Fryette Effects - SAS EF86 Overdrive

ppluis0 · Post by **ppluis0** » 22 Aug 2012, 00:17

All this is very strange.

Mat, you might want to start to build on a breadboard only that part of the circuit comprising the input jack to the pair of clipping diodes.

If that setup works, you would have more nodes to measure, compare, and get some idea of what may be happening.

Unfortunately I haven't no more clues to try to guide you ...

Keep us informed to see if someone else on the forum is interested to work with you in this debug.

Cheers,
Jose

mat · Post by **mat** » 24 Aug 2012, 05:36

Ok, found what was wrong with the IC. There was no connection between pin2 and the 22pF cap. Now the voltages seem to be ok.

One thing left that make me wonder is the high voltage. On the schem it says 300V+ before plate resistor but I have 235V on that spot. I think this is because of the transformer I used. Would it be ok to swap the 10K/1W (or put in parallel) on the power filter caps to say 1k/1W ones to get higher voltage to the tube ? What would be the advantage - more headroom? Tube would be happier? At least the voltage on the plate would be closer to the original.

We are middle of moving our house (1.9.) I have packed all recording equipment but will record after we settle down to the new home. I'll record something today with my iPhone (if the circuit actually works).

Mat

mat · Post by **mat** » 24 Aug 2012, 11:52

iPhone sound sample (mindless noodling):

I did put 2 3K3/3W resistors in parallel with the existing 10K ones and here are the voltages:

pin1 103V
pin3 2,425V
pin6 125,8 to 171,5 (bias pot at extreme settings)
pin9 3V

B+ before the plate resistor 263,8V

The pedal seems to be noisier than the first try. I'll try to move the wires to get it to behave

Do You think the EF86 can handle the bias range ?

Eaven I have got a flu/fever - I cant stop playin the pedal

BIG, BIG thanks to all contributing/helping me on this process

ppluis0 · Post by **ppluis0** » 25 Aug 2012, 00:26

mat wrote:Ok, found what was wrong with the IC. There was no connection between pin2 and the 22pF cap. Now the voltages seem to be ok.

Congratulations !!!!!

mat wrote:One thing left that make me wonder is the high voltage. On the schem it says 300V+ before plate resistor but I have 235V on that spot. I think this is because of the transformer I used. Would it be ok to swap the 10K/1W (or put in parallel) on the power filter caps to say 1k/1W ones to get higher voltage to the tube ? What would be the advantage - more headroom? Tube would be happier? At least the voltage on the plate would be closer to the original.

Mat, remember that some values I draw in the schematic were merely assumptions. If the supply voltage isn't exactly the same as the original design I think that the sound will be almost the same.

Obviously there will be different headroom so you must be tweak the potentiometers accordingly, but that's all.

Cheers,
Jose

ppluis0 · Post by **ppluis0** » 25 Aug 2012, 00:43

mat wrote:iPhone sound sample (mindless noodling): http://soundcloud.com/mattit-1/diy-ef86od-pedal

HEY... nice sound clip !!!!!

mat wrote:I did put 2 3K3/3W resistors in parallel with the existing 10K ones and here are the voltages:

pin1 103V
pin3 2,425V
pin6 125,8 to 171,5 (bias pot at extreme settings)
pin9 3V

B+ before the plate resistor 263,8V

The pedal seems to be noisier than the first try. I'll try to move the wires to get it to behave

Do You think the EF86 can handle the bias range ?

After a little of math based on the numbers you posted, seems as the current flow across the tube varies from 0.9 to 1.4 mA, wich is lees than the maximum limit for this model, rated at 3 mA at max.

mat wrote:Eaven I have got a flu/fever - I cant stop playin the pedal

BIG, BIG thanks to all contributing/helping me on this process

Hope you recovers soon, and good luck with your moving.

Cheers,
Jose

mat · Post by **mat** » 25 Aug 2012, 15:01

Hi Jose,

Thanks for confirming the bias range. Now that I have played the pedal for few hours I noticed that there are lots of different sounds in it. The bias pot really changes how the pedal reacts playing. The bias at full has noticeable lower volume and really creamy breakup. The gain at zero the pedal is really cool boost pedal. On the end of the soundclip I adjusted guitar volume at 5 and played that chord melody - after that I maxed the guitar volume and played few chords and the most beautiful- (at least to me) melting-little-tweed-amp-sound came trough

Also really cool overdrive sounds can be found in there with twisting the three knobs. I''ll test the pedal tomorrow at the band rehearsals with my Dumble nonHRM and Marshall2555 clones

Mat

ppluis0 · Post by **ppluis0** » 25 Aug 2012, 18:22

Hi Mat,

Perhaps you can undo the last modification at the high voltage supply unsoldering the 3K3 resistors, and play again to see what happens.

