ELK Viking 50 bias problems?

[Madison]

Repairing it.
Replaced first dropping resistor, power tube coupling caps,/grid resistors, bias cap, power tubes.

These items got it stable enough to play.
I sounds okay but the power tubes are HOT! (not red plating)

This amp has a non adjustable fixed bias FWIU.
I am reading 70 mA with my bias checker.
That can't be right??
Should be about half I'd think.

I've been messing around with the two resistors (1.5K and 75R) on the bias and all I can do it just make it red plate by increasing the 75R one.
Cn;t seem to get it to do down.

Also, I decided to hook up a pedal to the reverb/tremolo.
The switch for the vibrato, it only adjusts the speed,............it doesn't kill the it.
Bad caps galore?
It works okay until you get above 7 on the speed pot and dies.
It's not set up like a typical Fender providing for more confusion.

I don't have a schematic.
Did a search and came up with nothing.

Any ideas?
Thanks all.

[Madison]

2

[Madison]

3

[Madison]

4

[Willabe]

How's your finger doing?

      Brad   

[Madison]

How's your finger doing?

      Brad   

Thanks for the concern.
4 more weeks, wires come out.
Hurts like heck; keep bumping it and try to use it when gripping items....... about ready to tell them to pull the damn wires now.
Have a very local show coming up in a week.
Still going to attempt to play classical tune(a Bee Gee song.....kill me, kill me now!) and Oloha Oe on steel guitar.

[Willabe]

Be patient, wait on having them pull the wires. It will seem like forever but it will be over soon. You want it to heal as best as possible for full normal use again.

I'll take some heat for this but, I like some Bee Gees songs. They could write a hook and knew the craft of song writing well. Just a matter of taste in music. In high school in the 70's we hated them because of the disco thing but their early songs are very different.

I heard them in the 60's on AM radio and my brother had a greatest hits album from that time that was good. OTOH, if Stayin Alive comes on the radio I don't mind listening to it. IMO it's a well crafted song and put together well in the studio.



                   Brad       

[jjasilli]

Can you draw a schematic of the bias supply.  Don't forget that because the bias circuit draws near zero current, it has to be a voltage divider.  The series resistor(s), by themselves, do nothing for bias voltage attenuation, because there's not enough current to cause a voltage drop across them.  (Though they do isolate bias supply filter caps).  To drop the tube's current draw you need more bias voltage on the grid - G1.  To do that, increase the value of the shunt resistor, so less bias voltage bleeds to ground in the divider circuit.  Personally,I would use Doug's trick:  temporarily install a pot wired as a variable resistor on the top end of the shunt resistor.  Increase the pots resistance until you get power tube diss right.  The hard wire the shunt resistance accordingly.

[Madison]

Can you draw a schematic of the bias supply.  Don't forget that because the bias circuit draws near zero current, it has to be a voltage divider.  The series resistor(s), by themselves, do nothing for bias voltage attenuation, because there's not enough current to cause a voltage drop across them.  (Though they do isolate bias supply filter caps).  To drop the tube's current draw you need more bias voltage on the grid - G1.  To do that, increase the value of the shunt resistor, so less bias voltage bleeds to ground in the divider circuit.  Personally,I would use Doug's trick:  temporarily install a pot wired as a variable resistor on the top end of the shunt resistor.  Increase the pots resistance until you get power tube diss right.  The hard wire the shunt resistance accordingly.

Here is a drawing of the bias system.
I am a little confused about where I might insert a pot, if it is that easy?

Willabe,
About the same way I feel about them.
I have to play something familiar for the audience at this show.
Every time I play "How deep is your love"  I don't announce what it is.
On the intro, people at first the get a really confused look on their face and yell "the Bee Gees!!" as soon as the first lyrics "I know your eyes.............."
Like an expression of "are you freaking kidding me!?".I look at them like " I sheet you not!"
People seem to smile in any event.
Fun song to play BTW!(much rather play it than listen to it!)
Lots of chord changes.
Hopefully I can pull it off.

