5F2A with a 6L6 prebuild questions

[big-daddy-59]

I;m just passing on info from this forum and others on the size of the first cap when using a glass rectifier. I'm sure there is a spec for it but I'm still pretty much a noob at this as well. I don't understand all the abreviations on the spec sheets for instance. I can just pass on what's worked for me and some lessons learned the hard way by blowing stuff up.

[tubeswell]

Is there some spec sheet you got your info about " not go over 40uf with a 5y3".

Most 5Y3GT datasheets of yesteryear say 20uF as a maximum capacitance. But the info on those old datasheets is there to ensure that you get a long life out of your tube in your TV, radio or whatever. 40uF is what most geetar amp peeps push a 5Y3Gt to in SE amps in order to get most of the spongy flab out of the supply. 20uF will work fine however, as will 16uF.

I would like to understand more about the power section electrolytic cap's and how the "uf "relates to the circuit and tubes. I understand how caps work in a tone stack and also how by-pass caps work, but filter caps, I don't know. Will hi "uf's" stress the rectifier tube?

F is the standard unit of capacitance. uF is standard shorthand for 'microfarad' and 1 uF is one-millionth of a farad. (although in the old days they used to notate this is 'mF', which is confusing because in shorthand for other units of measurement where a lower-case 'm' is included, people usually understand this to refer to a 'milli' of something, meaning 'one-thousandths' of a unit of measurement. Similarly in the old days they used to denote picofarads as 'mmF' meaning micro-microfarads).

1 nF (nanofarad) is one thousand-millionths of a farad. 1 pF (picofarad) is one million-millionths of a farad. The order of units-of-capacitance from big to small is F, mF (millifarad - 1 thousandths - okay big deal - but where do you see this being used?) uF, nF, pF. The more capacitance, the more stress is on the rectifier because more charge appears on the electrodes inside the rectifier tube.

The first filter (a.k.a. the reservoir cap) cap's job is to 'smooth-out' the 'gaps' between the peaks of AC swing in the rectified power rail (after your typical Full-wave rectifier has done its job of putting all the VAC bumps on the same side of the 0-reference point voltage potential).

The reservoir cap does this by storing charge on the upswingin' part of the cycle and releasing this back into the power supply on the downswing side of the cycle. This effectively removes the 'bumps' in the rectified AC power supply - leaving a sort of 'smaller-sinewave' of power at a higher potential than 0, (but looking nowhere near as bumpy as the pre-filtered supply does).

The resistor or choke (on a typical CRC or CLC filter) then further 'reduces' the 'waviness' of the 'rectified-and-capacified' power signal (each in its own respective way).

Then the output filter cap on your typical CRC or CLC filter takes care of the little bit of ripple that the other bits can't deal with, leaving you with a pretty smooth DC voltage several hundred volts above 0 (ground) potential.

There's plenty of sites on the net where you can looking up AC power supplies and rectification and filtering types with pretty pictures that better explain all of this BTW.

[tyru007]

I've only built a few amps so far but I spent a lot of time on a BF Champ (AA764) 15 watt SE build with various tubes.

I used the Hammond 270DX power supply (285-0-285 at 120 VAC) with a Hammond 125ESE. I ran it using parallel output tubes, tube recitifed, SS rectified, with and without a choke, etc.

Here's my take.

The circuit does not appear to sag. Except for a voltage drop, I get the same response using a tube or diodes to rectify the voltage.

The SE is sensitive to power supply ripple. At the very least, put a resistor between the first two caps to create a PI filter and increase the size of the caps (as tubeswell describes above). The 10-25W wirewound resistor 100-330R would work well.


I'm currently using 4 UF4007s to a 40 uf cap -- 150 mA, 4-5H choke -- 20 uf cap prior to the plates (only because I have the choke).

I have tried a EH 6V6, TAD 6L6GC and a JJ KT77 with JJ and EH 12AX7. I like the EH 12AX7 with KT77 the best so far.

I would recommend adding 1.5K to 5.6K grid resistor to the power tube.

