AA764 with 6SL7 tubes. keep FMV TS or swap to James ??

[kagliostro]

The idea is about an AA764 but with 6SL7 tubes instead of 12AX7 and 1625 instead of 6V6

I've think to maintain the components around the 6SL7 the same of the 12AX7 on the original AA764 (is correct to do ?)

but 6SL7 tube has less gain (70) than the 12AX7 (100)

so as to compensate a bit of the lost gain is better to swap from FMV TS to a James TS or the difference will be subtle ??

Franco

[sluckey]

I built a Fender preamp using 6SL7 in my "Rocky" conversion project. Works fine and has plenty of gain. Notice the different plate and cathode resistor values.

This schematic shows a 6SN7 for V1 but V1 is really a 6SL7. See page 4.

     http://sluckeyamps.com/rocky/rocky.pdf

[kagliostro]

Thanks Steve

Some time ago I've seen that conversion and was thinking you used 6SN7 tubes

The AA764 vibroChamp original schematis has V1a Rp 100K + Rk 1.5k and V1b Rp 100k + Rk 1.5k (plus 47R)

Is your council to use the values on page 4 of your schematic

V1a Rp 270k + Rk 4.7k and V1b Rp 270k + 4.7k (plus the 47R) with the FMV TS ?

One other thing, the raw control is important or can be shun ?

Thanks

Franco

[sluckey]

Is your council to use the values on page 4 of your schematic

Those higher values work well. Ampeg used higher values for the 6SL7s. I just copied the values used in the original Rock Ola amp. I never used the standard Fender values. Those may work well also. Don't know.

The raw control is not necessary but it lifts the tone stack which decreases TS loss and allows you to over drive the power amp for some nice tones that you usually don't get with that tone stack. If you remove the 6.8K and use a 100K audio taper raw control, the first half rotation acts just like a mid control and the last half rotation acts like a raw control. I like it. But just connect the 6.8K to ground if you don't want the raw control.

[kagliostro]

Thanks again

All started some time ago when I was asking some indications for an amp that I'm planning

you pointed my attention on the AA764 vibrato circuit (I was asking for the right way to do tremolo in SE amp)

so, the amp I was planning was (V1) 6SJ7 > (CF) 6SL7 1/2a > James Tone Control > (Gain stage) 6SL7 1/2b > (power Tube) 1625

total 3 tubes (6SJ7 + 6SL7 + 1625), adding one tube (for tremolo) they became 4 tubes (6SJ7 + 6SL7 + 6SL7 + 1625)

but now I've seen the AA764 vibroChamp and with the same number of tubes I can have a two channel, an AA764 + 5C1 front

remembering Geezer (I hope he is well, is a long time I don't see new post) and Tubenit's TOS that uses a pentode as OD

this is (around) the new idea



BTW, do you think that a 40cm x 20cm (15.74" x 7.87") chassis will be adequate (there are a lot of pots) or I can go with a smaller one ?

Franco


p.s.: Do you recognize the input jumper switch ? 

[sluckey]

do you think that a 40cm x 20cm (15.74" x 7.87") chassis will be adequate

That's about the minimum size I'd use. And I'd space the pots/jacks 1.25" on centers. You'll have to put power switches, indicator lamp, fuse, etc., on the back.

p.s.: Do you recognize the input jumper switch ?

I do. Do you actually have some of those now?

[kagliostro]

To have those C&K is a bit complicated here, didn't find it at the local vendor or in other easily way, Mouser has, but my friends that have some time buy items from Mouser say that the shipping is 20€ (too much for my taste), however I've find some other brand ON-ON-ON switch and I'll order those

That's about the minimum size I'd use. And I'd space the pots/jacks 1.25" on centers. You'll have to put power switches, indicator lamp, fuse, etc., on the back.

