Green Bullet: EF86 + 5E3 in U.L. mode.

[chocopower]

Hi!

I have an old Shaftesbury combo.



Some times i use the power stage with a POD and a 2x12 cab with great results.

I was trying to bulid a head with an EF86 preamp and use as inspiration the Dr-z28 preamp.

This a previous schematic. I have joined the Drz preamp and the Shafthesbury power amp.

Any commentary or properness will be welcome. Thanks in advance.

UPDATE: NON GOOD WORKING SCHEMATIC. MODDED ONE DOWN THE POST

[snoof]

Whenever I use pentodes as a first gain stage, the coupling cap ends up being much smaller.  Like .0047 or the like...

[kagliostro]

I think Geezer and Tubenit can help you

see the HoSo56 amp, use the search engine here in the forum

Kagliostro

[jojokeo]

Looks good to me and will work however I tailor a couple of things such as:
1 - w/ single input - resistor no higher than 33k and I actually like 15k here.
2 - I like a bias resistor value btwn 2.2k & 3.3k w/ those other values being the same.
3 - I like a higher B+ voltage than normal to help lessen microponic issues.

[tubeswell]

If you've got Merlin Blencowe's book handy, take a look at chapter 3 where he recommends using lower plate resistor (lower gain) in EF86  to reduce microphonics. (He designs one with a 100k load line and an HT of 250, 60-70V scr, and 1.4V cathode bias). (I don't know how much more I can tell you without disclosing too much about his book)

[snoof]

it does help w/ microphonics, but it also changes the sound.

[imaradiostar]

following the Merlin school of thought- add a large grid stopper to the split load triode- I tried it on a recent design and thought it worked well.

jamie

[sluckey]

I would suggest moving the node D filter cap to the chassis that the EF86 is mounted.

[chocopower]

Sluckey: Why should i move node D filter cap? Could it generate any oscillation or hum problem?

The original amp, use this node for tremolo, wich is locate in the preamp chasis. For the rest of the preamp there are 2 more filter nodes.

thanks everybody for yours comments. I,ll test it.

[imaradiostar]

I just noticed you're using the rt 66 pre- I wouldn't mess with the pentode or tone values too much, at least not until you've heard it as-is. It works pretty well the way you have it. It won't ever do pretty sparkly blackface fender clean sounds but it'll do a whole range of slightly to very dirty sounds that are very useful. It's amazing how well it'll show the character of the guitar you're using while creating interesting harmonic content the seems to "round off" the rough edges of a given guitar's tone. I'd say mine favor brighter sounding guitars but your experiences may be different. It also works well with pedals. I've found that an amp like this with a distortion pedal or two and my h2o chorus delay makes a very portable small gig rig that covers a wide range of sounds.

jamie

[sluckey]

Sluckey: Why should i move node D filter cap?

It's a good idea to have the filter cap located close to the circuit it supplies, especially on a split chassis. According to the schematic, node D only supplies that V1 pentode. Are you saying you have more circuitry that's not shown on your schematic?

[jojokeo]

I agree w/ Sluckey in that it's just good design to have the cap close to it's associated circuit if it's possible - but it's not a must have. I've done it various ways but after the first three caps or a cap can, then the remaining one or two go on the board near their counterparts (pun intended) so design the layout accordingly.

[chocopower]

Imaradiostar: Use the RT66 the same preamp that the Dr-z28 ? I search the net without result. Finally i decide use the Z-28.
I´m looking for a simply little amp and the Drz´s are the kind i like.

sluckey / jojokeo: This is the original preamp schematic (sorry, not finished). Is located on the top of the combo.
Is a cool amp, but the elliptical speakers have a poor bass response and a low output.
I wanna use it as a modular amp. The combo for home/small gigs and the power amp section+new preamp+2x12 cab for medium gigs.
As you can see, the filter cap node for the preamp (except for the tremolo circuit) is located on the preamp secction. My idea is use the tremolo filter node (located in the power amp secction) for the new preamp. This way, i can use the power sec. as a Lego. The conection is done using a octal tube socket for the B+, ground, AC power and heaters. The signal use a RCA socket.



