Recto/B+ question

[Matty_V]

I tossed in a 5v4 in my amp to see what that did for clean headroom and its giving me the exact same voltage as the 5y3 at each node. Shouldn't that be noticeably higher? What am I missing? Also something that's peculiar. I'm using the 40x20x20x20 cap can doug sells, and the b+ drops to 0 in just a few seconds after turning off the amp. That doesn't seem right. I resoldered all the connections and no change.

Any help is much appreciated.

[phsyconoodler]

What amp is it and what bleed resistor have you got in there?

[Matty_V]

Its MY amp  No bleed resistors in there.

[sluckey]

I tossed in a 5v4 in my amp to see what that did for clean headroom and its giving me the exact same voltage as the 5y3 at each node. Shouldn't that be noticeably higher?

Maybe not. It depends on the 5Y3. NOS 5Y3s should give less B+, but the newer ones don't.

I'm using the 40x20x20x20 cap can doug sells, and the b+ drops to 0 in just a few seconds after turning off the amp. That doesn't seem right.

Pull all tubes except the recto. Now how fast does the B+ drop when turned off?

[Matty_V]

Its the 5y3gt doug sells. Odd the two rectos had the exact same numbers.

Regarding the B+...If all tubes are pulled, the B+ will essentially not drop. If you pop in V1 it will drop ~15V/second, and will drop more the more tubes are in. If all preamp tubes plus PI are pulled and just power tubes are in, the B+ will drop to 0 in 3-5 seconds.

Problem or not? Just seems fishy seeing as every disclaimer I've read about tube amps says filter caps may contain lethal voltages even after being off for a while.

[sluckey]

Its the 5y3gt doug sells.

That's not a NOS 5Y3.

...says filter caps may contain lethal voltages even after being off...

Key word in that phrase is "may".

[phsyconoodler]

Looking at your schematic it looks like you are missing a 100k resistor at the plate of V1 and you show a .5amp fuse on the AC hot coming in.
  And you have no negative feedback but you have 82k/100k plate resistors on the PI tube.Imbalanced but can sound interesting.

Never seen a master that works like that one either.

[Matty_V]

Are you referring to V1b? Its a cathode follower so no plate load resistor on that one. I think I just wrote the wrong value in the schematic for the .5A. I thought having the 82/100K on the PI balances things out because of unequal gain?

I do want to throw in some nfb to cancel out some highs. I find myself with the treble down most of the time.

That master works pretty awesomely, although I think I've reused that pot too many times and its acting screwy atm.

[phsyconoodler]

Normally you would use the 82k/100k WITH negative feedback so it balances the PI when feedback is used.Without NFB you normally use two 100k's.
   Hmm..cathode follower,but I'm talking about the plate on V1B.
And 250 ohm cathode resistor for a pair of EL84's? What voltage do you have?Normally an EL84 amp cathode biased uses a 125 ohm cathode resistor for a pair of tubes up to about 350v.

 

[Matty_V]

Doh! Ok I'll re-evaluate the 82K/100K situation after work.

Why would v1b need a plate load resistor?

I used the 250 ohm to cold bias them. I had a 130 ohm on there and upping it to 250 was one of many things I tried to help tame the circuit. It will probably change once everything is said and done and I have everything the way I want it. Maybe if slap on nfb I'll want to drop it back to 130.

[phsyconoodler]

Not to be a pest,but 1K screen resistors are kind of overkill for EL84's too.
  This amp a bit of a gain monster?

[Matty_V]

No no. I don't mind any criticism of my design. I'm learning as I go. I'll fiddle with the screen resistor values at some point to see what that does. You can have the screen supply and plate supply on el84s close to the same potential without any problems? This thing is an animal, but you can roll the pre down and boost the master and it cleans up very well. It would work for many musical tastes.

Any advice where I would start with some nfb? I don't want a presence control, just a fixed value that cuts some of the highs. They get pretty ear piercing when you crank this beast.

[phsyconoodler]

Leave the feedback off and add a vox style 'cut' control.Rolls off the highs nicely.Or simply voice the preamp differently.If you have to roll the treble pot all the way down,you are in the wrong zone anyway.
  Maybe up the V1 cathode bypass cap to 5uf or so.And add a bypass cap to the power tube cathode resistor.100uf/50v will thicken things up a bit.

  As long as your B+ is not too high,the screens don't need to be dropped that much.

[Matty_V]

I don't have to have it ALL the way down, but in the lower half or so. The cut control sounds interesting, I'll educate myself on it. I did have v1 bypass cap as 25uf and it was muddy as can be. I'll see if i have a 5uf on hand to try out. Perhaps an on-off-on switch could make for some nice bypass cap switching options there...or could be totally unnecessary. Its voiced just fine until you really start pushing the power tubes, then the treble gets to be too much. I removed the power tube cathode bypass cap cause it made them sound really sizzly.

My b+ @ node A is 368V

[sluckey]

Why would v1b need a plate load resistor?

It doesn't. It's fine just as drawn.

