than a 0.47u one.
The character is the same--quite a nice sinus (shorting cleans some HF obviously)

Romy the Cat wrote:

If I were you then I would use elimination techniques. Get rid of the output buffer and see if the noise still there. Replace the air caps in feedback with mike or alike and see if noise is there. The air caps are in a way antenna and they suck anything from air.

Ok, this gave some info:
0) disconnecting the air caps completely reduces 50Hz by some 10dB, but the resulting noise is very ugly, irregular, high peak-to-peak (a bit higher than with the loop closed), the spectrum suggest a white noise; the RIAA eq is out so all the dirt is present in the full glory; where the fuck does it come from? a cheap chinese 12AX7 noise?

1) shorting the buffer (either the plate or the cathode to gnd) or the output changes absolutely nothing
2) but taking the buffer out completely and taking the signal from the second stage's plate via 0.47uF reduces the ham by some 20dB;
the noise is quite "nice", the noise floor drops by some 15dB or so
3) simultaneously disconnecting the air caps from V2's grid reduces ham by additional 3-4dB

Moreover, if the scope is at the output, the phono switched off, touching the 330pF gives a super strong ham
(touching the 150pF has a very-very mild effect in turn). Caps as antena ok, but why the buffer?
I keep scratching my head...

Cheers,
N-set

Cheers,
Jarek
STACORE

01-14-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	183
Post ID:	17709
Reply to:	17422

Exorcisms

Paul, after a month of vacations I've tried you advice:

Paul S wrote:
If you can do it, try isolating both cases/chasses from your "ground plane(s)". Use one inclusive "ground plane" for both the PS and the gain/RIAA sections, and ground this "ground plane" to the house "neutral" wire. Do NOT connect any shielding to the "ground plane". Then ground the cases/chasses, all internal shields, the IC shields, TT/motor, arm and step-up ground lugs to a special, dedicated ground rod (not the house neutral or ground!). This special ground rod is the "bleeder" I keep referring to.

but with the chasses floating (I've realized I have a two-pin instalation in my apartment and no visible pipes
to connect to). No effect.
I continued the exorcisms with "Grounding and shielding techniques in instrumetation" in one hand
and a scope in the other. After some less idiotic arrangement of transformer screens and watching the AC polarity
I got only a minor 5dB reduction in 50Hz ham, but it's still audible.
The only quiet condition still remains as before: i)the signal earth (the Cu plate) connected to the PS chassis
ii) the signal box shielded.
All the rest doesn't work, including taking the input stage out completely, disconnecting air caps, etc etc. Shit...

Disconecting the signal earth and leaving it floating results in an ugly 100Hz pickup, related to the
mains Xformers/rectification. It resists all the exorcisms, it only wants the Cu plate to be connected to
the PSU chassis, driving me crazy.
Sorry for spamming you with all this...
Cheers,
N-set

Cheers,
Jarek
STACORE

01-15-2012

Post does not mapped to Knowledge Tree

Paul S
San Diego, California, USA
Posts 2,658
Joined on 10-12-2006

Post #:	184
Post ID:	17710
Reply to:	17709

Floating?

N-set, I am confused by this/your most recent post. If you are saying you have discovered a "quiet condition" along with the desired curve and gain, then you are done, no? If it works, it works; good for you; be thankful; leave it alone; congrats!

If you are still looking for a solution, surely you realize that all my suggestions require a special, dedicated ground to use as a "bleeder"; there are no floating "variants".

If you need a ground [path], seek and ye shall find.

Best regards,
Paul S

01-22-2012

Post does not mapped to Knowledge Tree

steverino
Posts 367
Joined on 05-23-2009

Post #:	185
Post ID:	17731
Reply to:	17415

Twitchy signal tube and amp interactions

While I have found the Amperex globes act more consistently than the Bugle Boys I wouldn't call the Bugles uniformly bad. I wonder if they just don't match up with current amps as uniformly as the later Globes? I have a stockpile of GE longplates from the late 50s and early 60s. They are very good tubes IMO a bit analytical but have clean treble and overall transparency. While they work in most amps they simply can't function in a Herron phonostage that I have. They sound different there and actually started to make a bit of noise after a month or two in there. I though they were defective until I stuck them into a different amp and they worked fine. On the other hand Amperex and Mullards from the late 60s and 70s work fine there as do amazingly enough the 12AX7 LPS from Sovtek (the only Russian signal tube that sounds at all satisfactory to me as long as they are "picks of the litter"). I'm sure there are a lot of amp and tube interactions that are never officially "diagnosed".