More amp pics:
http://www.vintagehofner.co.uk/britamps/restof/shaft1.html

[chocopower]

Hi again!
Amp working.
The only problem is that treble pot need to be on max to get a good highs level.
Time to tweek the tone stack... Looking at it, is very similar to a a Baxandall.
If i want more highs is logical decrease the value of the .0047uf cap betwen treble pot and ground? I tried a .0033 and seems to work, but mids goes too much...
Maybe rise the .0068uf bypass cap in the vol. pot is a better solution?

[jojokeo]

Hi again!
Amp working.
The only problem is that treble pot need to be on max to get a good highs level.
Time to tweek the tone stack... Looking at it, is very similar to a a Baxandall.
If i want more highs is logical decrease the value of the .0047uf cap betwen treble pot and ground? I tried a .0033 and seems to work, but mids goes too much...
Maybe rise the .0068uf bypass cap in the vol. pot is a better solution?

Personally I'm not a fan of the Bax stax but before you go changing it I have three suggestions:
1 - put the 470pf & 470k resistor in parallel not series after the pentode. I'd also change that 470k resistor to 220k instead so that when you do this, you won't lose quite as much signal (only ~32% is about right instead of 50% - too much) & this could be exactly what you need giving you a more useable range w/ that pentode. You could also instead try a 100k resistor w/ no cap accross it too - you'll get more signal w/ more high end but a little bit of high end will be knocked off. It will also act as a nice grid stopper for your next stage preventing blocking distortion. Pentodes typically don't sound too good from too much gain/drive going into the following stage w/out doing this (taking away a bit of signal) because from about 8 to 10 on the control pot and you should find it a worthwhile change.
2 - if even more high end wanted, put 470pf cap across 500k vol pot, lugs 1 & 2.
3 - if you don't like #2 above w/ the added slight hiss background noise, put a 470pf cap accross the 470k resistor down towards the end going to the power amp and check that out for your tastes.

[chocopower]

upppsss. sorry. I think there are a confusion..I talking about the first schematic, on the top of the topic.

UPDATE: NON GOOD WORKING SCHEMATIC. MODDED ONE DOWN THE POST


The one you talk is the original preamp... I build a new and simplified preamp to use with the stock power amp. As you can see in the pics, the power amp with OT an PT are in the base of the combo. Original preamp is in the top. I´m just gonna use the power section with a new build preamp.

[jojokeo]

Sorry for the misunderstanding.
Things to try in order to "brighten / enhance" the high end: remember that by reducing bass and/or mids can give the signal an aparent increase in highs.

1 - use either a 33k or 22k input resistor, not 68k.
2 - change your .02s to .01
3 - try a 100k slop resistor, not 150k and raise the 250pF to 500pF or .001uF
4 - put a 2.2 or 4.7uF cap on the CF 1st stage cathode.
5 - short or remove the feedback loop

[chocopower]

Sorry for the misunderstanding.
Things to try in order to "brighten / enhance" the high end: remember that by reducing bass and/or mids can give the signal an aparent increase in highs.

1 - use either a 33k or 22k input resistor, not 68k.
2 - change your .02s to .01
3 - try a 100k slop resistor, not 150k and raise the 250pF to 500pF or .001uF
4 - put a 2.2 or 4.7uF cap on the CF 1st stage cathode.
5 - short or remove the feedback loop

Hi! I already had made some of your solutions-
1.- i,m using a 33k imput resistor. I just had a mistake in the draw.
3.- i rise the 250pf cap to 750pf, but signal get so much (to much up-mids...). Today i,m gonna try 500pf.
     Thanks for the "100k slop resistor". I was trying to understand this tonestack and that helps a lot.
5.- As you can see in the schematic i install a 100k pot in series with the 5k6 original NFB loop resistor. with pot at max, no  NFB loop.

