Unbalanced SEX, please help

Fred_P · 23704

0 Members and 1 Guest are viewing this topic.

Offline Fred_P

  • Jr. Member
  • **
    • Posts: 37
on: April 24, 2010, 10:21:54 PM
While using my SEX with the HD600s since a couple of weeks I had the impression, that the right channel is noticeably louder than the left. As I do not have a signal generator, I measured the output voltage on the 10 Ohm 5W loads using a 1kHz signal from a test CD.
The results confirmed my observation: 3.2V vs 4.2V at full volume.
The relative difference remains similar at different volume levels.
Reversing the channels and even the tubes did not change the situation.
The measured DC voltages are shown in the attachment.


Any suggestions please on how to balance it out?



Offline mrarroyo

  • Full Member
  • ***
    • Posts: 89
Reply #1 on: April 25, 2010, 02:45:20 AM
Are you sure it is the amp and not the source? Have you tried a different source or used your source with a different amp?



Offline Fred_P

  • Jr. Member
  • **
    • Posts: 37
Reply #2 on: April 25, 2010, 03:42:26 AM
Good idea, let's make sure there is nothing wrong with the source.
The values with a different source: 3.4V vs 4.3V

If it was the source, I would expect the output values to stick with the input channel when reversed, which they do not. The right channel always produces more voltage than the left, regardless of whether the source is connected in the correct or the opposite way.



Offline bernieclub

  • Full Member
  • ***
    • Posts: 114
Reply #3 on: April 25, 2010, 03:56:41 AM
I would check the tracking of the volume pot....maybe way off balance.

Bernie Zitomer


Offline JC

  • Sr. Member
  • ****
    • Posts: 485
Reply #4 on: April 25, 2010, 04:53:51 AM
If it were me, I think I would like to verify the DC Voltages on each channel with the inputs shorted, or at least with no signal present/volume control at minimum.  Then, see if the Voltages "follow the tube" when you swap tubes.

Jim C.


Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #5 on: April 25, 2010, 05:32:04 AM
If I read your post correctly the voltages on the plates and bias voltages of the two channels are different.  Were you reading  AC volts or DC volts?  I assume the readings of 3.2V and 4.2V is AC, right?  The other voltage measurements should be DC volts.



Offline Fred_P

  • Jr. Member
  • **
    • Posts: 37
Reply #6 on: April 25, 2010, 02:17:47 PM
Bernieclub,
The output voltages were measured at maximum volume setting, where the volume pots should not make much difference. But to make sure, I measured the resistance between input and middle of the pot at maximum volume, and it is 2 Ohms for both channels.

JC,
I measured the DC voltages with no input signal, and minimum volume setting of course.

Now I swapped the tubes and measured again.
The DC voltages did follow the tubes, with exception of the cathode voltage at the pre stage (pin 6)
Of course I did not measure the exact same voltages, and by "followed" I mean only
that the absolute value of the difference was roughly the same,
but the direction of the difference became the opposite.

The AC output voltages also followed the swapped tubes.

Grainger,
Yes, that is correct. First I measured DC voltages with no input signal.
Black figures are for left channel, red figures for right channel.
All values except for the output voltage are DC.
Then I connected the source, and set volume to maximum and measured the output voltage in AC.


OK, so can we conclude now, that the difference is only caused by the tubes?
If so, what is best way to handle it?

Before we go to the possible options let me mention something,
that might be of importance:
There are two visual differences between the 2 stock tubes.
1. the first has no marking at the bottom, the second has the following in red: "Zenith, made in USA"
2. when switched on, and looked from the top, the first one has a wire across that glows, while the second does not have this.
The second one with the marking gives the higher gain.

And now for the possible options:
1. order a matched pair (if available)?
2. order some more tubes and try replacing them until there is a good match?
3. mod any values to achieve the same gain with the existing tubes?
4. .....?



Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #7 on: April 25, 2010, 02:57:29 PM
Contact the Queen and ask.  A telephone call is always best.



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19845
Reply #8 on: April 25, 2010, 03:46:56 PM
The tube that made the red voltages looks pretty weak across the board (as long as those voltages follow the tube).  I'd give Eileen a call and she will send you a replacement.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline Dyna Saur

  • Full Member
  • ***
    • Posts: 75
  • Yippie-Ki-Yay....
Reply #9 on: April 25, 2010, 03:51:30 PM
The speaker connections should have zero Volts DC on them, and carry only AC audio signal content.
2 watts into 8 ohms would be around 4VAC maximum.  

The biasing scheme is reminiscent of the old Western Electric "split rail" power supply, and is essentially a hybrid between fixed and cathode bias.