I lose a lot through the Fender BF tone stack and the 6L6 or KT77 did not seem to break up as early. I added back some gain by changing the plate resistor on the first gain stage to a 220K and I have a bypassed cathode resistor on the second gain stage. You may/may not need extra gain with the Tweed tone stack.

With the B+ at 400 to 450V, you may need another filter stage to drop the preamp plate voltages.

At 450V you may need to use filter caps in series for the first two filter stages so that you have a filter rated for the high B+ voltages. Two 350V 40uf caps in series would give you a filter with 20uf 700V rating.

[Gerard]

All great points Tubeswell and Tryu007. I thank you for your comments. What will adding a grid resistor to the power tube do? and the difference sound wise from a 1.5k to a 5.6k? grid resistor would be?

[MWJB]

"The circuit does not appear to sag. Except for a voltage drop, I get the same response using a tube or diodes to rectify the voltage." - That isn't typical, I normally find that the envelope changes with each different rectifier. On a regular champ with 1x6V6 you might not get a huge change in idle voltage from one recto to the next because current draw through it is not that great.

[tyru007]

All great points Tubeswell and Tryu007. I thank you for your comments. What will adding a grid resistor to the power tube do? and the difference sound wise from a 1.5k to a 5.6k? grid resistor would be?

Used to prevent parasitic oscillation and blocking distortion. It will not effect tone. Fenders typically use the 1.5K on 6L6s while marshalls use the 5.6K with El34s. I'd go with the higher 5.6K.

For more info go to:
http://www.aikenamps.com/InputRes.htm
http://www.freewebs.com/valvewizard/gridstopper.html

MWJB - I compared a rectifier of 4 UF4007 diodes and a 220R 25W inductive wirewound resistor to a GZ34 rectifier so that they both gave about the same B+. Granted that the GZ34 does not produce much sag and the 220R resistor does add resistance to change in voltage. I did not detect much diffence in these two methods of rectification for the SE amp.

[MWJB]

"I compared a rectifier of 4 UF4007 diodes and a 220R 25W inductive wirewound resistor to a GZ34 rectifier so that they both gave about the same B+. Granted that the GZ34 does not produce much sag and the 220R resistor does add resistance to change in voltage. I did not detect much diffence the the method of rectification for the map." By trying to emulate the voltage drop of the tube, with the diodes & resistor, you have muddied the water somewhat regarding a comparison between tube and SS rectification. Diodes alone don't drop the voltage, it would have been a better comparison if you had A/B'd the diodes vs the GZ34 without the dropping resistor. What you did was similar to Weber's copper cap recto & some folks feel that these sound just like tubes, so I'm not surprised that there wasn't much difference.

[tyru007]

MWJB - I may try subbing differnet rectifier some time if I can get my hands on a few types just to see. I just wanted to know if a diode + "sag" resistor = GZ34 rectifier, which it appears it does for an SE amp. I have left mine with the solid state rectifier.

Gerard - One thing I found out worked well for reducing the hum from the filements was to raise the ground voltage reference for the 6.3V power supply. This along with good power supply filter made my amp almost silent at full volume.

If your 6.3V supply has no center tap, create a artificial center tap using two 100 ohm resistor then tie the center tap to either the cathode of the power tube (at about 20-25V) or create a 30-40V supply using a voltage divider with 270K then 22K resistor to ground off the B+ (450V) supply (also acts as your capacitor bleed resistor. The voltage at the connection of the 270K and 22K should be about 35V.

I have mine tied to the cathode of the power tube. For some reason I get a slight hum at the speaker when the amp is in standby, which goes away when the standby switch is turned to play.

[Gerard]

All great points!! Thanks. Now it is off to my build.

