SE paralleled dissimilar tubes headache to bias !

[Colas LeGrippa]

Hi,

I am just trying to find my problem a cure without the right remedy. I feel I'd need a diode somewhere but I can't say more. Figure a SE amp. one 6550 on one side and one 6V6 on the other. Each tube has its own cathode resistor. When I try to bias HOT both tubes to their limit, I just can't. I can easily bias the 6550 alone at 117mA / 300V , but when both tubes are plugged in, the bias goes wrong, like if they'd have declared war to each other,,,,,,,,,,,,, The PT is strong enough so is the OT.

Colas LeGrippa

[John]

The tubes are in parallel, so the resistors will then be in parallel also, which will halve the resistance.

At least, I think that's right... but don't be betting money on me.

[Willabe]

I wonder if it has to do with the OT's anode R and very different current draws at different plate resistance for the 2 tubes?

Also are you feeding their screens from the same B+ node as the OT/plate? If so could that cause some kind of feed back with the OT/plates/screens?


             Brad      

[HotBluePlates]

.... one 6550 on one side and one 6V6 on the other. Each tube has its own cathode resistor. When I try to bias HOT both tubes to their limit, I just can't. ...

Each tube, given its own cathode resistor, should settle in to whatever idle current is right for the tube and the conditions.

So what is your power/output transformer and cathode resistor for each tube?

[Colas LeGrippa]

Each tube, given its own cathode resistor, should settle in to whatever idle current is right for the tube and the conditions.

That, I was not quite sure about, since dissimilar tubes were used. So, if the voltage drops too much, trying to raise the tubes current, the power transformer is too weak then. The OT is a Mercury Magnetics AXIOM 4.5 K and the PT I just know the voltages,  ( 290-0-290 ) other specs are unknown but it fed a pair of 6L6 before. To know for sure, I will swap the 6550 for an EL34 and see if I can bias both tubes ( EL34 and 6V6 ) right on the hot spot. The guy ( Peter ? ) at MM assured me the OT could easily handle a pair of 6L6, so it has to be the PT that just can't do the job. One question : would swapping the recto tube for a SS give more power to the PT ?

Thanks, bro.
I love you all !!!!!!!!

Colas[/color]

[Colas LeGrippa]

suite)

the cathode resistor for the 6550 is 200 R, and the 6V6 is 360 R . This is the best I can do and in so doing the tubes reach  80%  their limit ( 100% dissipation ).  If I try hotter ( lower resistors values ), the voltage drops to the point where it becomes colder !

Colas

[Willabe]

So, if the voltage drops too much, trying to raise the tubes current, the power transformer is too weak then.

That should be right, Ohms law.

the PT I just know the voltages,  ( 290-0-290 ) other specs are unknown.  The guy ( Peter ? ) at MM assured me the OT could easily handle a pair of 6L6, so it has to be the PT that just can't do the job.

That's 1 of the problems with MM and I've bought their iron. They won't come out and list the specs on their web site or tell you what they realy are on the phone.

You said that the 6550 was biased at 117mA? A 6V6 SE would be about 40mA? So 117mA + 40mA = 157mA, not includind the preamp tubes? A pair of 6L6's in PSE would be about, at most, 50mA each? Probably 40mA's each?

So, 157mA - 100mA (2x6L6 @ 50mA each, PSE) = 57mA, best case, or 157mA - 80mA (2x6L6 @ 40mA each, PSE) = 77mA, worst case, that's at least, best case, 1/3 more than the PT's B+ can supply?



                      Brad      

[PRR]

Get two meters, one for B+ and one to read cathode current. Bias the 6550 alone to 56 Watts (42W+14W). A good 6550 will stand 56 Watts for many minutes. As you should already know, from a cold start the current is zero, at 6 seconds it starts to rise, at 12 seconds it is near full current. Go 15 to 20 seconds, take readings, and shut down fast.

I estimate 150 ohms under 6550 with 300V plate and screen.

If a "290V" PT won't hold 300V, then it isn't big enough.

> OT is ...4.5 K

The best-guess load for 300V and 186mA (42W+14W) is 1.6K, way lower than 4.5K. This may be part of your problem: a higher Z OT will have higher DC resistance. Possibly as much as 400 ohms. Which gives a hefty 75V DC drop just in the OT.

290VAC should give 370V DC. 370V in a 4.5K load is 83mA. THAT's where you should be aiming. Higher current will not give larger output power.

370V at 83mA is 31 Watts. If you allocate 1 part to 6V6 and 3 parts to 6550 (12W:35W ratio), that's 21mA and 62mA. Possible cathode resistors are 1400 under 6V6 and 480 under 6550. And yes, this uses 65% of the tubes' rated dissipation.

[sluckey]

Figure a SE amp. one 6550 on one side and one 6V6 on the other.

There's only one side in a SE amp. Which OT are you using and exactly how do you have the circuit connected? I'm imagining plates tied together, screens tied together, grids tied together (maybe with separate grid stopper resistors), and cathodes with individual resistors. A pic would be very nice.