Romy, Have you ever wondered whether the Parisian might have foisted that box off on you knowingly? That seems like a suspiciously low price for 180 Amperex tubes.

01-22-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	186
Post ID:	17732
Reply to:	17710

Toilet ham

Paul S wrote:
N-set, I am confused by this/your most recent post. If you are saying you have discovered a "quiet condition" along with the desired curve and gain, then you are done, no? If it works, it works; good for you; be thankful; leave it alone; congrats!

I somehow feel that my quiet solution is a half-ass solution, i.e. the need of a complete shield.
Redoing some chassis work/connections (my assumption that all the chassis is on the same potential was actually wrong)
did quiet the things a bit.

Paul S wrote:

If you are still looking for a solution, surely you realize that all my suggestions require a special, dedicated ground to use as a "bleeder"; there are no floating "variants".

Yes, I understand that.
Yet, there is something I don't get in your method: while the chassis is on a dedicated ground and the signal
ground on the line neutral, there will be in general a potential difference between those two points. The signal ground-to-chassis
capacity closes the loop. Why not connecting both the chassis and the signal zero to the same rod in a star-like fashion?

Paul S wrote:

If you need a ground [path], seek and ye shall find.

I think I've found it--in the toilet!! A small piece of cold water tubing. Perhaps it'll do the trick.

Cheers,
N-set

Cheers,
Jarek
STACORE

01-23-2012

Post does not mapped to Knowledge Tree

Paul S
San Diego, California, USA
Posts 2,658
Joined on 10-12-2006

Post #:	187
Post ID:	17733
Reply to:	17732

The Shield

N-set, I don't know much about the theory, but maybe you have focused on "uniform ground potential" while neglecting the effects of EMF? (Not to mention the EXTREME difficuty of maintaining a "single", non-reactive "ground path" in the first place...) Again, I keep the chassis/shield(s) and the signal grounds electrically separate just because it reduces noise. Yes, the differences in potential are there in both cases; but for me the bleeder can be arranged to address the audible part of the problem. Again, the signal ground is the line neutral, and the chassis/shield goes to the dedicated ground (and NOT the line ground). And remember, you want to solder or tightly clamp your bleeder to clean, solid, well-grounded copper, not chrome, nickle, etc.

By the way, FWIW, it may be the case that your AC system includes a ground wire, after all, despite the 2-prong plugs/outlets. In many older systems a ground wire is stubbed inside outlet, junction and switch boxes; and in some older systems the ground wire is stubbed at each box, but it is "made up" just outside the box. Of course, the problem of line noise via inductance (if nothing else..) is usually present, in any case, when the house ground is considered, as such.

Best regards,
Paul S

PS: You might well find it best to 2-prong your amps and to only "bleed" the front end...

01-23-2012

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,159
Joined on 05-28-2004

Post #:	188
Post ID:	17734
Reply to:	17731

It is hard to say.

steverino wrote:
…. as do amazingly enough the 12AX7 LPS from Sovtek (the only Russian signal tube that sounds at all satisfactory to me as long as they are "picks of the litter").

Yes, this is exactly my finding 12AX7-LPS are surprisingly better sounding tube then the rest of Russian 12AX7 crap. Russian very much like to substitute the 12AX7 with 6N2P (slightly lower gain and permanently paralleled filaments). I absolutely hate the 6N2P (and 6N1P - western equivalent is ECC88/6BQ7A). They are absolutely dear in sound, even if you get some very early production batches. The 12AX7-LPS is good but the still in my view very much loose to more sophisticated in sound for instance Telefunkens. The 12AX7-LPS are dryer, less colorful and more predictable sort of speaking. The 12AX7-LPS bass is surprisingly rich but it more sounds like dry coughing. I like when coughing is deep with moisture and blood sort of speaking, of cause I am talking about bass….

steverino wrote:
…. Romy, Have you ever wondered whether the Parisian might have foisted that box off on you knowingly? That seems like a suspiciously low price for 180 Amperex tubes.