[Gerard]

Up and running (almost)

Hi, I just finished my 5f2a with a 6l6gc. When I first started it up, it sounded like garbage! rattling/buzzing distortion. So I disconnected the negitive feed back wire. That made it sound alot better. (do I need the change the value of the negitive feedback resistor? or is something else wrong?)
2nd, When the volume is turned up past 9, the sound cuts out (oscillation?). What pins of the 6l6 do I need to install the grid stopper resistor on?
I am running this amp with a gz34 , 6l6gc and 12ax7 . hammond 123ese O.T. It has 440v on the B+ . 510 ohm cathode resistor
How do I make this amp less muddy with H/B pickups?
I have not adjusted any voltages yet, I could use some help! here's the voltages

6L6GC
#3-436v
#4-395v
#8-31v

12AX7
#1-212v
#3-1.70v
#6-215v
#8-1.66v

[MWJB]

Go bigger on the cathode resistor, 510ohms is borderline at 436vdc. 680-800ohms.

No, you shouldn't need to change the NFB loop. Something else is probably wrong.

Grid stopper can go from pin 6 to 5. Connection to 220K grid load connects to pin 6. Though, this shouldn't really be necessary.

Your input jack doesn't look right? 1meg should go from hot to ground, with a jumper from ground to the switch.

You seem to have all your grounds daisy chained, separate the preamp & power amp grouns, run the preamp grounds to the input jack, power amp to a PT bolt. That 0.0047cap should be grounded to the back of the tone pot.

[big-daddy-59]

with a single input the grid resistor should be 33k not 68k. I can't tell from your pictures what the value is. Preamp voltages are a little high still, might want to increase the value of R11.

[MWJB]

Lose the black ground wire that runs from your input jack to the speaker jacks, the speaker jacks ground through their connection to the chassis at the socket nut. If you have insulated them, don't.

[Gerard]

working better

I took your suggestions and did the following:

corrected the input jack 1m resistor wiring

grounded the tone cap at the pot


removed the ground wire from the input to the output jacks

grounded the volume pot to the input jack ground

changed the cathode resistor to 680 ohms, now idling at 51ma

changed the input resistor from 68k to 22k (is that ok?)

the hum has always been very quite even at wide open throttle. no problem with 60hz hum.

the other problem I was having with the buzzing distortion with the NFB circuit hooked up was the O.T wires. I just reversed them from tube to B+. That took care of the buzzing distortion.

When I built this amp, I thought it would have more output (loudness). I have built a mission 5e3 and that to me is quite loud! I know this is a small practice amp , but I thought it would be louder.
I have tried all types of tubes in it , 5u4gb, gz34,5y3gt, 6v6,6l6gc,12ax7,12ay7, 12au7,12at7, I havent got a favorite yet, and yes I will watch the voltage when switching tubes.

The only problem I have now is some sort of parasitic oscillation . When the voulme is past 5, there is a slight vibrato effect, just slightly noticable. The amp is still usable but I need to tace that problem down. Which wires should I look at first .Which wires would I need to shorten or reroute?

Would you recommend using a shielded wire from the volume pot to pin #7 of the 12ax7?

Any thoughts on putting a cap on the second cathode resistor of the 12ax7?

Thanks for all your input !!

[MWJB]

Yes shield the wire from volume pot to 12AX7 pin 7, route it (and any other grid wires) away from plate wires (12AX pins 1 & 6). The orientation of your tube sockets seems to encourage long parallel runs of plate & grid wires? Would it be feasible/easier to rotate the 12AX7 by 180 deg?

Also the wire running from the 220K grid load to the power tube grid is rather long, you could also shield this, or what I would do is remove the coupling cap that connects to 12AX7 pin 6 and see if it will reach from the turret where it meets the 100K plate resistor straight to the power tube? Keep it away from the heater wires, mount a new 220K directly on the power tube pin 5 & ground to PT bolt.

What about the PT centre taps & grounds are these still soldered to the chassis?

A bypass cap at pin 8 of the 12AX7 will shunt a large proportion of the signal there to ground, lessening the effect of the NFB loop (it will still have some effect). It will also increase gain, not something I think you want until your oscillation issues are sorted.

Connect a true RMS AC meter across your speaker terminals, play the amp. Voltage will jump about some but try and determine an average, square the average then divide by speaker load in ohms...I'd expect about 8W?