Plate voltage is too low on output valves.

[phylomatic]

I've got a question guys. I've recently built an AX84 October and had great success with it. It works and pumps out good volume at about 9-10 watts each tube. I'm sure most of you are familiar with the AX84, if not you can check it out at www.ax84.com. Mine is the october and I have very little variance with what he used.

My problem/question is this. I had trouble getting the grid's mA up to 29 so I had to play around with some resistor values that set up the B+ values for the circuit. I know that has a technical name but I forgot what it is. Anyway, all my voltages matched perfectly with the schematic but all I could get was 14 mA and the amp was WAY to quiet even though it sounded ok. It wouldn't ever break up so it always sounded clean and no louder than say a tv a moderate level. So, after I changed some values of those resistors I finally got the bias set perfect but the plate voltage dropped too much for me. I'm using JJ/Tesla 6V6 tubes and they aren't that expensive so I would like to get that optimal 70 percent dissipation because I'm not worried about having to buy new output tubes. From what I hear, the JJ/Teslas are built like a tank anyway. They've survived a first time amp builder's build so that's proof enough for me. I've posted my schematic and I would like to know what is affecting the plate voltage so much and is there any way I can get it up without sacrificing the bias mA that I have now. I posted this on the AX84 forum and someone suspects the screen resistors, but I'd like a second opinion because I have to special order high wattage resistors because I don't just have those lying around like I do the smaller wattage.

Here is my conversation below on the other forum with the replies.

'Plate voltage on output valves is low.'
Author: Phylomatic (registered user: 7 posts )

My problem started with having too low of an amperage reading on the grid of the output tubes. That seemed to be a common problem and some tweaking of a few resistors resolved this situation for me, like it has so many others.

I won't go into specifics with the voltages because my eyes start to cross when I read about it so I'll post a link to the schematic that I scanned with my current voltages and the resistor values that I changed.

My problem now is that with the resistors that I've changed I suspect I've negatively effected my plates voltage on the output tubes. They are now off by about a little more than 10 percent, which I know is satisfactory, but what bugs me is that all my voltages were just about perfect, except for the ones going to the biasing pots, and of course the biasing amps were off. I was getting only 13 mA instead of the 29 that I got from my calculation. Anyway, I know it doesn't make that much of a difference. I would like to know why the plate voltages are so out of whack now after the changes I made. Is there a better way to get the biasing set than the changes I had made, and because I'm a bit of a perfectionist, I would like to have the plate voltages closer if possible. Any help would be appreciated.

Here is the link to the schematic

[link] //I'll post the attachment

I put a big red square around the resistor values that have been changed. Also, I wrote the new voltages in permanent marker so they'll show up better. Also, before I forget, I ruined the UF4007 diode so I had to use a IN4007 I bought from Radio Shack out of convenience. I suspect I messed up the rectifier, or even more so, I suspect that the IN4007 is causing this mess. If you guys agree that its the diode, could you give insight on why that is?

I'm using JJ/Tesla 8 pin tubes, and the 9 pin preamp tubes are the new Mullards. All of them work wonderfully, and I've tried (not on purpose)to destroy the JJ/Tesla output tubes, but these things are still working. This is my first amp build, I know I should have started with the easier one, but I did my homework and it actually came together relatively trouble free and I didn't shock myself (knock on wood). I did zap about a thousand multi-meter fuses though testing voltages without moving it off the mA setting. Thanks again in advance.
views:

Reply to this Message
Author: Harrison Ford Prefect (registered user: 12945 posts )
No, subbing a 1N for a UF wouldn't do that.

But running the current up will drop the voltage if the PT is getting loaded. Did you use the same PT as originally spec'd?

But I think the real problem is your screen current. You're pulling 10mA per tube for that, which is 3.1W. Dunno about the JJs, but the old standard for 6V6s was 2.2W max per screen, and typical screen current was 2-4mA at Ep in the 300V range. I'd try bigger screen resistors. You may need to bump the screen voltage back up a bit if it drops too much from that.
views:

Reply to this Message
Author: Phylomatic (registered user: 7 posts )

Is there any reason to use UF over 1N?

The power transformer is exactly the same as the one spec'd.

I'll give it a try. I suspected the screen current, but put it as the very last thing to check. I'll let you know the results. By the way, the JJ/Tesla 6V6 (this is the new ones, not NOS) is spec'd at 13mA at 250V typical.
views:

Reply to this Message
Author: MasonAtom (registered user: 259 posts )
UFs tend to cut down on high end noise/switching hash and fizziness compared to the 1n4007s. At least, that is my experience in the builds that I have done. Really simple and cheap mod that makes a noticeable difference.

Mason

Reply to this Message
Author: Phylomatic (registered user: 7 posts )

I've got a question about your screen resistor calculation.

Here is what I'm assuming, and I could be completely wrong or my brain isn't working right.

Voltage = amps times resistance
Power = Voltage * Amps

If that's the case then I show a voltage drop of 10 volts across my screen resistor which give me 10/1000 or 10 mA like you said. If Power is Voltage * Amps than I calculate just .1 Watts or 10/1000...am I doing this calculation wrong?

So, you can also see I have a bit of confusion with his calculation if someone could help me out with that also. Thanks guys. I love this forum because you always seem able to help me usually without posting. I fixed a Peavey amp of mine that I would have never known what to do if it wasn't for this place so I appreciate it.