Yes, of course, I forgot, there are also power switch, indicators etc. etc. ... it will be better I go for a longer one, maybe less deep (16-17cm instead of 20 ..... maybe), or I go for a higher measure and stack two rows of pots, something like in the amp was build from a my friend

http://www.diyitalia.eu/viewtopic.php?f=22&t=10281



Grazie

Franco

[HotBluePlates]

The idea is about an AA764 but with 6SL7 tubes instead of 12AX7 and 1625 instead of 6V6

I've think to maintain the components around the 6SL7 the same of the 12AX7 on the original AA764 (is correct to do ?)

but 6SL7 tube has less gain (70) than the 12AX7 (100)

so as to compensate a bit of the lost gain is better to swap from FMV TS to a James TS or the difference will be subtle ?? ...

What is your planned bias for the 1625?

You already know the 1625 is basically an 807 with a different heater voltage, which is basically a 6L6.  That implies a higher drive voltage is needed compared to using a 6V6.  Knowing the planned bias will also tell you how much drive signal is needed at the 1625's grid, allowing an assessment of the preamp's signal levels.

I'd have made some guesses about the 1625's bias, but only the screen voltage is shown on the schematic...

[kagliostro]

The PS is still in work, for G2 I've think to 250V and B+ will be around 350V and .... I'm not able to establish the drive signal required 

on the 807, if I remember correctly, a 16-17V signal is required, but I've no idea about the 1625 and math and me didn't go arm in arm



Franco


p.s.: The IRF820 is a capacitor multiplier, not choke or voltage regulator

[HotBluePlates]

The PS is still in work, for G2 I've think to 250V and B+ will be around 350V and .... on the 807, if I remember correctly, a 16-17V signal is required, but I've no idea about the 1625 ...

The data sheet you attached shows -18v bias for the plate & screen voltage you're considering, so that will probably be close enough.  When figuring signal levels for the rest of the amp, if you want to easily drive this amp to distortion then assume that you want to get an 18v peak signal with a relatively-weak signal at the input jack (20-70mV peak, or comfortably-less than your pickup's actual output).

You can get more-exact with gain predictions by plotting an operating point on the curves and measuring internal plate resistance at the operating point.  But let's just take the data sheet value of 44kΩ and assume it will be close.

Gain = 70 * [100kΩ/(44kΩ+100kΩ)] = ~48

2nd Gain Stage:
18v peak / 48 = 0.375v peak required at the grid.

Volume Control:
Assume 10% taper and a setting of ~8, so a loss of 0.3.  0.375v peak / 0.3 = 1.25v peak input to volume pot is needed.

Tone Stack:
The Duncan Tone Stack Calculator shows a loss of about -12dB overall and -26dB to mids for a typical Fender tone stack.  Convert to a loss ratio as 10^-(Loss/20).  So -12dB = 10^-12/20 = 0.25 (a 1:4 loss), and -26dB = 10^-26/20 = 0.05 (a 1:20 loss).  You can check your math by entering 20 * log * (1/20) into a calculator and getting about -26dB.

So for full output we need 1.25v peak / 0.25 = 5v peak for bass/treble, and 1.25v peak / 0.05 = 25v peak for midrange.  Though, to be fair the blackface Fender sound equates to this mid-scoop and relatively clean mids.

1st Stage:
5v peak / 48 = ~104mV peak, so this will be relatively clean or at least only overdriven by hot pickups with the Volume control set high.

You may wish to use a 25kΩ pot (or higher) for the Mid pot and/or add a resistor-to-ground which can be shorted out.  This would go between the Mid pot's ground connection and the actual ground.  It will for a Mid Boost which will reduce the overall signal loss in the tone stack, as well as boosting midrange.  Then you have the option of a blackface style clean sound and a more mid-heavy and distorted sound.
Some of the above also depends on what "perception of headroom" you want to have, meaning what volume pot setting you want to be the onset of distortion based on the strength of the pickups you plan to use.  The same amp circuit can seem "too clean" to some player if it doesn't break up until 9-10 on the Volume pot, but also to have "no headroom" to other players if the amp distorts by 4-5 on the Volume pot.

[HotBluePlates]

OOPS!!  The above totally neglected the negative feedback around the output stage on this amp!

The data sheet claimed 10.8w for the condition with 350v plate, 250v screen and -18v bias.  10.8w RMS across 8Ω is √(10.8w * 8Ω) = ~9.3v RMS or 13.14v peak at the speaker.

