Voltage Divider to drop B+

[seedlings]

I have about +400VDC on my 5F1 build. I'd love to have that down to around 350V. Zeners...changing to 6L6 or EL34... how about this easy voltage divider. Let me know what you think of the picture.

Voltage Divider (http://i206.photobucket.com/albums/bb54/seedlings/VDivider.jpg)

Thx

[donnyjaguar]

If you think about it, you shouldn't need the lower resistor as the current draw will be more or less consistent. Raise the value of the dropping resistor to compensate.

How would I do it? Easy, make the power supply into a choke input filter. That will drop it down but you'll still have the energy storage capability of the choke to smooth things out.

[seedlings]

Can't remember exactly how I came up with 23mA. I should've used 100mA since that's what my PT is rated for. The additional amperage means I will probably be making a toaster anyway.

...Also forgot to consider how the load will interact.

[tele_player]

A voltage divider isn't really appropriate in this application, load curent is too high.

Anything you do is going to dissipate some heat.

Simplest solution would be a CopperCap rectifier, assuming you've got a tube rectifier.

[Trout]

I have about +400VDC on my 5F1 build. I'd love to have that down to around 350V. Zeners...changing to 6L6 or EL34... how about this easy voltage divider. Let me know what you think of the picture.

Voltage Divider (http://i206.photobucket.com/albums/bb54/seedlings/VDivider.jpg)

Thx

400V on the anode of a 6V6GT is perfectly safe and would last decades biased properly and with the correct screen voltage. Fender typically slammed 6V6's with huge voltages and those amps have lasted decades.

I have run 450V on 6V6GTs for very extended periods of time the past couple years without issue.

Another approach if your worried about longevity would be to use JJ 6V6S's. They are built to handle much higher voltages than the garden variety 6V6GT.

Trout

[epluribus]

Seedlings, are you thinking of the touch characteristics--the way you can shape dirt response with voltage levels? I have a Kalamazoo Model II that's loosely based on a 5F1, and it uses a 3.3k resistor early in the power rail to set voltage levels throughout the circuit, as I understand it. Sounds like maybe this is what you're talking about.

Here's the schemo. It's R12.

The resistor appears to attempt to keep plate voltage above screen votage in the power tube, and oddly enough it's only a 1/2 watt resistor. Note how they experimented with values, originally 1K, then 2.2K, and finally 3.3K like mine. I ended up having to add a 2.2K screen resistor in series with the 3.3K to keep Vs about 3 volts below Vp. Wiring it that way kept it from reducing voltages further down the power rail. (BTW, at the moment they're both the 1/2 watt variety, soon to be 2W. It's not factory spec, but the smaller ones strike me as insufficient. Funny, it's run fine since 1966 that way... :rolleyes:)

I too have wanted to run even lower voltages throughout the amp, just to see how it impacts the sound and feel, so color me tuned in to what you find out here.

--Ray



http://img470.imageshack.us/img470/1416/kalamazoom2schemld5.jpg

[tele_player]

As you say, that 3.3k is only to keep the screen voltage below the plate voltage, otherwise bad things happen. 1/2W resistor is way more than you need there, that's easily calculated.

[epluribus]

As you say, that 3.3k is only to keep the screen voltage below the plate voltage, otherwise bad things happen. 1/2W resistor is way more than you need there, that's easily calculated.

Hey Tele Player, thanks. Steep learning curve still--rookies. :rolleyes:

As for Seedlings project, I note that when I put different resistors in there at R12, I can move the voltage structure of the entire amp--about 30V - 40V or more if I use a 10K vs. nothing at R12. (With the screen resistor in place R12 isn't needed.)

As for what that does for the amp, it indeed has an impact on the tone and feel. Generally, at lower voltages, the amp gets dirty very early and the dirt gets quite thick and harsh. (The rattiness being down to lousy gain structure I think, some tubes saturating way harder than others. The resistors and caps have been bench checked for drift.) At higher voltages, the amp stays clean longer, the threshold of breakup is wider on the pick, and the dirt is smoother, lighter, and easier on the ear--but not as Tweed-y (BTW, I don't care about loud at this point, though the voltage change has minimal effect on volume.)