I´ll test the 2.- and the 4.- point. Some voltage reading could be a good information too for you and others.

And last question... The way the treble pot work is really diferent. is more like a guitar tonet cut that a usual amp treble pot. Is that normal?


thanks a lot!!

[chocopower]

1.- 15k imput resistor installed.
3.- 100k slope resistor installed. Cut a bit the basses..

Some voltage readings:

"D" node: 280v
Pin 6: 70v (maybe a bit low, isn´t?)
Pin 1: 80v (shouldn´t be lower than anode voltage???)
Pin3: 1.7v

Just for fun, i added a 100pf cap between the imput of the tone stack and output of treble pot. It works nice. With this cap, the treble pot is more usable and give a nice mid bump in highest setings.

upsss... and some really important. I avoid the 470k grid leak resistor in the output stage module. In orinal amp had his sense, but in this design is "paralleling" the 1M volumen pot...

[chocopower]

OK...

After try several mods i decide eliminate the treble/bass tonestack and try the 5E3 one.
Now this is the schematic. Is working, stable and just need some adjust looking for a more balanced sound (bass, treble, gain, distorsión, etc.....)



Jojokeo sugested me in this topic (http://www.el34world.com/Forum/index.php?topic=9133.new#new)avoid the tone pot due to the fact that i need dial it at max to get a good treble level. I´m gonna try it and install a cut control after the P.I. to control the extra highs that, sure, i´m gonna get. :-)

[jojokeo]

Nice to see your updated schematic. Do you know that you're cutting your bass response more when you switch in the other cap values in series after the pentode?  Just checking because when caps are in series it's like resistor's being in parallel. You are already low and cutting the bass by a lot w/ the .001 coupling cap going to the others which will reduce even more bass further and I get a sense that you are not happy w/ bass response? Try at minimum a .01 and you could go higher to a .02 for your "basic" or "standard" setting and switch in series another .02 which be equivellent to a .01, then a .01, a .005, or others depending on how much bass you ultimately want to cut out. Hopefully you'll like the tonal response a bit better?

Another thing is it also seems you're still wanting more highs too? You should have plenty and if not, what kind of speaker are you using? Also, are you playing humbucking pickups too? These both can help to make it seem that you don't have enough highs but when you plug in a single coil your highs could be too much. Do you test your amp for these two things to be tried or considered before making too many changes that may not be necessary or possibly too drastic?

[chocopower]

Nice to see your updated schematic. Do you know that you're cutting your bass response more when you switch in the other cap values in series after the pentode?
Hi! Yes, i know. I using the same sistem used in 70´s Oranges. They call FAC control. I used it just because i had a cheap 4 position rotary swicht with those caps in it. I used it in a Orange clon i rebuild last year. I know i need twek the values.
Ahh! there are an error in the schemo! the coupling cap after the EF86 is a .01uf, not the .001 i draw (CORRECTED).... The bass response is quite good. At low volumen i dial the swicht at bypass mode. This way i have punch at bedroom levels. Like i "Surround" in a hifi!
If i rise the volumen i move the swicht cutting the excesive bass. I just need to find the ideal values for a proportional and more gradual bass cut in each position.

Another thing is it also seems you're still wanting more highs too? You should have plenty and if not, what kind of speaker are you using? Also, are you playing humbucking pickups too?
I´m testing the amp with a Jazzmaster and a 2x12 cab with a V30+Greenback. I think they are not the ideal speakers for this amp, but is the only one i have at this moment. At low volumen i can dial the tone control at middle or 3/4 position, but with the amp dimmed, i need put it at maximun and still sound a little dark. No so much with bridge pickup, but a little dark for the neck pickup.
I tried too, a Lespaul type guitar with JB and i missed some more highs. I think that if i eliminate the tone control i could get a lot more highs and control them after the PI with the Cut control, as you sugested.