Treating the output stage as an opamp with feedback, the closed-loop gain is 1+(R_series/R_shunt) = 1+(1500/47) = ~33 (closed-loop)

The open-loop gain implied from the earlier post is
13.14v peak (speaker) / 0.375v peak (2nd Stage grid) = ~35 (open-loop).

It looks like your feedback loop won't do much of anything, even at the max feedback setting.  In a way that's good because the earlier analysis is unchanged (except that I assumed the Master Volume was at maximum).  And maybe PRR will double-check me, but it also appears that the original VibroChamp values applied almost no negative feedback to that amp (unless I did it wrong, it appears the closed-loop gain of the VibroChamp is higher than the open-loop gain, which means the feedback isn't reducing gain).

[kagliostro]

Thanks HotBluePlates

I've read your post three times and my brain is still computing

only one thing, the AA764 NFB uses only a 2.7k resistor instead of the 10k pot + 1.5k resistor of my schematic

this changes something on your considerations or you based it on the 2.7k value ?

---

About chassis I've think to one other option, a big Faceplate, higher than the chassis with rows of pot on it

as in some PA amp, this way it will be possible to "shrink" the chassis lenght



Franco

[HotBluePlates]

only one thing, the AA764 NFB uses only a 2.7k resistor instead of the 10k pot + 1.5k resistor of my schematic

this changes something on your considerations or you based it on the 2.7k value ? ...

I calculated based on your schematic, and the setting of the feedback pot for maximum NFB. That's why I used the 1.5kΩ and 47Ω values in the calculation.
Let's calculate the feedback a different way... By seeing how much feedback voltage is applied.

The AA764 will output ~4w across a 3.2Ω speaker, so speaker voltage is √(4w * 3.2Ω) = 3.58v RMS or ~5v peak.  If we assume the 6V6 does this with an input equal to the bias of 21v, and the 12AX7 stage before it has a gain of ~61, then that's 21v peak/61 = 0.344v peak at the 12AX7.  The feedback loop reduces the speaker voltage to 5v peak * [47Ω/(47Ω+2.7kΩ)] = 0.086v peak.  So that's 20*log (0.086v/0.344v) = -12dB below the applied signal level.

As calculated for your amp, the speaker voltage is 13.14v peak at the full ~11w output, and we found the input to the 2nd Stage is 0.375v peak to get that output power.  The feedback loop is a voltage divider, which applies 13.14v peak * [47Ω/(47Ω+1.5kΩ)] = 0.399v peak

Hmmm... That's more feedback voltage than signal voltage.  To get a feedback level 12dB below the applied signal level in your amp, you'd need 0.375v * 10^(-12dB/20) = 0.094v peak.  Using standard values, a 6.8kΩ series resistor and the 47Ω resistor-to-ground would comes very close (exact would be more like 6.5-6.6kΩ).

Again, someone who's more knowledgable on feedback should double-check me with these.  I'm a little confused that 2 different approaches came up with such different results, which probably means I'm screwing something up. 

[kagliostro]

I was wondering that all the situation can be changed if I adopt the Rp and Rk values used by Steve in his Rola conversion

with those values the gain will stay (I hope I do may math correctly -- 70 * [270kΩ/(44kΩ+270kΩ)] = ~60 that is more than  ~48

Franco

[HotBluePlates]

I was wondering that all the situation can be changed if I adopt the Rp and Rk values used by Steve in his Rola conversion

with those values the gain will stay (I hope I do may math correctly -- 70 * [270kΩ/(44kΩ+270kΩ)] = ~60 that is more than  ~48

Sounds like a good idea!  And you could always start with the larger series feedback resistor I came up with and reduce it if you find your max-feedback setting isn't polite enough.

[kagliostro]

Sounds like a good idea!

Oh, YES, Steve has often good ideas

And you could always start with the larger series feedback resistor I came up with and reduce it if you find your max-feedback setting isn't polite enough

This is one other good idea

Thanks

Franco

[PRR]

> applied almost no negative feedback

A fair number of git-amps have almost-no NFB at nominal load, but at speaker bass resonance where impedance is 5X-10X higher they have enough NFB to keep the output from rising the expected 5X-10X.

