bias problem

[jekaiser]

Help!    New to the forum.   

What would make a double ended el34 grid biased output section not conduct unless there is a signal on Pin 5.   I have the bias set at about -55 vdc and the plate voltage is about 534 vdc.   This amp is a Daily,  (made by Dennis Daily a few years back)   I can't find a schematic anywhere on the net.   

This started out as a frying sound in the amp.  I shut it off and found a short in the phase inverter and the the el34's were pretty weak with an intermittent short showing on both tubes.   I replaced the defective tubes and intstalled a 1 ohm precision resistor ahead of the cathode ground's.   Fired it up and there is virtually no voltage drop across the resistors.   I also checked the voltage drop across the output transformer between the center tap and the plates and same result.  When a signal (guitar strummed) is introduced, the Fluke shows current (via a voltage drop across resistor or transformer) and on the digital read out spikes.   I've seen spikes of 50+ mv.   The amp does amplify but it doesn't sound right.   

Any hints would be greatly appreciated. 

[sluckey]

I think your bias voltage is too high. Crank it down to about -40vdc, then monitor the voltage across your 1Ω resistors while decreasing the bias voltage to get the desired tube current. Watch for red plating in the EL34s.

[Willabe]

What would make a double ended el34 grid biased......

Double ended? Are you saying it's a push/pull (PP) 2 x EL34 output section?

[jekaiser]

To Willabe:   Yes.   Push Pull.   Two tubes.   Cathodes grounded.

To Sluckey:   Cranked it down as low as it would go.  -54 .    It will go to -64.     Anyway, cranked it down and checked.  One tube was dissipating about 12 watts the other about 31.  Swapped the tubes and the readings stayed with the sockets.  The readings did not follow the tubes.[/b]   The tubes check dang near equal as close as my tester will show.  The tubes are new and were purchased as matched/burned in.   Plate voltage was 528 at the time of the measurements.   I have seen the plate voltage drop to 465 for no apparent reason. 

CORRECTION.  THE READINGS DID FOLLOW THE TUBES

[Willabe]

The tubes check dang near equal as close as my tester will show. 

Sorry, but what tester?

[sluckey]

Cranked it down as low as it would go.  -54 .    It will go to -64

Sounds like the bias circuit is set up for 6550s or KT88s. You need to change a resistor in the bias circuit so you can get the voltage adjustment range down to -30v to -50v. Then the EL34s will be happy.

[jekaiser]

The tester is a Precision Apparatus Co, model 612.        I purchased this amp used, and played it almost every day, 4 or 5 hours per day for about6 months before it popped.  It sounded good.   Now it doesn't.    Looks like I need to get it to someone who has had more training/experience/equipment than I have.   I repaired radios in the army back in the 60's and that's looks like just enough to get me in trouble.  But I hate to let something get the best of me.   

[HotBluePlates]

The tester is a Precision Apparatus Co, model 612. ...

That tester is an emission-type tube tester. This type treats all tubes as diodes and checks how much electron emission the cathode can support. You can not match tubes with this tester.

What is happening in your case is each of your output tubes has a different transconductance (Gm). The U.S. unit for this value is "micro-mho" or "mill-mho" where the latter is equivalent to the European unit of transconductance of "mA/volt". You'll notice the Euro unit is the inverse of your Ohm's Law equation for resistance, and why the U.S. calls the unit a "mho" ("ohm" backwards).

Anyway, if you have 2 tubes with very-different Gm, they will idle at different currents with a given grid bias voltage. Your emission tester won't clue you in to the Gm difference (because that's not a factor in an emission-type test), which is why both tubes look similar on your tester but behave very differently in-circuit.

Separately, an emission test is hard on the tubes and could wear them out prematurely.

[jekaiser]

Thank you HotBluesPlates:    Very informative.    I purchased the tubes as "matched and burned in"  but unless they know what you know, I might have purchased a Granny Smith and a Red Delicious. 

So, If I may solicit your expertise once more, how then am I to accurately bias the tubes if at Idle they are performing so differently? 

[sluckey]

You can use two bias pots to adjust the tubes for the same idle current. But if... "One tube was dissipating about 12 watts the other about 31.", it would make sense to get a closer matched pair.

That still does not address the problem of your bias adjustment swinging between -54v and -64v. Wrong range for EL34s. That's the answer to your initial question... "What would make a double ended el34 grid biased output section not conduct unless there is a signal on Pin 5."

[HotBluePlates]

+1 to all of what Stuckey just said.

[jekaiser]

I can't thank those who jumped in on this and  you sluckey and hotblueplates enough.   I will attempt to bring the bias voltage in to range and install two bias pots.   May take me a while to gather up the necessary parts.  I'll have to pull the single pot out to find the value...    And who would you recommend as a vendor for a "matched" pair.   Obviously the vender I chose is selling more than tubes.......

[sluckey]

Post some hi rez pics of the amp circuit. We may be able to figure out the bias circuit. Then we can make better and maybe more detailed recommendations.

[jekaiser]

Having a difficult time getting down to the 1024 limit.   

[jekaiser]

Hope you can see this.   I noted the bias pot and where the wire emerges from the board to supply bias voltage to the two tubes. 

[jekaiser]

HOT DANG DO WAP A DILLY!    IT'S FIXED!!!!!!!

Thank you (in no particular order) to sluckey and HotBluePlates !!!!!  And to everyone.       I doubled the resistor ahead of the diode in the bias circuit and put a 250K pot in series with the original 50k bias pot.  That brought the bias voltage down to between -30 and -40 ish.   Set the bias for 44ma across the cathode 1ohm resistor and with the plate voltages being about 30 volts apart, (the output transformer was like 37 and 44 ohms) that worked out to about 19.8 watts on one tube and 21.2 on the other.   The bias voltage ended up at -39.4

The amp sounds better than it ever did, (imagine that.....,,)  You can go to sleep tonight with the thought that you helped another lost soul.   And from the looks of the number of posts you guys put out, you are not unaccustomed to accolades.   Thank you, thank you, thank you, thank you!!!!!!!!!!!!!!!

[Willabe]

How many mA's did you measure through each tube?

[jekaiser]

right at 44.   One was 44.1 and the other 44.2 and it would vary a couple of 10ths