The guy ( Peter ? ) at MM assured me the OT could easily handle a pair of 6L6

Did you tell MM exactly what you are doing with this OT? What was their response?

[Colas LeGrippa]

The best-guess load for 300V and 186mA (42W+14W) is 1.6K, way lower than 4.5K.

That's no big deal. Just hooking up a 2 ohm speaker cab to the 8 ohm tap will almost correct the impedance mismatch.

That's 1 of the problems with MM and I've bought their iron. They won't come out and list the specs on their web site or tell you what they realy are on the phone.

  You're right: I asked the sales person ( no engineer for sure  ) at MM how much current could handle the OT and  he would never tell me. ''the only spec we have is 4.5k'' is what he answered me, before pushing me to order it ( and what I did ).........I'll go back to my bench now trying to find a solution with all of your clues.

Thanks, bros, I love you.Colas

[Colas LeGrippa]

Interesting discovery

Definitely, my PT is too weak. I have done the bias on the 6550, just a bit cold at 30W . I have swapped the 6V6 ( 14W ) for an EL84 ( 12 W ) but all I could get is 18mA from that PT , roughly 6W according to the voltage, for feeding the EL84. When I raise the bias on the EL84 and lower it on the 6550, I get more highs, like if the EL84 was unable to give all its guts at lower bias settings. So, by playing with different bias values,  we can change the tonal characteristic of the amp and it gives ineresting tone changes. You are right if you argue that lowering an output tube bias will reduce the total output power, give poor performance, and maybe open the door to parasitic oscillation and more. But, ( and I have performed tests all day long ...thanks to my patient lady, woman, lover, girlfriend who has cooked a delicious turkey for supper meanwhile... but that's another butterball game... ), but, I repeat, in my case, it gave a very interesting tone. The EL84 biased at 6W and the 6550 just below 30W, give a perfect balance in my amp. El 84 too hot, = too piercing highs, 6550 too cold: sterile tone. That beautiful tone I've obtained by adjusting both bias couldn't have been obtained by inserting a high or low-pass filter somewhere in the signal path. It has to do with the tubes bias, and I realised that a good tone is not obtained automatically by biasing the tubes at their max power and tweak around after. The bias is a part of the tweaking ( I think I do not teach you something new here, for me it is new though. )

Colas

[jjasilli]

I suspect the general rule is true that, within reason, hotter bias sounds better than colder bias.  Most systems work best when performing at or approaching their maximum (from below). 

EL-84's are a special case -- a common complaint is that they sound piercing or glassy. I've had success with lower plate voltage and larger grid stoppers.  By lower plate voltage I mean within spec, say 280 - 300VDC, as they are often run at higher voltage.  (E.g., Traynor & Trainwreck.) Next time I'll play with bias too.  Good suggestion!

BUT, I'm confused about paralleling these tubes.  My understanding is that paralleled tubes should have the same output impedances (plate-to-plate resistance).  Otherwise current draw will get screwy.  Que passa?


P.S.:  Hope you enjoyed the turkey!

[Colas LeGrippa]

BUT, I'm confused about paralleling these tubes.  My understanding is that paralleled tubes should have the same output impedances (plate-to-plate resistance).  Otherwise current draw will get screwy.  Que passa?

 

Well, first I thought that my problem was related to the current drawn by 2 different tubes, but HBP confirmed that each tube would draw its own current, according to its own cathode resitor, which appeared to be right according to my different tests and experiences. To tell you the truth, I never understood the concept of plate-to-plate resistance, but the 2 different current seem to get along quite well !

The only math I used here is finding the total reflected impedance according to the total current drawn by the 2 tubes vs the high voltage as suggested by PRR, which gives approx 2k ( 280V / 133 mA ).

P.S.: the turkey was excellent, so was the cook.

[HotBluePlates]

My understanding is that paralleled tubes should have the same output impedances (plate-to-plate resistance). 

Remember for d.c., the tubes only see/know the d.c. resistance of the OT. Impedance will only come into play when a signal is applied, and a non-optimum impedance will likely just result in less power output (and more or different distortion).

One question : would swapping the recto tube for a SS give more power to the PT ?

Yes.

Aside from the current capacity of the PT, the power supply has its own impedance. Mostly, this is the d.c. resistance of the secondary, plus the primary resistance of the PT, referred to the secondary (this is the primary d.c. resistance reflected to the secondary, just like you'd normally calculate for secondary to primary for the OT).

However, other things can add to the power supply impedance. A tube rectifier will have some added impedance (why we like them to add sag/drop voltage). Any choke/resistor in series with the supply ahead of the plate will increase power supply impedance. And of course, the OT d.c. resistance is in series with the supply impedance as far as the output tube plates are concerned.

All these things will cause the output tube plate voltage to drop with increased current draw, according to ohm's law. The 6550, drawing the bulk of the current, will contribute most to this effect.

Of course, if the PT just can't supply the current (without its own sag due to its ratings), then you'll be limited by it no matter what you do.