Thanks

[stratele52]

Madison look at any Fender Blackface amp schematic, you'll see how to wire a pot for bias . Bias circuits are ( mostly ) all the same in any amp.

Here ;

http://www.thevintagesound.com/ffg/

[Madison]

Madison look at any Fender Blackface amp schematic, you'll see how to wire a pot for bias . Bias circuits are ( mostly ) all the same in any amp.

Here ;

http://www.thevintagesound.com/ffg/

Still curious why this one doesn't work.

Beyond that.
Couple of things-

I don't think 25V is going to supply bias to a couple of EL34s?
Usually 70V ........I think.

I'd have to go with a Marshall type bias set up then?

Thanks!

[stratele52]

25 AC is not enough to have DC bias voltage around minus 35 DC volts bias voltage.

Do you read 25 AC with no load ? No bias circuit connected ?

Yes marshall tybe bias circuit can bring you more voltage . I use it when power transformenr have no bias winding

[jjasilli]

Yes, the tube charts say about -35 V_g1: http://www.mif.pg.gda.pl/homepages/frank/sheets/129/e/EL34.pdf

I agree, with Stratele52, to check the source voltage of the bias supply at the PT tap with the bias circuit disconnected.  However, it's not likely that the amp was designed with the wrong bias tap voltage; more likely that the bias components drifted in value; or that the bias circuit was re-built incorrectly in the past.

The schematic you drew is "upside down", which is fine, but can be confusing because it disguises the voltage divider which is at the corner the 75R series resisitor/100K shunt resistor.  Usually a voltage divider is shown with the series resistor on the upper, or top, leg of the drawing.  The test pot could be placed on either end of the shunt resistor. 

[jjasilli]

Further thoughts:  is there evidence that this bias supply was worked on or mucked up in the past?  I'm  thinking maybe this:  the PT's bias tap (unloaded) is about 50 VAC, but the half wave bias rectifier is cutting this in half to about -25VDC, which is not enough.  If so, then ways to go:  1) use a bridge rectifier (backwards!) in the bias circuit which will result in plenty of (-)rectified bias voltage, like -50VDC or more.  Measure the resulting voltage & divide it down to about -35VDC; then fine tune to taste.  2) As suggested in other posts, source the bias voltage off the HT line. Follow known schematics.

[Madison]

Further thoughts:  is there evidence that this bias supply was worked on or mucked up in the past?  I'm  thinking maybe this:  the PT's bias tap (unloaded) is about 50 VAC, but the half wave bias rectifier is cutting this in half to about -25VDC, which is not enough.  If so, then ways to go:  1) use a bridge rectifier (backwards!) in the bias circuit which will result in plenty of (-)rectified bias voltage, like -50VDC or more.  Measure the resulting voltage & divide it down to about -35VDC; then fine tune to taste.  2) As suggested in other posts, source the bias voltage off the HT line. Follow known schematics.

I'll check unloaded voltages.
I don;t think this has been mucked with.
Resistors, caps, diode look period correct, but who knows.
I'll report.

[Madison]

No load.

 PT=25VAC

after bias rectification -34DC.

Hmmm.

[PRR]

The bias supply was *just-barely* spot-on, when it was built, for older wall-voltages, for the specific tubes originally installed, and the warranty you got (slim to none).

With higher voltages, and the wide spread of today's tubes, it's almost-enough. You could leave it that way; carry spare tubes and sweat-proof shirts.

Or voltage-double the winding and divide down to get happy tubes.

[Madison]

Can I put a Fender type bias (ie aa165) and source it off one side of the HT tap like a Marshall?
I have it all wired up right now like that off the 25V tap with the same -70 results.
Would like to move one wire over to the HT.
Will I blow up something?

[PRR]

If you have -70V, why do you want to change?

It probably needs -35V to -45V.

[jjasilli]

Yes. I'm not sure how you're getting -34 or -70.  But IMHO -34 is good enough.  OTOH I like the elegance of PRR's solution of a voltage doubler.  And, yes alternatively you could tap off an HT leg.