I do not know, the tubes faking are very separate subject and I do not want to be involve into thinking about it, partially because I do not have a lot of knowledge about it. I did but the look like Tele with long plate that has all necessary attribute of being original but they sound very bad, Were they very well make fake or it was the bad sounding batch? It is hard to say.

The Cat

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

01-23-2012

Post does not mapped to Knowledge Tree

steverino
Posts 367
Joined on 05-23-2009

Post #:	189
Post ID:	17735
Reply to:	17734

Fake bugle/teles

Romy said "Were they very well make fake or it was the bad sounding batch?"

Some of the fakes are quite sophisticated supposedly. I thought that the washability of the label was one of the better ways to test for them. The fakes tend to have sturdier labels printed on the tube surface that don't wash off easily.. It's also possible that these were simply a collection of discards.

01-23-2012

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,159
Joined on 05-28-2004

Post #:	190
Post ID:	17736
Reply to:	17735

Never again.

Yes, the fragility of the marking is a very good sign on some German tubes. Still, I do not see why the people who would like to fake them nowadays could not reproduce the washable marking. They have learned how to make the diamonds under the bottom of the Tele’s pins, so to make fragile paint is not really a difficult task. The signal tubes are not really a big deal. If you once catch a good butch and buy a few 12AX7 for phonostage then they most likely last to the end of your life. The typed like 437A or 417A do die relatively fast but 12AX7 with it’s 1mA on plate will live forever. So, I feel it worth to pay expensive piece once in life for good new 12AX7 and do not worry about it ever

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

02-01-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	191
Post ID:	17773
Reply to:	17733

Bingo!

Paul S wrote:
N-set, I don't know much about the theory, but maybe you have focused on "uniform ground potential" while neglecting the effects of EMF? (Not to mention the EXTREME difficuty of maintaining a "single", non-reactive "ground path" in the first place...)

That was it, Paul! I've ass-u-med that the all points of the chasses will be at an equal potential. How wrong (maybe right at HF)!
I've rearranged connections within the CHASSES in a starlike-fashion: now every chassis has it's
own "chassis bolt" where different parts of the chassis/screens etc meet.
The signal ground (big Cu plate) is connected to the signal chassis bolt, which is
the grounding terminal for the TT. The connection to the PS chassis and the mumetal cases of the SUT's
also meet there (the upper right corner):

With all in place, the SUT's connected, the chasses closed, the inputs shorted
and the PS far away the ham is gone. Huray!
The latter constarined will be addressed by closing my mains transformers in steel coffins and also possibly
the SUT's--there is enough space. The earthing--my lovely Cu toilet tube does not seem to a have an
earth connection-there may be some plastic part somewhere, so I'm floating...
The next step is to cook a decent TT-pre cable and make that interface ham-tight.

Cheers,
Nset

Cheers,
Jarek
STACORE

02-08-2012

Post does not mapped to Knowledge Tree

FJC Corpas
Madrid-Spain
Posts 4
Joined on 02-08-2012

Post #:	192
Post ID:	17797
Reply to:	7480

Total current per channel

Hello all, I´m new in this forum although I have been following it for a long time. I´m building a phono based in this design and I would like to know what´s the current draw per channel. I think is 5mA but I would like to be sure.
Thanks and Rgs, FJC

02-08-2012

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,159
Joined on 05-28-2004

Post #:	193
Post ID:	17798
Reply to:	17797

Stay with capacitive filtration for B+ for this init.

Yes, 5mA is about right. Each 12AX7 eats around 1mA (some time a few fraction of mA more). So, if you have 3 tubes for 1.2-1.2ma each then it will, be around 3.6ma or for all intended purpose you can call it 5mA most likely you ask this question become you want to do what I did – the input choke and bleeding resistor. Well, I am not sure that it is worth direction. I think the RCRC filter works juts as fin in this configuration. I did not detect any advance (for THIS current) when I went for LC and I would advise to state with regular and simple capacitive filtration.