[Colas LeGrippa]

Thanks , I love you all !

[PRR]

> My understanding is that paralleled tubes should have the same output impedances

Anything else makes the designer's head hurt.

Actually, the tubes will do what they do; and plate-impedance alone is not enough.

BUT: Colas' basic idea is (pardon me!) "crazy".

When I saw it, I thought of this:


Plowing with one big Belgian horse and one miniature horse.

Obviously this plow is not going to go straight.

I don't think Colas wants to "go straight" either.

That plow, to go straight, you are going to have to rein-in the big horse to the same strength as the little horse. It would really work better if you sent the little horse back in the house, and let the big horse work naturally.

Or you could plow "bent". Maybe if you wanted tight crop circles, the big/little horse team would be excellent. (Like we used to build circle-track race cars all offset.)

Colas has use of many "normal" amps that will "plow straight" or "plow strong". But he likes to search outside the well-harvested fields. He's got a fine intuitive understanding, he learns how to work the reins, and he's not discouraged when his experiments don't turn out wonderful. So if he wants to hitch a Belgian and a Mini together, why not? (At worst, he has a 6V6 *or* 6550 Champ which may be old-hat but useful.)

[Colas LeGrippa]

BUT: Colas' basic idea is (pardon me!) "crazy".

well, great discoveries have been made by crazy people who held on tight to their dreams, like Einstein and thousands others who were considered as crazy people, so thanks for the compliment !!!!


PRR thinks that electrons from 2 different sources act as 2 horses of different strength.  I think electrons flow like water. Imagine two water hoses, one small and one large, out of which 2 different current are coming out and mixed together to make a wheel turn. Ech flow of water contributes to make the wheel turn because they are part of total water flow. The wheel will keep on turning steadily, with no imbalance at all.

I don't think Colas wants to "go straight" either.

PRR may go the way he wants. Probably from the old school, he's got the knowledge and seems to understand perfectly all the theory related to valve amps. He helps us all here, day after day, restlessly and we can rely on him. But, can he understand ( at firt glance ) something he's never seen before nor studied in ''straight'' books ?

For me, a guitar amplifier is a musical instrument ( which is totally different from a Hi-Fi amp ). Experimenting paths where few people have gone is ineteresting because it gives me the opportunity of fully understanding the ''mechanic'' of tubes.

So if he wants to hitch a Belgian and a Mini together, why not?

Why building a straight Champ, when I know in advance that I will succeed and what it will be, tone wise ? What do I do after ? Leave it in the closet ? Give it away, or take it apart ?

[Willabe]

Colas has use of many "normal" amps that will "plow straight" or "plow strong". But he likes to search outside the well-harvested fields. He's got a fine intuitive understanding, he learns how to work the reins, and he's not discouraged when his experiments don't turn out wonderful. So if he wants to hitch a Belgian and a Mini together, why not?

Alvin you missed the best part.


                 Brad     

[Colas LeGrippa]

I didn't miss it : it touched me right to the bones, this sure is a nice comment !

[Colas LeGrippa]

Hi  bros,

I am back with my dissimilar tubes after completing another day of experiments. At half volume with an overdrive  pedal, sounds great, I've already said that. BUT, sometimes, when I step close to the amp with my guitar, I can hear my amp trying to talk to me, rather shout to me, imitating the noise that my car's starter does when I turn the key in the ignition as the engine's already running..........pretty much the same. In the amplifier's language, it means: ''Stop me, I can't take it anymore''.......so I read HBP, PRR , Sluckey and all other's posts one more time concerning this thread and I came to the conclusion that I had to run the 6550 alone..[color=red]FOR THE MOMENT[/color]. The bias is much more stable and so is the tone. BUT THERE MUST BE A WAY TO DO IT with just one OT . I gotta to find a way of isolating the 2 tubes with transformers or diodes to achieve that. I'll get back to you .I love you all

Colas

[HotBluePlates]

You probably need grid, screen and/or plate stoppers, connected right at each socket.

That suggestion assumes I understand you correctly that the amp tends to break into oscillation when you use both output tubes.

Keep plate/screen wires away from grid wires.

[Colas LeGrippa]

That suggestion assumes I understand you correctly that the amp tends to break into oscillation when you use both output tubes.

when I use  ONLY ONE output tube as well.......................my prob with 2 different tubes comes prob. from oscillation, and not from the use of two different tubes at the same time. I just discovered that today, Jesus-Crowd.

I'll go back to my f..........amp and re-dress the leads for the 3 rd time now...........this time if it doesn't work, I put this piece of shit in the garbage, I call MM and ask for a reimbursment ( wow I'm getting good in English ). No joke, i have NEVER had such a pain in the a..........mplifier. One week spoiled, trying and trying, lungs spoiled too with more fume inhaled than my truck exhaust running for an hour ( badly said but you can understand ).

By chance, my PP 50W runs like a cadillac ( with just a small rattle.....bad tube, bad cap ? )

Colas