A caveat with changing the voltage available to the power rail. If you note on your 5F1 schemo, Fender wired your Champ the same way Gib/Kzoo wired mine...the voltage divider for the power tube screen is on the power rail, which affects the relative voltages all the way down the line. What that means is that when you tune the proportions to keep screen voltage below plate voltage on your 6V6, you change all your other voltages too, changing the signature feel and tone of the amp.

To combat that, I started out by breadboarding all the resistors on the power rail, and ran 'em instead through 3/4W 10K 15-turn Cermet trimpots I picked up at RatShack for about six bucks a pack. (A process I'm still tweakin'. :)) First thing I did was get that initial voltage divider out of there and use a screen resistor on the screen supply branch of the circuit so it wouldn't shift all my other voltages. Then I replaced that divider (at R12 on my Kzoo, same spot on your 5F1.) with something I liked better, tone and feel-wise.

Tweaking the other B+ resistors downline, (R10, R7, and R4 on the Kzoo, same spots on yours.) I then set the plate voltages on all the other tubes more the way I like 'em, having to re-check the bias on the 12AX7 triodes, natch. Since I was fiddlin' with bias, this was also a good point to spec bypass caps for each stage, which has a huge impact on the tone and gain of each stage--cool stuff for sure.

Once I got the tubes breakin' up the way I want 'em, the next trick is to get 'em to all saturate at the same time, more or less. (Different times, actually, depending on what you want.) That I haven't gotten to yet, but it strikes me as a case of setting the relative signal levles into the grids of each gain stage, along with tweaking the coupling cap values. Also, things like impedance and total resistance issues have to be researched, I would think. Once again, thank heaven I finally learned about breadboards.

After all is said and done, it seems to me I'll end up writing down the values I ended up liking for just a few resistors and caps and such, and soldering them permenantly in place of the originals. Ultimately I would guess the finished amp will have about twenty bucks worth of caps and resistors in it, and look like almost nothing was done at all. I suspect it'll be a mod that would take about half an hour to do a second time, but it'll play vastly different.

So back to your question in your OP...the 5F1 appears to me to use a resistor for a voltage divider already, and the choice of values is entirely up to you. Just keep in mind that it'll change all sorts of things downline, not all of which are simple plate voltages, which is great if you get a kick out of tweakin' like I do. :NUTS

Other tech types, please feel free to disagree and QC my findings--like I said, I'm a pretty rank rookie at this point. But I think I'm on the right track...

--Ray

ps...Randall Aiken has a terrific article about these issues on his website.

http://www.aikenamps.com/

Go to Tech Info/Advanced/Designing common-cathode triode amplifiers (http://www.aikenamps.com/CommonCathode.htm) .

He gives you an intuitive grasp of what many of these circuit mods will do, from a tone and feel standpoint--if you read between the lines a little bit. Better yet, he also tells you how to actually calculate the values involved so you've got some hard theory to go with it. (Ultimately so you can make some intelligent predictions of what your mods are doing.) Great stuff, exceedingly eye-opening.

[seedlings]

epluribus, I paid $60 for a short stack of books where I didn't learn as much as in your post and those preceeding.

Can't wait to get home tonight to mess around with that R12!

It also just occurred to me that I may have 400V where that schem says +295 instead of where the schem says +280! I've read and re-read and fiddled so much that now I can't remember. If that's the case, my 6V6 probably has like 380V or something like that.

~novice~

Also, how "hot" should a power tube be to the touch? I've looked on line and over 100C I guess is OK and quite comon. Mine's HOT.

Wish there was a better word than "thanks"
-CHAD

(Money?)

[epluribus]

Hey Seedlings, 'preciate the kind words.

Just keep in mind that I'm really new at this too, and though I too have done a ton of reading, I'm only just now learning to apply it, and with sometimes halting success at that. Be sure to check me out on my "findings," as in my book they're provisional till proven guilty, at least at this point. (As Tele Player kindly pointed out. Thanks man!)