[Madison]

I am getting -34V where the two 220Ks meet.
I am getting 70mA on my bias checker.(Hoffman designed)
So I think the 70mA is high right?
I thought that should be about 35mA?

Pardon, I hope this is more clear.

[PRR]

I'm confused voltage or current.

The doubler will let you tune-to-taste.

[Madison]

I'll just go ahead and try the doubler now.
Thanks.

[jjasilli]

I am getting -34V where the two 220Ks meet.  You need to measure bias voltage on the lug of the tube socket's signal Grid pin - G1.  However it should still be about -34V on the pin, even after the 220K resistor. (Because there's no current draw to cause a voltage drop across that resistor.)


I am getting 70mA on my bias checker.(Hoffman designed); So I think the 70mA is high right?  Is this for one tube or two?  Back to the tube chart I posted earlier.  At idle two plate should draw 2 x 55mA = 110mA (plus screen draw).  For 2 tubes, 70mA is cold.  For one tube it's real hot.


Pardon, I hope this is more clear. Not yet.  

[Madison]

>>I am getting -34V where the two 220Ks meet.  You need to measure bias voltage on the lug of the tube socket's signal Grid pin - G1.  However it should still be about -34V on the pin, even after the 220K resistor. (Because there's no current draw to cause a voltage drop across that resistor.)


-34V before the 220Ks. and -26V after.
-26V on G1 pin

Have already tried replacing the 220K resistors, no change.

Bias checker is like this; one tube at a time.
70mA, and they are hot!
http://www.el34world.com/charts/biascurrent.htm

Going to try re soldering the pins.Something I haven't done yet.Maybe the original solder has worn/cracked.(shot in the dark)
or
Bad OT?

[stratele52]

>>
-34V before the 220Ks. and -26V after.
-26V on G1 pin

You must have -34 volts on tubes's pin 5 . There is a problem there with your bias voltage .

 Remove your EL34 and read your -34 volts at pin 5 on both tubes . If you read it , tube (s) grid is probably shorted (? ) . Put new tubes.

If you don't have your -34 with tubes removed , something is short to ground there OR bad solder or tubes contact

[Madison]

Well I guess the "good news" is, it's the same with the tubes removed= -26 on pin 5.
These tubes were new, they are pretty worn in at the moment but I don;t think they are shot....yet.
I have no more at the moment.

With experimentation.
I could get the bias to go to 30mA when I switched it to a cathode bias.
Problem was, I don't understand this tremolo, plus I got next to no output.
It doesn't jiggle the bias at the normal Fender spot on the 220K resistors.

[jjasilli]

Tremolo!!!      

If this is power tube bias tremolo, then temporarily disconnect the output of the tremolo circuit.  Get the bias circuit working right.  Then re-install the tremolo and de-bug that if necessary. 

It does appear from your description that the tremolo circuit is malfunctioning, and it may be bleeding away some of your bias voltage.  That's why it should be disconnected: to remove variables from affecting the main bias circuit. 

Also, you don't have to understand the tremolo circuit.  Just do a "dumb" rebuild.  If you make both a schematic & a layout diagram of the tremolo circuit you can cross-check for accuracy in the rebuild.  Do not use a meter to measure component values, because something is likely to have drifted in value.  Instead use the coding on the body of the component to get the stock value.  If the coding is a mystery and you can't find it on google, then post here.

[stratele52]

+1 JJasilli ,

As I write ,no -34 volts at pin 5 mean a short to ground 

Short answer to mean too munch load somewhere . Here on tremolo circuit ?

[Madison]

I don't understand the tremolo on this circuit.
It is not the typical Fender that comes in and modulates the  bias and 220K resistor junction.

It does not have an optoisolator either.

I think, it is actually modulating some other way.
A preamp tube?
Does this sound like any other amp you can think of?
I am betting this is just a copy of something well known.......I think all the ELKs were.

I WISH I had a schematic.
Sorry.

[Attila]

I had one of these a few years ago and found it was based on a AB763 ish style circuit maybe that will help, by the way a lovely amp once the caps are replaced... and you change the speaker if you have the cab as well 

[PRR]

We don't know the tubes; heck, I don't see a B+ voltage. (After _30_ posts?)