IIRC, the negative bias rail on my two SEX amps runs around 17 VDC, +/- a volt. This is determined by the 620 ohm WW resistor, and the current flowing through the 6DN7's power triode.   The VA stage triode  of one tube could also be weaker than the other, the 6DN7s were never intended as being considered as "matched sets"  (or even as being used as audio tubes) so there will be unit-to-unit  variations.  HAving two tubes which look different is a dead give-away that something is amiss here...     The glowing wire across teh tube is just the connection between the filaments, this is essentially a series-wired, rather than parallel-wired filament arrangement.

Try:  

1: swapping the input cables, to see if the "unbalance" follows this. 
2: swapping the tubes, check if it follows the "suspect" tube. (You have alerady done this)
3: the volume control could be mis-tracking. If you measure its resistance, it may not be exactly 100K    ohms, but the resistances should be within 5% or so of each , and the wiper-to-ground resistances should be similar.

HTH

/ed B in NC

Builder and owner of 1 pair "stock"  Paramounts, 1 pair of upgraded  Paramour IIs, two SEX amps (one upgraded, one bare essentials), two FPIII (one stock, one 6SN7 and step attenuator modified) , one somewhat modified  FP II,  one stock  Seduction, and two Quickies (and coming soon,  one Crack).  
« Last Edit: April 25, 2010, 03:54:06 PM by Dyna Saur »

ed brown


Offline JC

  • Sr. Member
  • ****
    • Posts: 485
Reply #10 on: April 25, 2010, 04:14:53 PM
It seems like the consensus is that one of the tubes is not emitting as strongly as the other.  The way to be sure of this would be to have the tubes tested on a tester that could put a real number to it, but not many people have access to a tester like that.

Some vendors offer matching services for a nominal fee, but it pays to ask how they go about matching them.  All techniques are not the same.  Also, I'm not sure how many are accustomed to matching 6DN7s, since they are a dissimilar dual triode designed for service in the vertical deflection section of B&W TVs.  IOW, whomever is matching the tubes will need to know what he is doing, and have to know how to adjust his test set-up for valid testing of this tube.

So, yes, a call to Bottlehead might well be in order for advice.  They may be able to find you a pair from their stash, or at least give you some suggestions.  If nothing else, they have no doubt run through a few more 6DN7s in their history than the average high-end audio emporium!

Just FYI, the point of suggesting the measurements I did was to try to determine if your discrepancies had to do with differences in the tubes, or differences, somehow, in the circuits between the two sides.  Sounds like you have narrowed it down to the tubes, which is certainly the easier to address of the two possible outcomes!

Jim C.


Offline Fred_P

  • Jr. Member
  • **
    • Posts: 37
Reply #11 on: April 26, 2010, 04:05:48 AM
BIG THANKS...
for the support and advice to all of you.

I mailed the Queen about the issue, and asked for her help.
« Last Edit: April 26, 2010, 04:08:18 AM by Fred_P »



Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #12 on: April 26, 2010, 06:29:30 AM
The voltage reading you get on the low channel is 25% lower than the stronger channel.  You must have a bad tube.  This is a much larger difference than normal tube variances.  Well, maybe you are at the three sigma on either end.



Offline Fred_P

  • Jr. Member
  • **
    • Posts: 37
Reply #13 on: May 12, 2010, 12:46:06 PM
The replacement tubes have arrived 2 days ago, not a bad speed from US to Singapore.
I replaced them at once, and measured the output voltages again:
Now it is 4.4V and 4.7V -> 0.57dB which is totally OK.
(It used to be 3.2V and 4.2V -> 2.36dB before, with the first set of tubes)

So the issue is solved.
Thanks again to everyone for their support,
and most of all to the Queen for her quick help.

One side effect may be worth mentioning is that the amp
sounds noticeably different with the new tubes, which are of a different brand.
It is funny how we discuss all the possible upgrade options and never mention trying different 6dn7s.
May be I just overlooked this topic so far, but if it is still unexplored it might be worth a separate thread.


 
 
« Last Edit: May 12, 2010, 01:34:41 PM by Fred_P »



Offline JC

  • Sr. Member
  • ****
    • Posts: 485
Reply #14 on: May 12, 2010, 03:03:22 PM
Hey, glad you wrote back after the new tubes got there; it is always instructive to hear the end of one of these trouble-shooting stories!

As I recall, Doc once indicated a preference for the 6DN7 in the SEX, but there hasn't been a great deal of chatter about it.  Maybe that is due to the nature of the tube.  The 6DN7 was originally produced for B&W TV service, and as far as I know, no one produces a current model.  So, the usual audio-board chatter about "which is the best" may just not exist.

Still, I would be very much interested in hearing your impressions of your "new" tubes, starting with any data you might have on the brand, appearance, etc.
 
Probably, it would be good to wait until you get a few hours of operation logged on them, since there is usually some break-in period with the tubes themselves during which they may undergo some sonic changes.

Jim C.