After a look at the EF86 datasheet (located at: http://frank.pocnet.net/sheets/010/e/EF86.pdf ), appears that this tube can work as low as 150 Vdc and the main difference in numbers is the availability of clean signal output (that is exactly the kind of signal we try to avoid

)

Please tell me what type of diodes you use to clip the signal due I don't remember exactly if they are mentioned in this thread. I found that the original unit employs BAT41 some time after to post the diagram.

Finally, since this tube has a great amplification factor, try to shield it or place far from the transformer inside your unit to reduce noises and hum respectively.

Cheers,
Jose

mat · Post by **mat** » 25 Aug 2012, 20:26

Perhaps you can undo the last modification at the high voltage supply unsoldering the 3K3 resistors, and play again to see what happens.

I might try that but not right now - I'll play the pedal and get to know it better.

After a look at the EF86 datasheet (located at: http://frank.pocnet.net/sheets/010/e/EF86.pdf ), appears that this tube can work as low as 150 Vdc and the main difference in numbers is the availability of clean signal output (that is exactly the kind of signal we try to avoid )

You are right. I'll try to snip the extra resistors out after the house move. Wonder what kind of voltages would the 12V/120V transformer give with the type of rectifier in original.

Please tell me what type of diodes you use to clip the signal due I don't remember exactly if they are mentioned in this thread. I found that the original unit employs BAT41 some time after to post the diagram.

I used BAT41's.

Finally, since this tube has a great amplification factor, try to shield it or place far from the transformer inside your unit to reduce noises and hum respectively.

Thanks for the tip, The tube is in opposite side of the transformer.

Would be nice if someone having the original pedal could measure the voltages around the EF86

Mat

mat · Post by **mat** » 26 Aug 2012, 21:02

Ok, tried the pedal with band today. Great sounds and cuts trough mix very easily - I noticed I could play with lower volume than normally.

One thing that I would like to tweak is the trebliness (or lack of middle frequencies) creeping in when the gain pot is past 12 o'clock. I noticed I have two parallel 22pF caps on the gain pot. Should I swap a bigger value cap to get rid of the high frequencies ? On lower gain settings the treble content is perfect.

Mat

mat · Post by **mat** » 04 Sep 2012, 07:02

I am planning to put a switch to get a variable value cap on the gain pot. Also as much as I like the small tweed style bass breakup, I would like to have an option to switch different (input?) caps for more tight bass response. So, any suggestion what cap to put switchable (for smaller value)? The 47nF input, the 10uF
before the diodes or the 22nF before the EF86 ?

Thanks,
Mat

mat · Post by **mat** » 05 Sep 2012, 05:01

http://mattitee.kuvat.fi/kuvat/EF86over ... as+mod.png

Mat

ppluis0 · Post by **ppluis0** » 05 Sep 2012, 16:33

Hi Mat,

Perhaps you can place a capacitor in parallel with the couple of clipping diodes as the MXR Distortion+, wich employs 1nF to tame some harshness.

Due the low frecuency mods, the switches you mention surely can work.

Cheers,
Jose

mat · Post by **mat** » 05 Sep 2012, 19:40

Thanks Jose for the tip !

I'll start experimenting with my dip-switchable-variable-cap-box

Mat

mat · Post by **mat** » 26 Sep 2012, 18:26

Hi again,

I settled for 1uF and 3.3uF + 10uF on the switch for the cap before the diodes and 22p on the gain pot and 100p + 330p on the switch for the gain pot. Seems to work really well for different usable sounds.

My friend tried the pedal for few days and insisted me to build an another one for him. I'm half way there for the second build. Will let You know how it turns out.

Btw. which pot would suit better for the volume jump fix ? A linear (100kB) or a logaritmic (100kA)?

Cheers,
Mat

ppluis0 · Post by **ppluis0** » 27 Sep 2012, 02:12

Hi Mat,

Nice to read that the tone switch you implemented is useful.

If you have serious chances to build another unit, why not to try another pentodes different that an EF86 ? (You always can rewire later the tube socket to accept an EF86 if the resulting sound is non pleasant...)

Personally I think that several pentodes made years ago for use in radio o TV sets can be purchased in NOS condition and very cheaply. After all our intention is to make some noise, so several tubes not intended for Hi Fi can have good results, and others may not suit.

Some tube codes that appear as candidates are: 6AG5, 6AK5, 6AU6, 6BA6, 6BC5, 6BH6, 6BJ6, 6CB6, etc., speaking of american tubes.