These both can help to make it seem that you don't have enough highs but when you plug in a single coil your highs could be too much. Do you test your amp for these two things to be tried or considered before making too many changes that may not be necessary or possibly too drastic?
At this moment the preamp is mounted just in a wood board , so i´m not afraid about make a lot of test. (really ugly. I´ll make some picks four you divertiment )
I´m playing too with the Ef86 resistor values just for learn how a pentode works. This time i want to use some maths and not just put something i saw out there.
I,m reading and trying to understand the information i found in the Valvewizard web. I have some problems with the inglish (Spanish public education... ), but is a good web for newbies like me.
Simple and with pics...hahaha

Thanks four your conments. I´ll keep you informated!!

[jojokeo]

All is sounding good w/ the amp. For vol & tone - if you have tone up full it should act as a bright cap put across vol pot on lugs 1 & 2 so high end is preserved there when vol control is not up full. Next thing that could limit high end is the 100K grid stop resistor. Since bass freqrencies cause possible blocking distortion more than high frequencies, use a capacitor across the resistor from 500pF - .001uF and see how you like it. This should help your high end because there's nothing else restricting highs but this.

[chocopower]

I have been comparing the power section schemo and is really close to any 5E...  Fender amp (5E3 deluxe, for example).

The most obvious diference is the UL output transformer, but this is somethig i like to keep just because is the only amp with this caracteristic i have.
the other noticiable diference is the power nodes distribution. The 5E3 and similar designs use the same power node for the cathodyne PI and previously gain stage, whereas my amp have a diferent one for each one.

Just for fun....  Could be a good idea eliminate one of the 10k/16uf node and feed the PI and previously gain stage from one of the 100uf node?
This way, al voltages will rise in the EF86 and 1/2 ecc83 gain stage.
Besides the problem of lack of treble, i would like to improve the headroom and this could bee a good solution for both problems.

chao!

[jojokeo]

Headroom issues now...I believe it's from the EF86 the way you're using it. Here's a few suggestions to help you dial it in better.

1. Put a voltage divider on the output and get musical gain through the whole control range.
2. Raise the HT voltage.
3. Put a switch on or eliminate the cathode bypass cap.
4. Lower the load resistor.
5. Raise the screen resistor.
6. Make a pentode/triode option.

[chocopower]

Not a issue... just having fun while i learn something.

today`s work:

.-Found that the two 1k2 screen grid resistor in the EL34 rise her value (1k4 and 1k5). Last time i check them was ok.
Change boths for a 1k, 5w new ones make a big improvement in clarity, touch feeling  and HEADROOM...

.- Try several options in the ef86 wiring layout.

    100k for plate load and upgrade the .1uf screen cap to .047 but instead to ground i conected it to cathode.
    Tryed 3 diferent configurations for the screen voltage. "D" node voltage is 280v.

1rst option:
390k from D node. Nothing new. I saw this configuration in a 70´s Philips datasheet. Stable, clean and bright with reduced mids. Lot of dinamic control. Hi-Fi sound like...
Plate voltage: 132v **
Screen voltage: 155v **

2nd option:
470k to screen and another 470k to ground. Voltage divisor. Less headroom, no soo bright and a lor more mids.
Plate voltage: 194v
Screen voltage: 94v

3rd option:
470k to screen. No voltage divisor. Compresed, a little harsh and the less headroom.
Plate voltage and screen voltage in similar values: 165v aproximately (i forgott write the exact values in this test )


** I read in a lot of places that screen voltage should be lower than anode one. In the first option this rule is really forgotten...but amp sound ok, without strange noises. I had test it four 2 hours...
I suppose there are too many opinions and, some times, erroneous information in the net.
Option 1 and 2 are the best for me. Both options sound ok, just diferent. Maybe i install a swicht before take a definitive decision.
As Hannibal says "I love it when a plan comes together!" or "Me encanta que los planes salgan bien!"

Jojokeo, thanks a lot for your conments. It is a great thing take place where to be able to speak about these things... Where I live, sometimes the people look at me as if i was.....




I´ll post a definitive schematic when i finish with all the tests...  (In 2-3 years, maybe.... )

[chocopower]

I edit the tittle.