As for the hot tube, I guess I never really paid that much attention--I look for stuff like redplating, blue glow, and other oddities that have been discussed around here. But my real telltale is my multimeter--I find tons of stuff long before it becomes visible or tactile. But yeah, most all my tube gear runs rather uncomfortably hot to the touch.

Happy solderin', and solder safe!

--Ray

[tele_player]

Note: my comment about the 1/2W resistor at R12 only applies at low resistances.

Learn the power formula, and its various algebraic equivalences. Start with Power (in watts) = I (current in amps) * V (volts)
Volts = I * R (Ohm's Law)

[seedlings]

Well, found someplace that carried a 60V zener. Did the trick for now. Much creamier, a little compressed on the attack and a little "out-of-tune-ness". Sounds great. I'm still waiting for the 5Y3, though.

CHAD

[tele_player]

Where, physically and electrically, did you put the zener?

[seedlings]

Where, physically and electrically, did you put the zener?

The banded, cathode end of the Zener goes to the star ground (must be a star ground. Everything grounded at the same point). The other end of the Zener goes to the Center Tap of the PT. Volt meter shows -48V across the Zener.

Physically, it's sticking straight up in the air from where it's soldered to my ground (which is a 6" length of 14ga copper wire bolted to the chassis, where ALL my grounds are soldered to), and the CT is soldered to the other end up in the air.

I'm not done changing things yet, so its all temporary.

CHAD

edit* now voltages are 340 at the 16uF, 290 at the first 8uF and 250 at the second 8uF
http://i206.photobucket.com/albums/bb54/seedlings/champ_5f1_layout.gif
http://i206.photobucket.com/albums/bb54/seedlings/champ_5f1_schem.gif

[epluribus]

Much creamier, a little compressed on the attack and a little "out-of-tune-ness". Sounds great.

Hey Chad.

Not a bad description, similar to what I got. I like those little added singing harmonics, the "out-of-tune-ness," you get at lower voltages, but miss the pick definition on the attack. (The compression you mentioned.) Interesting tradeoff, and part of what makes Tweed different than Blackface IMHO. The other big difference I notice is the thicker flavor of the dirt and the earlier, faster onset at lower voltages--easy to overdo, but terrific when you get the balance right. Nice to hear the project is progressing so well.

--Ray

BTW, voltages: Stock, mine were running 265 Vp at V2, 205 at V1a (the second half of the 12AX7), and 172 at V1b (the first gain stage). I'm running both sides of the 12AX7 a bit hotter now, (230 and 200V, respectively) for a bit more headroom relative to the EL84, and the additional pick detail, with the EL84 still at 265. Each stage responds a bit more evenly to the volume knob now. Might have to try the preamp at even higher voltages like you're running.

[JJman]

My 5F1 build also had hi voltages. I think the power tranny may not be for a Champ. Who knows. Since the PS current is pretty low in a Champ, I use a 1K 5 (or 10?) watt resistor between the rectifier and the PS. I think this is the same thing a Weber Copper Cap does (the ones with the voltage drop feature.) It works nicely. The 6v6 plate is now ~340volts (at 12watts idle) whereas it was ~420volts (at the same 12watts idle.) Obviously I changed the cathode cap to bias back to 12watts.

I forget the wattage that the resistor dissipates but it’s well below it’s rating and has never blown in about 8 years. I use one on my 2 x el84 amp too.

[epluribus]

Interesting, I just dug up the data sheets for the 6V6--both of these 5F1 builds run really hot out of the box. Seems the max plate voltage is listed as 315V. By contrast, a Kalamazoo Model II/'66 Gib Skylark (both constantly touted as a 5F1 in Gibson trim, arguably partially true.) runs an EL84 at about 265-280V, well within its 300Vp limit. Hmmm...

