Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: dude on April 08, 2017, 02:27:32 pm
-
I have many USA made older EL84s but have no idea if they are matched but to listen using my ears, not too reliable
So, to find tubes in my lot of USA made "used" el84 that are closely matched I came up with this method, bias each tube separately. I'm using an 18 watt amp
Just wondering if my thinking is correct?
Use two cathode biasing R's, since a 150 ohm is a good start for two tubes together, a 300 ohm for each EL84 and separate by pass caps.
Measure the voltage drop across the cathode R. Then divide this voltage drop by the valve of the R, (300 ohm in this case). That gives me current draw. Measure the plate voltage at the tube's plate to ground, subtracting the the voltage drop across the cathode R and multiply that by the current draw. This gives me the dissipation in watts. Try to find two tubes that are close or change either cathode R, up or down to match the tunes by dissipation. Do this for each tube.
I know I need to subtract the screens but that's probably less than a half watt in terms of dissipation. I'd be shooting for about 11 to 12 watts each tube.
I'm thinking having mismatched el84's in my 18 watt amp could be causing some unwanted distortion...?
Thanks,
al
-
Just wondering if my thinking is correct?
yup
I do 2SE, and 4SE, they each get their own Rk, then I || R's to "balance" current, typically within 5mA for 84's if I'm picky.
could be causing some unwanted distortion...?
with "unbalanced" tubes you get some Vk differences, which effects how much grid drive you need before distortion, so each tube distorts sooner/later. Is it bad? only the player will tell :icon_biggrin:
-
> 300 ohm for each EL84 ... Measure the voltage drop across the cathode R. Then divide this voltage drop by the valve of the R, (300 ohm in this case). That gives me current draw. Measure the plate voltage at the tube's plate to ground, subtracting the the voltage drop across the cathode R and multiply that by the current draw. This gives me the dissipation in watts. Try to find two tubes that are close or change either cathode R, up or down to match the tunes by dissipation. Do this for each tube.
Just measure the drop in 300 Ohms. If you are "matching" for push-pull, you do NOT need all that other math. It's same (near enough) resistor, voltage, etc. Two tubes 10.4V and 11.2V are plenty close enough to play. 9.5V and 13V are not too far apart to try.
-
> 300 ohm for each EL84 ... Measure the voltage drop across the cathode R. Then divide this voltage drop by the valve of the R, (300 ohm in this case). That gives me current draw. Measure the plate voltage at the tube's plate to ground, subtracting the the voltage drop across the cathode R and multiply that by the current draw. This gives me the dissipation in watts. Try to find two tubes that are close or change either cathode R, up or down to match the tunes by dissipation. Do this for each tube.
Just measure the drop in 300 Ohms. If you are "matching" for push-pull, you do NOT need all that other math. It's same (near enough) resistor, voltage, etc. Two tubes 10.4V and 11.2V are plenty close enough to play. 9.5V and 13V are not too far apart to try.
The Amp is push/pull. Just trying To figure a way to match all the old USA El84's I have, matching brands means nothing. Can I do this:
Two EL84's in amp with 150 ohm biasing R, pull one El84 warm up tube and turn on. I know the 150 ohm on one El84 will overheat in time (red plate) but will take a bit to damage tube. Check the voltage drop across the 150 ohm R after around 2 minutes or just as tube starts to red plate. Mark it down, turn off amp and try all the El84's. Match the tubes with the closest voltage drop. Anything with-in 1 volt drop, can be a good guess as matched, good enough for a cathode biased push/pull amp...?
al
-
... Can I do this:
Two EL84's in amp with 150 ohm biasing R, pull one El84 warm up tube and turn on. ...
You were on the right track the first time with individual 300Ω cathode resistors.
If you have to use a single shared cathode resistor (150Ω), then just install 1Ω resistors between each cathode & the non-grounded end of the 150Ω. You can measure/match any tubes put in the sockets.
... Two EL84's in amp with 150 ohm biasing R, pull one El84 warm up tube and turn on. I know the 150 ohm on one El84 will overheat in time (red plate) ...
They could start redplating as soon as the EL84's warm up & pass current. Better to just do it the right way.