So this may be influenced by what speaker you use, what box, note-choices, etc etc. If you put a low-resonance speaker (Altec 417) in a large open box and never play your bottom octave, it may play the "same" with or without the NFB connected. With a high resonance speaker or a small box, bopping around the lowest notes on the frets, you may get some over-strong notes until you apply the NFB.

Yes, put in terminals for the NFB resistor and adjust it after all else is sorted.

Or put a 5K pot across the speaker, use a low-value NFB series resistor, and turn the knob from zero to heavy NFB.

[kagliostro]

Thanks PRR

Or put a 5K pot across the speaker, use a low-value NFB series resistor, and turn the knob from zero to heavy NFB.

 

Do you literally mean across the speaker for across the speaker ?????



Be patient, I had a long day

I was to the local hospital with my father (class 1924) from 18:00  till 01:00 because he had an cerebral ischemia ( a stroke) and now he is hospitalized

I had an interval because my sister in law coming out of my house forgot the key on the lock (security door course) so I had to go home

jump from one terrace to one other and with a 1kg bat break a shatterproof glass (not easy to do also because the jalousie was lowered)

so my brain that usually is a bit slow to understand at the moment is really out of order

Franco

[PRR]

> hospital with my father

Be with your father.

My parents are here, visiting, and healthier than me. They are no longer young though.

Our best wishes to your Dad.

---------------

Then ground one end of the speaker winding.

[Willabe]

Yes, best to your fathers recovery. 

[kagliostro]

Thanks PRR & Willabe

This morning has made a new electrocardiogram and another TAC (computerized axial tomography) scan and the doctor said that the electrocardiogram is OK and there are not new problems on the TAC, today is slow to coordinate movements and fatigued, but speaks better than yesterday and the mouth is almost back to normal, we also did a little walk in the corridors and this is positive

Thanks again

Franco

[PRR]

> TAC (computerized axial tomography) scan

FYI: in the US this is a "CAT" scan. (I just had one.) I guess in Italian vs English we put the modifiers in back or in front. (Also "cat" makes it sound soft and friendly.)

Slow and clumsy is to be expected (though must be distressing for you and for him). Speech improvement so soon is a very good sign. He lost some brain cells; if not too many then the brain is fairly good about re-wiring around damage. (I am not a Doctor.)

[kagliostro]

Yes, if there aren't new episodes, we hope that he will recover, we must be patient about that

Thanks PRR

Franco

[kagliostro]

Well, today he was discharged from hospital and is at home

He is not 100% but not so bad, he walk with his walking stick and a bit of help, just to make him feel safe

and talking is very good  also if not perfect

I'm very happy

Franco

[Willabe]

That is great news K! 

[kagliostro]

Thanks Willabe


Franco

[shooter]

jump from one terrace to one other and with a 1kg bat break a shatterproof glass

You might consider more logical access to your home, least you increase your BP more than it should be

My dad had a couple strokes, came through well, only long-term, 10+yrs, less confidence, otherwise, wit, humor, still DAD.   fwiw, I've switched to the james on my last 2 builds, and have it penciled into the next.

[kagliostro]

Ciao Shooter

I must have a mod on the lock of my house, the mod consist on the insertion of a friction on the lock cylinder, this mod allow to open the door (with the key) also if someone forget the key in the side inside the lock

--

Some time ago I've read that the James has a less load effect than the FMV, but I don't know in what order of magnitude is the thing

Franco


p.s.: BP on the Urban Dictionary has more than one meaning ...... to which one do you refer ? 

[shooter]

BP

Blood Pressure, we're not 20 and bullet-proof anymore

I don't know about loading, but I seem to be able to dial about any tonal range I want with 2 knobs vs 3.  less parts, less errors kinda logic.

[kagliostro]

Blood Pressure ...... AH, OK now understand

...... we're not 20 and bullet-proof anymore

There is something of true on what you say   

--

In the AA764 there are only two Tone Controls knobs, the middle is set with a resistor, however it is anyway a FMV

Franco