The Cat

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

02-08-2012

Post does not mapped to Knowledge Tree

FJC Corpas
Madrid-Spain
Posts 4
Joined on 02-08-2012

Post #:	194
Post ID:	17799
Reply to:	17798

I like iron

Yes, this is I was asking for. Then quiescent current per tube is +or- 1.2mA, total 3.6mA through 6K resistor.
As you stated, the power supply you built is like put a jet engine to a Honda Civic, but I have the irons and I like the idea. If necesary I can put a pair of wings to the Honda Smile

Joking apart, theory says LC filter is indicated when big currents are required but I like the way the current flows in a LC filter, smooth and steady.

FJC

02-08-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	195
Post ID:	17800
Reply to:	17799

Choke ringing

FJC Corpas wrote:
... but I like the way the current flows in a LC filter, smooth and steady.
FJC

...modulo ringing of the choke at switching! the current waveform in the input cell of my PSU
is an incarnation of satan

PS if you want to have a separate supply resistors for two channles they should be 12k, 6k is for both channels.
you can have a look at the actual schematics of my copy.

Cheers,
Jarek
STACORE

02-08-2012

Post does not mapped to Knowledge Tree

FJC Corpas
Madrid-Spain
Posts 4
Joined on 02-08-2012

Post #:	196
Post ID:	17801
Reply to:	17800

Ringing

Well, I normally use tube rectifier with my LC filters and, I don´t know exactly why, I haven´t had ringing problems at switching. I think could be tubes switch in a gentler way than silicon. I recommend you use an appropriated snubber circuit across the choke. Ringing and other "satanic" glitches can be due also to inappropiate transformer (or choke) rating. In this sense center tapped tube rectifiers help to avoid core saturation in the PS transformer.
FJC

02-08-2012

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,159
Joined on 05-28-2004

Post #:	197
Post ID:	17802
Reply to:	17800

Ringing is a BS reason.

In my view the Ringing of chokes is a BS reason that was invented by manufacturers who do not want to spend money for input chokes. Any book has formulation to calculate the critical inductance and if you are meeting the conditions then it will be no ringing. You do not even need to calculate anything – juts put a scope to the diodes before the choke and bleed the current unit the voltage drop after the diode shutting down voltage does to zero. It’s it.

The Cat

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

02-09-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	198
Post ID:	17803
Reply to:	17802

Choke ringing again

Romy the Cat wrote:
In my view the Ringing of chokes is a BS reason that was invented by manufacturers who do not want to spend money for input chokes. Any book has formulation to calculate the critical inductance and if you are meeting the conditions then it will be no ringing. You do not even need to calculate anything – juts put a scope to the diodes before the choke and bleed the current unit the voltage drop after the diode shutting down voltage does to zero. It’s it.

The Cat

Not that it would be so important but, ringing of a choke (or actually choke-transformer-1st cap tank) has nothing to do with the critical inductance needed to maintain the constant current. What heavy bleeding does is exactly lowering the Q of the
parasitic tank which oscillates excited by the switching.

Cheers,
N-set

Cheers,
Jarek
STACORE

02-09-2012

Post does not mapped to Knowledge Tree

Romy the Cat

Boston, MA
Posts 10,159
Joined on 05-28-2004

Post #:	199
Post ID:	17804
Reply to:	17803

What ringing?

N-set, can you explain what you are saying? What kind ringing are you implying? Are you taking about mechanical ringing of the wire or about the after pulse ringing?

"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche

02-09-2012

Post does not mapped to Knowledge Tree

N-set
Gdansk, Poland
Posts 617
Joined on 01-07-2006

Post #:	200
Post ID:	17805
Reply to:	17804

Ring-ring

Romy the Cat wrote:
N-set, can you explain what you are saying? What kind ringing are you implying? Are you taking about mechanical ringing of the wire or about the after pulse ringing?