As you live in europe, any tube named EFxx perhaps can be used. (The first letter E indicates 6.3 volts heater, the letter F is a denomination for small pentode and the remaining numbers are the particular model of each tube). Some of them require 7 pin socket, others needs 9 pin socket and there are a family wich employ a basing named Rimlock that is different of the most common miniature tubes, so be aware.

If you are brave enough to accept this challenge I'm gladly offer to assist you to select a suitable tube. Just tell me what you can find at your local flea market, at a garage sale or in someone's attic or basement, and think what can be done.

Cheers,
Jose

mat · Post by **mat** » 27 Sep 2012, 05:02

Hi Jose,

I can buy 6AU6WC from Your list but will use EF86 for this build because it is still current production tube and the build is not for me.

I've been thinking of a build that has Dumble style clean preamp (12ax7 on V1) and odd pentode (V2) for the overdrive section on a smallest possible box (maybe in a recycled PC power enclosure).

My fried saved for me some odd tubes from the trash bin. Any suitable candidates for the V2 ?
EF93 tungsram
EF94 tungsram
EF95 tungsram
6AL5
EB 91 tungsram
E180F philips
E188CC philips
6AT6
EBC90 tungsram
EL86 philips
EL90 philips

Which one do You think would give more even volume curve for the SAS build. A log or linear volume pot ?

Mat

ppluis0 · Post by **ppluis0** » 27 Sep 2012, 14:50

Hi Mat,

Usually the volumen potentiometers are logaritmic taper, also named "Audio taper". For this particular build I guess that the potentiometers to be used in the gain and bias functions works better if there are linear tapered.

Regarding the list of tubes you have, let me investigate what is each model and where can be employed.

Cheers,
José

ppluis0 · Post by **ppluis0** » 28 Sep 2012, 03:36

mat wrote:Hi Jose,

I can buy 6AU6WC from Your list but will use EF86 for this build because it is still current production tube and the build is not for me.

Is a logical desicion to employ a current type of tube if the pedal needs to be repaired in the future. I agree !!

mat wrote:I've been thinking of a build that has Dumble style clean preamp (12ax7 on V1) and odd pentode (V2) for the overdrive section on a smallest possible box (maybe in a recycled PC power enclosure).

Please post that preamp schematic, or tell me where can be found in the web, or (to not violate any copyright rights) please send me a copy via pm to study how that circuit works.

To use in that preamp you have listed three tubes that have exactly the same pinout and are interchangeably amongst them:

The EF93 is equivalent to a 6BA6; the EF94 equals a 6AU6, and the EF95 is the same as 6AK5. These three types have different gain slopes so you can do a sort of "tube rolling" in the same way as when we try different dual op amps into a TS derivative finding what colour of sound produce each one.

(I think that change one tube for another will be more notorious that try an 4558, a TL072, or a NE5532 successively into a classical dirt box....

)

The 6AL5 equals the EB 91, and a pair of them can be used in this interesting pedal: https://www.freestompboxes.org/viewtopic ... =spark+gap

The E180F is another type of small signal pentode with a rugged construction but has a different pinout than EF86 or the previously mentioned trio.

The E188CC is a twin triode (also rugged, shock and vibration resistant, etc.), with an amplification factor midway the 12AU7 and the 12AT7. I think that you can employ it succesfully in a clone of a Butler tube driver but the connection at pin 9 isn't related to a filament as the normal audio triodes and must be heated using only 6.3 volts.

The 6AT6 and the EBC90 are substitutes and each one is like a half of a 12AT7

The EL86 can be used as output tube for an single ended amp delivering close to 5 watts. The EL90 can produce 4.5 watts and is equivalent to a 6AQ5.

One of these pentodes drived with one of previous mentioned single triodes can be used to build something like this: http://niteglow.webs.com/ The funny thing about this design is that employ off the shelf regular supply transformers (also for the output transformer !!!) instead to use specially made ones.

Hope you can use some of the tubes you have into any of these suggested projects.

Or perhaps in all[/b] of these projects ??

Cheers,
Jose

mat · Post by **mat** » 29 Sep 2012, 19:10

Hi Jose,
Thanks for the very good explanation of the tubes. I will definitely build something out of them.
Got big inspiration from Your post. I will mail You the schem in near future.
Mat

mat · Post by **mat** » 29 Sep 2012, 19:42

Btw. I noticed that I did not put two 100R resistors to groud from the heater taps on my SAS build. Are the heaters now 'floating' ? What could be the problem with that ?
Mat

Fryette Effects - SAS EF86 Overdrive [traced]