I working too in a JSchem schematic.

Stay on tune.

[chocopower]

Here we go!

I tested two diferent configurations for the EF86 and both sound good to me, so i´m gonna try a switch for more versatility.
The  "Clean/Overdrive switch" is not mounted and tested. If any one see any posible problem due the high voltages involved, or posible hum/noise problems, please tell me.

Ahh! This is my first JSchem´s. What a nice tool!!! My Windows Paint times are gone!

Un saudo!

EDIT: NON FINAL SCHEMATIC. UPDATED ONE, TWO POSTS DOWN.

[tubenit]

Thanks for sharing the schematic! When you get it all tweaked to your liking please give us an updated schematic.
At that juncture, we can put this in the Library of SCH schematics.

Best regards, Tubenit

[chocopower]

Hi, again.

Right now, this is the schematic.

Hanging points and conclusions:

1.-When i switch to "Overdrive" mode in the EF86, i need move 1 position in the rotary switch to cut the extra bass. As posible solution, i´m thinking in use a lower cap for this configuration (1uf to 10uf) and add the 22uf one in the "Clean" mode. Easy. Next day test.

2.- A 4 position rotary switch is not enought. Across the whole volumen range there are some positions when the 4 positions don´t make a good bass adjust. I´m gonna install a 6 position one with the DC30 similar or identical values.

3.- The cathode bias attenuation is "jodidamente" great!!!! 40 to 50% less volumen with practical the same sound. I just move back one position in the Bass Cut to offset the slight loss of bass.

4.- I avoid the tone control in the preamp. Most of times it was dimmed and stealing a lot of signal. The "Cut" control after the PI work quite well, but is not useful in the whole range. I have seen 1M and 250k pots making the work, but i´m afraid about the 250k cut too much signal even in the max treble position. Any experience about this?

5.- As i use a pre-existent power amp for this proyect I did not change scarcely things in it, but the great difference between filter stages continues being strange for my.
A and B: 100uf.
Then it decrese a lot for 16uf in the C and D nodes.
Something like this could be a more equilibrate configuration:
A - 50uf
B - 50uf
C - 32uf or stock 16uf
D - 32uf or stock 16uf.
Worth a try?

Thanks everyboby,

Un saudo!!

[PRR]

> I read in a lot of places that screen voltage should be lower than anode

Not true for pentodes.

Books start by explaining TETRodes, then explain the problem when plate/anode voltage falls below screen voltage. This is WHY the Pentode was developed, and why true tetrodes were used for just a few years until pentodes arrived.

With pentodes, you "usually" get better results with plate at least 20% of screen.Yes, a pentode will work fine with screen at 300V and plate at 70V. For more gain you want a lower screen voltage; for bigger output you may want a screen voltage as high or higher than plate voltage.

> "Me encanta que los planes salgan bien!"

Very true.

> A and B: 100uf.
Then it decrese a lot for 16uf in the C and D nodes.
Something like this could be a more equilibrate configuration:

Look at the currents. Node A supplies 100mA. Node D supplies just 0.5mA. It makes sense to use a smaller cap for smaller loads. We could probably do some hard thinking and compute "exact right" values for all caps. I've seen as little as 0.1uFd for your node D. However a 16uFd is low cost and works well.

[DummyLoad]

horita e visto a sus planos y creo que va hacer mucho bien ruido  - salud a ti señor! :-) tengo un pregunto, ?que es un gañan.

[chocopower]

> I read in a lot of places that screen voltage should be lower than anode

Not true for pentodes.

Books start by explaining TETRodes, then explain the problem when plate/anode voltage falls below screen voltage. This is WHY the Pentode was developed, and why true tetrodes were used for just a few years until pentodes arrived.

With pentodes, you "usually" get better results with plate at least 20% of screen.Yes, a pentode will work fine with screen at 300V and plate at 70V. For more gain you want a lower screen voltage; for bigger output you may want a screen voltage as high or higher than plate voltage.