In a classic Fender 50W, 70mA per tube is mighty hot. Get the grid bias more negative.

But now you say "switched to cathode bias"?? A 2-6L6 50W can't run well in cathode bias without raising the load, dropping the B+, and accepting 35W instead of 50W.

The designer of the ELK was capable of original thought. The same shop built many of the later Kustoms(??), which are bits of Fender and bits of somebody else. So without a schematic, a fuller description of this rig would be helpful.

[stratele52]

50 watts Elk shematic here ;

http://www.drtube.com/schematics/elk/twinamp50.gif

Some other here ;

http://www.drtube.com/guitamp.htm#Elk

[sluckey]

I don't understand the tremolo on this circuit.
It is not the typical Fender that comes in and modulates the  bias and 220K resistor junction.

It does not have an optoisolator either.

I think, it is actually modulating some other way.
A preamp tube?

The trem very likely modulates a preamp tube, similar to that Elk twin 50. I'd put the trem issue on the back burner and fix that bias problem on the output tubes. That's a much bigger issue.

-34V before the 220Ks. and -26V after.
-26V on G1 pin

-26V on pin 5 of the EL34 is the reason for them running so hot. That -34v should pass on to pin 5 without dropping thru the 220K. I'm surprised no one has mentioned leaky coupling caps between the PI plates and the EL34 grids. Leaky coupling caps are a prime suspect for low voltage on the EL34 grids when the bias voltage is correct at the junction of the 220K grid resistors. Locate the two coupling caps and disconnect one end of each. Do you now have -34v on each EL34 grid? If so, replace the coupling caps.

From what I've read about these amps, the resistors rarely fail, but the caps often fail.

EDIT... I just realized that you have already replaced the coupling caps. Disconnecting them is still a valid check to try to find out why the bias is only -26v on the grids.

[Madison]

Just gat back from the shop where the amp came from.
The guy has LOTS of MIJ vintages stuff and knowledge.
He also said ELKs were high on the failure list.
Unfortunately, he didn't have a schematic either.

Thanks for all your time and help!
 I will get back on this in a couple of days;need a break from ye Norse Viking.

[stratele52]

I fix some Elk amps before and many amp like these , low budget Fender copy.

There are few error on wiring or circuits .  Like some new worker do some mistake when building them. But once they are fixed they are tough.

[jjasilli]

The output of the tremolo circuit will probably be like one of these: 

* VibroChamp:  lead wire goes from intensity pot to small bottle tube.  If so, then this should not be affecting power tube bias.  But you can trace backwards from that pot to trace the tremolo circuit.

* Princeton:  intensity pot is a mixer with tremolo into one of the pot' s outer lugs; bias into the other outer lug; middle lug 2 sends lead wire to the two 220k resistors.  This tremolo circuit may muck with the bias supply.  Also, you can trace the term circuit backwards from the intensity pot. 

As sluckey says, it's a good ideas to temporarily remove the coupling caps to the power tubes.  There has to be an explanation for why the bias voltage has dropped almost 10V at the tube's grid pin, from the insertion point at the 220k resistors.  Something at this location must be the problem.

[Madison]

Back to this.
Pardon the delay, catastrophic computer melt down,motherboard failure, munched my hard disk, had to buy a used replacement.
My luck.

Anyways,

We don't know the tubes; heck, I don't see a B+ voltage. (After _30_ posts?)

In a classic Fender 50W, 70mA per tube is mighty hot. Get the grid bias more negative.

But now you say "switched to cathode bias"?? A 2-6L6 50W can't run well in cathode bias without raising the load, dropping the B+, and accepting 35W instead of 50W.

The designer of the ELK was capable of original thought. The same shop built many of the later Kustoms(??), which are bits of Fender and bits of somebody else. So without a schematic, a fuller description of this rig would be helpful.