[rockon1]

Dont know a lot about this stuff but Im wondering if perhaps you tried another type of 6V6? I do know a bit about tubes. The JJ is a "freak" amongst 6V6's. It can handle plate voltages of 500v similar to many 6L6's. It also biases up low often compared to others. I like to call it the JJ6VL6. Just a thought. Bob

[6267]

It's not the bias ma that matters, but the relationship between the anode voltage and the anode current that sets up the tone. So it's not a 17ma tone or whatever, it's a 17ma @ 420v or what have you. First you need to make sure that the voltage is nice and high with all tubes removed. I have often seen new builds or rebuilds where a solder blob, piece of resistor lead, or other metal junk has wedged itself somewhere causing a weak short to the B+...sometimes this blows fuses, other times it just eats up current. If the voltage is high enough with tubes removed, we can assume that the problem is with either the tubes, the power supply resistor choices, or both.

It could be the tubes drawing too much current, but you should be able to increase capacitance to reduce the sag. If it's a 5y3 rectifier, try using a 5z4.
Good luck.

10% is close enough for tubes, Leo's resistors were 20%. The tubes vary more than that, i.e. most tubes are only 10% within spec anyway. At best!

[phylomatic]

I have not tried other tubes. I looked at the data sheet and you seem to hit the nail on the head as far as this tubes values...but what I don't understand is why it sounds so bad at 13mA? Mine is a 6V6S so is there much difference in that than say a 6V6GT?

[txstrat]

I like to call it the JJ6VL6.

I agree. I've recently built a Deluxe 6G3 with JJ 6V6s and with all voltages as specified in the schematic the tubes run at 43ma with no problems so far.
No redplating, great sound.

[phylomatic]

Right, I knew that about the mA not mattering, but the relationship is the important thing. So, the Voltage on the plate is the same as the cathode as I understand. My test points are over the 1 ohm resistor between the cathode and ground and that's where I'm measuring 29 mA. With 29 mA on the cathode and I'm measuring about 317 volts on the plate so thats 9.2 Watts. The MAX dissipation on the 6v6 is 14 Watts. The recommended wattage is usually agreed to be 70 percent of the maximum dissipation of the tube so I should have just put a little more thought into it and just bumped the mAs up to about 31 and that would put the wattage at 9.8 watts which would be the "optimum", or what I would like to call my reference point.

I guess I've come to the conclusion, with the help of you guys, that I don't have a problem, if I want 70 percent I just need to bump up the mAs a bit to match the plate voltage...which seems so obvious now that I'm either completely way off base now, or I'm really embarrassed for having thought about it so long. Does this sound about right?

Thanks for your help guys, this forum is great and always fixes my issues, even if it is just talking it through.

[txstrat]

I wouldn't get stuck on the 70% mark.
Many Fenders are biased quite on the cold side and sound great.
You could tinker around with the mA until you find the spot you like.
I wouldn't care if it's 70%, 60% or whatever.

[Enzo]

On fact if you look on most Fender schematics of the era, you will see a note to the effect: all voltages within +/-20%. Even the parts in earlier Fenders were of 20% tolerance. Electrolytic caps had tolerances of +80/-20%. Seriiously. SO I agree, don;t get hung up on numbers. And certainly not 2ma of plate current.

[rockon1]

I wouldn't get stuck on the 70% mark.
Many Fenders are biased quite on the cold side and sound great.
You could tinker around with the mA until you find the spot you like.
I wouldn't care if it's 70%, 60% or whatever.

+1 Its like a mythical number these days! Anywhere from 50-70 is fine IMO if it sounds fine. Bob

[Bruce / Mission Amps]

EL84s.
Does dropping 10v across his separate, 1K screen resistors send a semi-red flag up to anyone else? That's is 10ma of screen current at idle.
That means 20ma of the PT is used up for two screens and resistors.

[MWJB]

They're 6V6s, but yes, it does still look like a lot of screen current.

[phylomatic]

EL84s.
Does dropping 10v across his separate, 1K screen resistors send a semi-red flag up to anyone else? That's is 10ma of screen current at idle.
That means 20ma of the PT is used up for two screens and resistors.

You're not the first person to notice that. I posted this question at the AX84 forum and someone was alerted to that same thing. It says the typical screen voltage is 13mA at 250V (6V6 datasheet) for a push-pull setup so it doesn't seem to be that unusual. Should I work on lowering the wattage at the screen?

On another note. I realize that now about not fretting over the mA on the grid. I wasn't really concerned so much as I'm trying to understand the relationship between all the components and I just didn't get when the plate voltage dropped that much. I would like to get it operating at the 70 percent range though just so I have a point of reference and not just say, eh that's close enough. It probably isn't at all noticeable audibly for all I know, but I'm kind of into details. I guess I feel if I'm going to spend the money then I'm going to make sure I do it right. After it's so-called "right", then I will adjust from there to suit my tastes if needed. Right now all I have to go on is what is termed typical. Now I'm worried about the screen....

[6267]

That does seam fairly ridiculous. Can you see the screen grid through the plate on a 6v6s? I wonder if there are winding problems or red/yellow/white glowing going on in there. I would certainly try different tubes if you're sure about all the resistor values and wiring etc. I've had some 6CA7s (the beam kind, from EH) that had some bad winding on them...you could see the screens pulse with the music . Could this be RF? Also, you said earlier the amp is quiet...what are you using as a preamp for this poweramp?

[phylomatic]

No, it used to be quiet, now it has a decent volume. I've come to find out that the plate and grid aren't really a problem but the screen seems to be the issue. Looking inside the tube nothing looks funny as far as I can tell.

I'm a little confused at the alarm people have over the screen. Can someone explain to me the issue if the data sheet says typical is 13 mA and I only have 10 mA on the screen?

[6267]

I guess I never really measured screen current in my output tubes (I rarely end up using tubes with well published specs in my design, and as long as the screen isn't glowing I try not to worry too much...). I'll go check my filmobeast (total redesign, quieter than a churchmouse, meaner than a junkyard dog) and see what my "6v6"s are drawing on their screens.