[OldHouseScott]

Is there really no coupling cap between V2's split-plate output and the grid of the PI? Seems like that would force the bias of the PI to be abnormally high.

[Matty_V]

There is (I hope). I just forgot to put it in the schematic. Good catch! If not I think that tube would be long gone by now 

[stingray_65]

I don't have to have it ALL the way down, but in the lower half or so. The cut control sounds interesting, I'll educate myself on it. I did have v1 bypass cap as 25uf and it was muddy as can be. I'll see if i have a 5uf on hand to try out. Perhaps an on-off-on switch could make for some nice bypass cap switching options there...or could be totally unnecessary. Its voiced just fine until you really start pushing the power tubes, then the treble gets to be too much. I removed the power tube cathode bypass cap cause it made them sound really sizzly.

My b+ @ node A is 368V

do give this a look while educating yourself on the cut control and cathode select switch

http://www.el34world.com/Forum/index.php?topic=5621.0

I use the cathode switch on nearly every amp I build, I find that the pickups play a HUGE part in the voicing of my amps.

when building a basicly Fender preamp the sound nice and bright with single coils and muddy with humbuckers. The switch lets me change to a .68uf cathode cap (ala Marshall) and that seems to make a world of difference. The 5uf is great for the single coils and the 10-25uf seems to sound best with Fishman 's found in acoustics.

Ray

[Matty_V]

Very cool! Thats a handy little gem to have. If I did go down that route I would probably mount the switch in the back. I've already got 5 knobs up front and don't want to drill any more holes up there. Or I could dump my MID pot. I wish doug stocked more kinds of switches.

For the time being I stuck a 100P between the plates of the PI and that canceled enough treble out, maybe even too much. But I do like adjustability of a pot (and so would a customer).

I've got too many options to think about.

[38Super]

Normally you would use the 82k/100k WITH negative feedback so it balances the PI when feedback is used.Without NFB you normally use two 100k's

This seems opposite to me.  If you are employing feedback into the negative input on the phase inverter, then the PI is acting more like a true differential amp.  If each grid is being driven, then you shouldn't need to build offset in to balance AC gain, side to side.

Without feedback, the PI's + input is being driven and the - input is at AC ground, through a capacitor.  This is the design I have seen the most.  In this configuration, the - input side of the PI is not being driven.  With constant grid AC voltage (0V), the - side of PI is operating in more of a common grid config, with input on the cathode, common grid, and output on the plate. 

In my understanding, this is why the AC gain balance is tweaked with an 82K plate load on the driven side and a 100K on the reference side.

One other comment to add is: with standby switch between 1st filter stage and choke, the choke will ring up when you switch the standby switch.  Coils like to have constant current or they'll protest with voltage transients to try to keep voltage constant.  (V=L*dI/dT)  Can move standby switch or place a reverse biased diode across coil to snub ringup.

rob

[Matty_V]

I got both of those bits of info from gerald webers video. For the PI he just says that one configuration has more gain than the other, so you use 82K/100K to balance the gains out. I switched the 82K to 100K and it sounds a bit congested, I think I'll swap it back.

Gerald advocates putting the first cap on the hot side of the standby, so its easier on the rectifier tube when the switch is flipped. Is this unnecessary? What is choke ringup and why do I not want it?

Thanks

[stingray_65]

What is choke ringup and why do I not want it?

Inductors when switched off will produce a large spike as their magnetic field collapses.

If you're old enough to have ever had a car with points ignition it's the same thing.

points = the switch, condenser = the cap and the coil = the choke.

Relays and solenoids suffer the same issue.

The remedy for this is to use a diode across the leads of the inductor to act as a shunt for the spike.

[Matty_V]

What diode should I use for this application? I've only got 1N4001s and 1N4007s on hand. Which direction would be reverse bias in this situation? In relation to my schematic should it be:

-----|<-----

or

----->|-----

Please admire my advanced drawing abilities. Can you describe in a little more detail whats going on. I get the idea but don't quite understand it.

I had to deal with points briefly on my '75 civic cvcc project until I stuck in a 2g civic distributor to do away with them. Tis a silly car. 

[stingray_65]

Ok Matt,

both those diode would blow in a spike, likely, at a lowish B+ of say 250V a spike of 3 or 4x is easily possible.

at best this is a patch for a poor design.

I HIGHly recommend visiting Merlin Blencowe's Valve wizard site.

www.freewebs.com/valvewizard/fullwave.html

He has a great chapter on how to design a power supply and quite a few good thoughts on standby switches.

He's not one to skin a cat one way either. Lots of options and he explains the pro's and cons of each choice.

His articles run a bit deep, I often have to re read them a few times before I get the jist, but his outlines and generalizations (and great diagrams!) really help.

[Matty_V]

Well I'm not into half-assing anything so I'll edumacate myself some more. I'd like to check out that site but it's not currently available. It says its exceeded its monthly bandwidth. Hmmm

[tonewood]

I'd like to check out that site but it's not currently available. It says its exceeded its monthly bandwidth. Hmmm

I've never seen his page down before. Makes me glad I have his two books. They really are great references.

Stringray, thanks for the info on the inductors.