-
For El84s what would be an acceptable reading across that 1 ohm resistor?
-
what would be an acceptable
for PP, something around 70% maxPlate current is a good start
-
https://www.tedweber.com/bias-calc (https://www.tedweber.com/bias-calc)
Franco
-
For El84s what would be an acceptable reading across that 1 ohm resistor?
12w / 400v = 30mA
12w / 300v = 40mA
Depending on your plate voltage, something between the two numbers above will represent 100% plate dissipation. Depending on the amp design (or desired longevity), you may want something less than 100% dissipation at idle.
-
Thanks for all the info, the reason I'm re-biasing this amp, over and over with several different EL84's is the frizzle, I'm still hearing with mild distortion and above. Just hearing that el84 frizzle after a mildly distorted power chord is hit hard, the tone is good but after a second or so I hear this distorted frizzle that's lasts a half second or so then fades in the distorted chord.
I read about the Paul Ruby mod years ago and this has amp has that mod. I recently change the 18 watt TMB using the attached schematic.
I tried tubenit's enhancing cap (250p) across the load plate R to the power tubes, some help but still the frizzle. The amp is stock as the schematic below except the mentioned Paul Ruby mod. And I'm using a 4.7uf preamp bypass cap on the tone/vol channel instead of a 1uf, no big deal. I also have a 1000 uf bypass cap with the biasing resistor (told this tightens up the amp a bit), could this 1000uf bypass cap be causing that frizzle? The amp is a Trinity 18 watt TMB, that I recently changed to another design by Trinity 18 watt Plexi MKll.
After the change, I'm now hearing the frizzle, when the gain and volume on the both channels are past 2 o'clock. I don't play with much more distortion, usually a little pass the start of breakup and sometimes a hair more gain for a lead. Single notes I don't hear it only on a hard hit power chord. It's a quick frizzle and just sounds bad to my ears.
Some changes I made with no help:
Changed the grid load resistors to 220K from 100K (power tubes), was thinking about changing the plate load R's to 100k/82K as I figure this unwanted distortion is in the PI. Balanced PI or not...?
Rather than try a lot of changes on my own, any help would be appreciated. This is a proven schematic by Steven at Trinity, could the Paul Ruby mod in the PI have a bad component from overheating that solder joint? Or do I not need it, was on the the amp's old schematic so I kept it. I shooting in the air trying to "guess" what be causing the frizzle.
Plate voltage is 338V at el84 plates. Cathode R is 150 and current across that is 11.8 vdc Dissipation is 12 watts per tube.
Appreciate any help,
al
-
I understand why no one wants to commit on this thread anymore, it's "been beaten to death" here and elsewhere. I search the Frizzle Problem with EL84 amps and the causes. Found a few threads here were snubbers were used and Ruby's zeners, changing OT's, larger grid Rs and tons of info. This problem is inherently common in 18 watt amps with EL84.
I fixed the problem, my zeners were 7Vs , too low as the cathode bias was around 11v at idle. Changing the zeners to 12 volts (only size I had on hand) fixed the frizzle when the el84's were driven hard. Everyone of you guys had a great deal of input on this subject in other threads which I found by a simple search. I apologize for asking before I search the topic.
Take care,
al
-
I took a painting class where we painted" in the round", you find quickly there are many views of the same subject, with just as many good outcomes. Glad you're up n running.
-
You don't need to add 1 oHM resistor to the cathode. Knowing the exact value of the 150R resistor will be fine. For example, if the 150R measures 152.8 R, use 152.8 in your calculation. Though it is more rapid to use 1 ohm resistors but you have to put in the very precise ones ( 1% or less).
Colas
-
Doing some reading about putting a Zener across the biasing R (in parallel) in this amp to limit the voltage and prevent blocking distortion on high gain settings.
Why is it that using a string of two 5 watt, 10 and 5 volt Zeners (total 15 volts) lets the voltage go pass 15v? Before the zener the voltage would go as high as 19 or 20 volts when hit hard but with the zeners, stops at 17 or 18v, not 15v? The zener string cathodes go toward the tube. Is this what you get, a 15 volt zener can go a bit beyond its rating?
al