Here is the tube line up from left to right:

EL34
EL34
12ax7
6AQ8 (this is written on the chassis.Some one stuck a 12ax7 in there.I have a 6AQ8.I put it in there with no change)
12ax7
12ax7

Voltages on EL34s

Pin3 435V
Pin4 430V
Pin5 -26V

I disconnected the coupling caps to the EL34s again,just to verify.
I read like 300-500 mV wiggling.
Just to be sure I stuck a couple of big .47 caps in there.
No change.

Checked the grid voltages with no coupling caps attached, no change

Removed all preamp tubes, no change.

Here is the really odd thing.
I put in temp filter caps on the power tubes.
The voltages flipped.
Pin 3 426V
Pin 4 431V
Pin 5 still -26

The first dropping resistor is 130R.
I replaced it with the same value increase to 10 watt when I started in on the project.

I'd like to cathode bias it just to be done.

 

[jjasilli]

Though a schematic for your Elk model seems hard find, here's one that seems to be a lower wattage version of your amp:  http://www.drtube.com/schematics/elk/es-30.gif  It seems to have your preamp and PI tube tube line-up, except for the power tubes:  http://www.drtube.com/guitamp.htm  So the schematic might be useful.  

1.  The trem circuit is likely identical.  One-tube 12ax7 wobbling the tone recovery stage.  

2.  a) Pull tubes.  

b) Disconnect the bias supply from the 2X 220K junction, at the bias end.  What is the bias output voltage?  

c) Pull the coupling caps; 220K's and grid stoppers (if any).  

d) run the bias lead wire from the bias output to the the junction lug of the 220K's (no resistors present yet).  What is the bias voltage at that point?  

e) re-connect 1X 220K resistor at the bias end only - no connection yet on the power tube side.  What is the bias voltage on the power tube side of this resistor?

f) same as e) for 2nd 220K resistor.

g) re-install 1X grid stopper.  What is the bias voltage on the tube pin lug?

h) same as g) for 2nd grid stopper.

i)  reconnect 1 coupling cap.  What is the bias voltage on the tube pin lug?

j)  same as i) for 2nd coupling cap.

Cathode bias.  IMHO cathode bias is all about the power tube HT supply voltage.  Specs say 375, but you've got more like 430.  Unless you do a 50 volt drop (which is fine), your tubes will run hot in cathode bias.  Cathode bias is dynamic, so that the tube always wants to draw the same idle current regardless of the size of the cathode resistor. E.g., because cathode bias is hot, you increase the size of the cathode resistor to limit current.  But the larger cathode resistor pushes up the plate voltage on the other end of the circuit, which exerts a stronger pull on the cathode's electrons and re-establishes about the same current draw.  And vice-versa.  The only way to defeat this circle is to lower the B+ supply voltage to the plates & screens.

[sluckey]

Pin 5 still -26

I'd like to cathode bias it just to be done.

Don't give up on it yet. The bias circuit is just too simple to admit defeat. It should be easier to fix than convert to cathode bias.

I just noticed that the grid stoppers are mounted on the sockets between pins 5 and 6. What is the voltage reading on pins 6?

You have eliminated everything except the tube sockets. Try this. Pull the output tubes and don't put them back in until you have the correct bias voltage on pins 5 of both sockets. Unsolder the yellow wire from both pins 6 and let the end of the wires dangle safely. Measure the voltage at the end of each dangling yellow wire. If the readings are now about -34 volts, then replace the tube sockets.

[Madison]

Measure the voltage at the end of each dangling yellow wire. If the readings are now about -34 volts, then replace the tube sockets.

Still -28
Sockets are good

1.  The trem circuit is likely identical.  One-tube 12ax7 wobbling the tone recovery stage. 

I think so too, that is net on the list.Thanks!

2.  a) Pull tubes. 

Check

b) Disconnect the bias supply from the 2X 220K junction, at the bias end.  What is the bias output voltage? 

-34 VDC

c) Pull the coupling caps; 220K's and grid stoppers (if any). 

Check

d) run the bias lead wire from the bias output to the the junction lug of the 220K's (no resistors present yet).  What is the bias voltage at that point? 

-34 VDC


e) re-connect 1X 220K resistor at the bias end only - no connection yet on the power tube side.  What is the bias voltage on the power tube side of this resistor?