I referr to the after pulse ringing. The parasitic tank formed by input
choke inductance + its parasitic capacitance+transformer leakage+parasitic cap of the transformer winding
gets excited at diode switching (the current momentarily falls so the big inductance of the choke produces a
voltage glitch,trying to maintain the current flow; this glitch excites the mentioned tank).
I observed in my PSU HT choke ringing at some 30kHz, which penetrated the closeby filament supply (the first cell).
Moreover I have an improvised, very simple laptop "scope" with a floating ground, so I could also see the current
waveform in the input choke circuit. You can see it in the post #172, red trace. Perhaps I'll try to bleed the supply heavier
and see if it damps the heavy oscillations.
Cheers,
N-set

Cheers,
Jarek
STACORE

	Target	Threads for related reading	Most recent post in related threads	Forum	Replies	Views	Started
»	New	Where are our good phonostages?..	Omnigon Tubes...	Analog Playback Forum	61	665456	05-31-2004
»	New	The Expressive Technologies SU-1..	“too bright” or “resolution” or “details” with SUT prim...	Analog Playback Forum	33	376156	12-30-2004
»	New	EAR 834P Modification Guide..	The cap will not change volume in the pass band...	Analog Playback Forum	45	639217	02-09-2006
»	New	My Analog Playback: the fat lady has sung..	My analog setup update....	Analog Playback Forum	9	119486	04-04-2006
»	New	Phono stages with SU-1..	SU-1...	Analog Playback Forum	4	66080	11-23-2007
»	New	Chasing utopian better phono interconnect...	Did I miss something?...	Analog Playback Forum	6	110895	06-05-2008
»	New	Buying a last cartridge...	Lucky you...	Analog Playback Forum	80	810368	09-05-2008
»	New	EAR 834P mods..	The cap will not change volume in the pass band...	Analog Playback Forum	45	639217	11-04-2008
»	New	AMR PH-77: just another phonostage or more?..	Oh, yeah... the sound of the Thing Itself......	Analog Playback Forum	11	132387	07-05-2009
»	New	An interesting Russian pnonocorrector...	Uber-tweeky phono topologies deconstructed...	Analog Playback Forum	9	99319	03-01-2010
»	New	Expressive Technologies SU-1 and cartridge output and i..	MV and Ohms...	Analog Playback Forum	2	33631	07-07-2011
»	New	How to run MM-type cartridge into MC phonostage?..	Quite interesting....	Analog Playback Forum	6	67371	11-13-2011
»	New	The tales of two phonostages: Allnic and 834PT...	The tales of two phonostages: Allnic and 834PT....	Analog Playback Forum	0	23388	12-21-2011

	Target	Threads for related reading	Most recent post in related threads	Forum	Replies	Views	Started
»	New	Where are our good phonostages?..	Omnigon Tubes...	Analog Playback Forum	61	665456	05-31-2004
»	New	The Expressive Technologies SU-1..	“too bright” or “resolution” or “details” with SUT prim...	Analog Playback Forum	33	376156	12-30-2004
»	New	EAR 834P Modification Guide..	The cap will not change volume in the pass band...	Analog Playback Forum	45	639217	02-09-2006
»	New	My Analog Playback: the fat lady has sung..	My analog setup update....	Analog Playback Forum	9	119486	04-04-2006
»	New	Phono stages with SU-1..	SU-1...	Analog Playback Forum	4	66080	11-23-2007
»	New	Chasing utopian better phono interconnect...	Did I miss something?...	Analog Playback Forum	6	110895	06-05-2008
»	New	Buying a last cartridge...	Lucky you...	Analog Playback Forum	80	810368	09-05-2008
»	New	EAR 834P mods..	The cap will not change volume in the pass band...	Analog Playback Forum	45	639217	11-04-2008
»	New	AMR PH-77: just another phonostage or more?..	Oh, yeah... the sound of the Thing Itself......	Analog Playback Forum	11	132387	07-05-2009
»	New	An interesting Russian pnonocorrector...	Uber-tweeky phono topologies deconstructed...	Analog Playback Forum	9	99319	03-01-2010
»	New	Expressive Technologies SU-1 and cartridge output and i..	MV and Ohms...	Analog Playback Forum	2	33631	07-07-2011
»	New	How to run MM-type cartridge into MC phonostage?..	Quite interesting....	Analog Playback Forum	6	67371	11-13-2011
»	New	The tales of two phonostages: Allnic and 834PT...	The tales of two phonostages: Allnic and 834PT....	Analog Playback Forum	0	23388	12-21-2011