After a lot of tests, i agree... Obviously, i´m not a Tech, but as you says, i could ear for my self as the higher screen voltage, the higher volumen and heardoom.
Thanks for your theory aclaration.
-----------------------------------------------------------------------------------------------------------------------------------------
> A and B: 100uf.
Then it decrese a lot for 16uf in the C and D nodes.
Something like this could be a more equilibrate configuration:

Look at the currents. Node A supplies 100mA. Node D supplies just 0.5mA. It makes sense to use a smaller cap for smaller loads. We could probably do some hard thinking and compute "exact right" values for all caps. I've seen as little as 0.1uFd for your node D. However a 16uFd is low cost and works well.

mmm.. i have to go deeper in this just beacuse is a bit out of my capacity. i have to study....
The 16uf nodes in my amp are made with two series 32uf caps (when i did the cap job is the ones i had in the workbench), so make a test with 32 uf is easy.
I´m more afraid about the double 100uf nodes....  But it have been working with them since 1963, so i supose is ok..
And as all you say a lot, "if it ain't broke, don't fix it!" (ore something similar...).

thanks!!

[chocopower]

horita e visto a sus planos y creo que va hacer mucho bien ruido  - salud a ti señor! :-) tengo un pregunto, ?que es un gañan.

Hahaha!!

Hello! A "Gañan" is a Spanish local TV joke. Search Youtube........

[frus]

> I read in a lot of places that screen voltage should be lower than anode

Not true for pentodes.

what about beam power tetrodes?

[chocopower]

here we go..

last one (i hope so...)


http://picasaweb.google.es/davidageitosfuentes/GreenBullet?authkey=Gv1sRgCN6Bwp3Ql576Lw#5531371600280562306

[chocopower]

Some voltages measured.

I´ll like some conments/corrections about it, mainly in the EF86.

I know bias is really cold for EL34, but this way i can use 6V6 too.
thanks in advance.

[tubeswell]

Hi chocopower

I note some similarities between your tone stack an the matchless DC30 EF86 channel, except the latter has largish (1M5? or 5M1??) resistors between each of the caps for switching noise JM2CW

[chocopower]

I note some similarities between your tone stack an the matchless DC30 EF86 channel, except the latter has largish (1M5? or 5M1??) resistors between each of the caps for switching noise JM2CW

Hi,

I tested the 220k/1M5 used in the DC30 but i didn´t like it so tested diferent values.

Yesterday used it in a band practique for very first time.
While i was really proud about the sound while play it alone, the sound mixed with all the band isn´t soo satisfactory.
I´m using a Marshall cab loaded with a V-30 and a Heritage(greenback like).
To much bass saturation and fuzz.
I´m gonna try a lower value for the EF86 cathode cap, like Tubenit use in all his designs.

[kagliostro]

What about to connect a pot (connected as rheostat)

between the cathode bypass condenser and ground ?

if the pot is shunted than the cap apply all his effect

increasing resistance give less effect to the bypass cap

Kagliostro

[chocopower]

What about to connect a pot (connected as rheostat)

between the cathode bypass condenser and ground ?

if the pot is shunted than the cap apply all his effect

increasing resistance give less effect to the bypass cap

Kagliostro

Wich value for the pot/trimmer? I supose 10 times the cathode resistor...

[kagliostro]

Try with a 5k or 10k pot

Kagliostro

[chocopower]

hi there!

new schematic.

Now is totally usefull for single and humbucker guitars.

[tubenit]

That is VERY COOL !    Thank you!

With respect, Tubenit

[chocopower]

Thanks!

The UL/Normal switch is ALWAYS in the "normal" posittion (i dont know how to call the "no UL" posittion), so i suposse i´ll quit the switch and wire it as it should be.
I decided close the amp and leave it in my band´s local so i could stop to change things on it (for a while, at last...)
Next time ill put my hands on it i´ll post the voltajes and wanna try your enhanced caps in the P.I. Could be a good idea test it in a cathodine.