#One resistor -28

f) same as e) for 2nd 220K resistor.

#Two resistor -28
Wait a minute!!

 

I am going about this from the "wrong end"!!!!

Back.

Okay.
I have changed the 220K resistors down to 20K.
I am reading hovering at -31 to -32 on pin5
If I go any lower on those resistors, say to 1.5K, it will read -34 but I loose tons power and tone.

So I am now reading with the 20K/1W resistors.
pin3 421
pin4 428
pin5 -31 to -32

Am I good ya think?
It sounds great!
Loud and chunky.

[PRR]

What kind of meter are you using? Digital, or a needle-meter?

[Madison]

What kind of meter are you using? Digital, or a needle-meter?

Digital.

[PRR]

34V at bias supply but 28V after a 220K resistor is 34V-28V= 6V drop. 6V in 220K suggests that you have 1Meg (1,027K) of "something" hanging at the grid to ground.

A few electrician DMMs do have 1Meg input, though most electronic-tech and many $3 DMMs are 10Meg input.

[Madison]

Confusion

34V at bias supply but 28V after a 220K resistor is 34V-28V= 6V drop. 6V in 220K suggests that you have 1Meg (1,027K) of "something" hanging at the grid to ground.


Where would I test and how for "something"?
Just the MM from the grid to ground or ?
Bad tubes?
Sockets?
Lost.......


A few electrician DMMs do have 1Meg input, though most electronic-tech and many $3 DMMs are 10Meg input.

If I am understanding this, my $20 meter ranges up to 2Meg.

[jjasilli]

A few electrician DMMs do have 1Meg input, though most electronic-tech and many $3 DMMs are 10Meg input.

If I am understanding this, my $20 meter ranges up to 2Meg.

No, you are misunderstanding PRR.  He is not talking about the maximum range that your meter can read.  Rather he is talking about the input impedance of your meter.  That was a problem for early SS meters - low input impedance.  This loads down the circuit under test causing a voltage drop and thus a false voltage reading.  (Kinda like the Heisenberg uncertainty principal at the macro level - the very act of measuring causes a change in the item being measured!).  So, VTVM's persisted for awhile, because they have a very high input impedance of 10M - 20M.  Eventually SS meters were developed with high input impedances, so VTVM's eventually all but disappeared (except for a niche group of curmudgeons who shall remain nameless).  Post if you want more info (about input impedance, not curmudgeons).

Something is causing a voltage drop (attenuation = subtraction), or a voltage division at the point in the circuit where you get the 6v drop.  Anyway PRR is suggesting that your meter might be giving a false, low voltage reading.  If not the meter, then something else is causing a drop of 6V, or a division by a factor of  5.7 (=34 / 6) at that specific point if the circuit. 

1.  Try a different meter;  and/or

2.  Run the bias output wire to a breadboard outside the amp.  Connect the black (common) lead of your DVM to the amp chassis; the red lead to the free end of the bias lead wire at the breadboard.  The far end of the wire @ the breadboard should still read -34V.  At the breadboard, connect one lead of a 220K R to the free end of the bias lead wire.  Connect the black (common) test lead of your DVM to the amp chassis; the red test lead to the free end of the resistor.  What does the meter read?

[sluckey]

Measure the voltage at the end of each dangling yellow wire. If the readings are now about -34 volts, then replace the tube sockets.

Still -28

Your results to this test mean your meter is almost certainly lying to you.

[jjasilli]

Yes, my outboard breadboard test is meant to help verify this. 

Also I inadvertently stated the wrong voltage division factor: 28/34 = .82

[Madison]

Ah, I sort of get it; contaminating the results by doing the test itself.
Thanks for the explanation JJ !!!

Okay did the test with the same -28 results.
Going to run down to the local home center and grab whatever MM they have.

[Willabe]

Going to run down to the local home center and grab whatever MM they have.

The instruction booklet should have a spec list in it and I've seen it written on the packaging before. Open up the package and look for the input Z before you buy it.


              Brad