Let's see...If (as a generalization) high plate voltage = more headroom, then it would seem these 5F1 circuits are running with far more headroom, with the dirt coming on much later on the volume knob--compared to the Gib/Kzoo version. Also, according to Dave Hunter's book, high Vp also reduces the touch of the amp, (stiffens it) which experience would also suggest. The Kzoo goes from clean to way-over-the-top dirty just on pick attack alone, almost too much for itself. I have a hunch your builds are much more controllable, partly because of the high Vp. Hmmm...

--Ray

[seedlings]

I may put the two 12V zeners in there too, just to mess around. I'd like to see how it plays at 325V. Maybe this could be another mod? Throw some zeners in there in a heavy-duty selector switch to vary the voltages. That'd really shake up the feel. Probably only work in cathode biased amps, though. Check out these Zeners. (https://taweber.powweb.com/store/vdump.htm) Mine just looks like a regular diode.

CHAD

[Old Tele man]

Interesting, I just dug up the data sheets for the 6V6--both of these 5F1 builds run really hot out of the box. Seems the max plate voltage is listed as 315V.
--Ray...check a little closer on those data sheets and you'll notice that the 6V6GTA is also spec'd for "maximum peak positive voltage (absolute maximum) = 1200 volts...in Vertical Deflection Amplifier - Triode Connection." [...the duty cycle is the voltage pulse not to exceed 15% of a scanning cycle, ie: 30-freame, 525-line, TV-circuit].

...so, the tube CAN handle high voltage *if* done in short pulses and not continuously. Luckily, Class-AB1 operation isn't totally continuous when operating, but rather basically 50%-duty cycle...one tube ON the other tube OFF and back-n-forth like a see-saw...the only real 'continuous' operation is at idle when BOTH tubes are conducting, but doing so only very little.

...with Class-A operation, however, they're ALWAYS conducting...so the above info is moot.

[epluribus]

...check a little closer on those data sheets and you'll notice that the 6V6GTA is also spec'd for "maximum peak positive voltage (absolute maximum) = 1200 volts...in Vertical Deflection Amplifier - Triode Connection." [...the duty cycle is the voltage pulse not to exceed 15% of a scanning cycle, ie: 30-freame, 525-line, TV-circuit].

...so, the tube CAN handle high voltage *if* done in short pulses and not continuously. Luckily, Class-AB1 operation isn't totally continuous when operating, but rather basically 50%-duty cycle...one tube ON the other tube OFF and back-n-forth like a see-saw...the only real 'continuous' operation is at idle when BOTH tubes are conducting, but doing so only very little.

...with Class-A operation, however, they're ALWAYS conducting...so the above info is moot.

Hey OTM.

Been on the road a couple days, just got in to read this. Interesting point, seems I vaguely recall reading something like that but haven't gotten as far as following up on it. I would think, too, that as long as you don't get too far beyond the max dissipation curve you have a certain amount of room to play with higher Vp. Apparently a lot of mfgrs. think so. :)

As for the impact of voltage on tube response, I just finished computing a bunch of load lines to see if the charts bore out the test-drive. Changing only B+ and finding the appropriate internal plate resistance for that voltage, the available grid voltage swing gets significantly bigger as you just go above the max-rated Vp for the tube--aka much more headroom. Even more interesting is how the effect can tail off as the V goes higher.

Ultimately a very useful thing to draw for yourself on paper--sheds a lot of light on some of those gritty low-voltage circuits. Also very interesting to note that the bias on the 12AX7s in the Kalamazoo/Gib family is exceedingly high when viewed against the characteristic charts. Even worse when they start stickin' bypass caps in there. Either a different Rk/Ck or a different B+ in the preamp would make a world of difference here, I suspect.

For those of us tinkering with KalGib 5F1 variants, this might be a real useful thing to sit down and sketch--might make lots more sense out of the listening tests. Might even apply to the Champ kits we've been talking about.

--Ray

[Old Tele man]

...with MOST octal tubes, their max Vp rating results from their pin-to-pin spacings at the base, as witnessed by the 6BG6 which has it's plate connected to a "cap" instead of a base pin with 700 Vp, but which can also handle +6.6KV and -1.5KV "pulsed" voltages...and this tube is basically just a 6L6 designed for RF applications where "plate-cap" connections are VERY common.