AA864 Musings

[sluckey]

100 Ohms from home with this project.  Sorry.

You can replace that 100Ω with a wire jumper and start playing music. Doing so simply means there will be no NFB until you can put a 100Ω in the amp.   

[dwinstonwood]

OK...

I'm always nervous about this. Let's see: [After adding the jumper] With no tubes in I'll flip the switch on the dim bulb tester. If nothing loud, flame-like or smelly happens I'll turn it off and plug it into the wall. Then, I'll hook up the multimeter to to measure ACV on the heater pins. Repeat the process to measure pins 3 to ground, and pins 4 to ground on the Octal sockets. Do the same for pins 1 & 6 on the 9-pin sockets. (I've already measured the cathode pins to ground and I have the correct resistance measurements for each socket.)

What I don't know is the negative bias voltage measurement I should want to see  on pins 5 with the bias pot turned to maximum negative voltage. I want to be sure it's low enough before installing tubes.

[sluckey]

What I don't know is the negative bias voltage measurement I should want to see  on pins 5 with the bias pot turned to maximum negative voltage. I want to be sure it's low enough before installing tubes.

Without the lamp limiter, turn the bias pot for maximum negative voltage on pin 5 of both 6L6s. Hopefully you will have at least -50V. If so, then plug the tubes in.

[dwinstonwood]

Thanks!

[dwinstonwood]

I'm going slow with this.

7.0VAC on the heaters

6L6 #1 pin 3 - 490.5 VDC
6L6 #2 pin 3 - 492.7 VDC

6L6 #1 pin 4 - 493 VDC
6L6 #2 pin 4 - 493.2 VDC

However - with no tubes in - I'm measuring in the 480's on the 9-pin plate sockets (1&6). Is something wrong there? I don't want to proceed if that's bad.

Thanks!

[dwinstonwood]

However, I am getting a range from -47.85VDC to -36.04VDC on both #5 pins.

[sluckey]

However, I am getting a range from -47.85VDC to -36.04VDC on both #5 pins.

May need to tweak that range a bit. But for now leave it set to -47.85 and plug in the tubes and speaker.

[dwinstonwood]

Cool. I'm eating lunch first - mainly to clear my brain and re-focus, and then on to installing tubes...

[dwinstonwood]

It will only take me about 3 minutes to swap the 470 with a 1K. It's a quick fix (assuming that's the best resistor in the bias circuit to tweak).

[dwinstonwood]

With the original 470 Ohm bias resistor still in the circuit I'm measuring across the 1 Ohm resistors and then, the plates:

Tube #1 (outer tube) - 0.049VDC (49ma)
Plate - 463VDC
22.687W - 75%

Tube #2 - 0.047VDC (47ma)
Plate - 463.2VDC
21.77W - 73%

I guess my math is right? I multiplied the plate voltage by the milliamps.

[dwinstonwood]

I'm wondering if my 12_7's are running too high on the plates:

1st 12AX7: pin 1 - 282.9, pin 6 - 281.3
2nd 12AX7: pin 1 - 286.2
12AY7 PI: pin 6 - 261.3, pin 1 - 253.2

At any rate... it sure sounds great. On "3" with the gain on "5" and my guitar volume 1/2 way up it's loud in my room. I'm playing it through the 6G3 75 watt speaker.

Oh, and at the lower volume (I haven't cranked it) it's extremely quiet. Wow. I need to chill for a bit and just enjoy the result.

[sluckey]

It will only take me about 3 minutes to swap the 470 with a 1K. It's a quick fix (assuming that's the best resistor in the bias circuit to tweak).

No. That would just decrease the bias voltage and you need to increase it to about -55V when the pot is at one end. Measure the voltage at the junction of the diode and the 10µF cap. What have you? We'll be messing with that 15K and/or 27K that's connected to the bias pot to get the bias voltage a bit higher.

As for the high voltages on the little tubes, you'll need to increase the size of the 1K and 4.7K dropping resistors on the B+ rail. But first, measure the voltage at nodes A, B, C, and D. What have you?

BTW, Good looking amp.

[dwinstonwood]

Thanks sluckey. Yeah, I went back and re-read Hoffman's bias instructions and I see know that raising the 470 would make the circuit less negative.

So, should I take these measurements with the tubes in or out?

[dwinstonwood]

The diode/10uF cap junction measures -69.3VDC. That's without the tubes in. I'm guessing it shouldn't make a difference.

[dwinstonwood]

With the tubes in:

cap can A - 468.4VDC
cap can B - 468.1VDC
directly in front of/before first 22uF cap - 460.3VDC
directly in front of/before second 22uF cap - 440.6VDC

I might not have measured at the right points.

[dwinstonwood]

The schematic shows 410 at the third node and 380 at the fourth. So, I'm roughly 50-60 volts high on these nodes. Nodes one and two are 422 and 420, respectively, so about the same amount high.

Also, the schematic shows 235 on the first two triodes, 240 on the third, and 220 & 205 on the PI.

It seems like all of these current production transformer manufacturers attempt to make exact replicas of the originals, which is fine, but they should have - IMO - engineered today's wall voltages into their products. I know it would drive the prices up, but it would be nice to have both "vintage correct" and "current-day" secondaries. With 10% to 15% lower secondary voltages you could build vintage amps with vintage-correct plate voltages.

[dwinstonwood]

I've been trying to work on this, but I can only get so far.

The three preamp triodes all have 1500K cathode resistors. If I divide each of the schematic's cathode voltages by 1500 and add it all up I get a total of 3.9ma of current draw for the preamp.

I want to drop about 50v. So, that indicates an extra 12.8K of resistance.

The problem is I want to drop the PI voltage, too. And, I don't know how to calculate the 12AT7 current draw from the LTP in the schematic. Plus, that tube comes before the preamp, so I really need to drop that voltage first before I can calculate the preamp resistance I need.

There's always the blind stab in the dark method. I could replace the 1K with the 4.7K, and replace that with a 10K and see where that gets me.

[dwinstonwood]

Well, if the voltage drop across the 1K resistor is 8V (468 to 460), then the PI is drawing 8ma. So, 50/.008=6250.

I think I have a 6.8K 3W.

Then, I only need to drop the preamp voltages by 30V. So, 30/.0039=7.69K. I think I also have an 8.2K 3W.

That may be a better place to start.

The JJ 6L6GC's can take up to 500V on the plates, so 468 seems ok. But, the screens are rated at 450V. Could I just increase the screen resistor values? It can't be that simple.

[sluckey]

Quit thinking of this amp as an AA864. What you really have is an AB763 with an extra gain stage. More like a Super Reverb with the iron you have. Don't worry about the screens. They are fine.

Now, for the fun tweaking  phase...

Here's a plan. Tackle the bias voltage first. Pull the 6L6s until you have approx -55V on pin 5 with the bias pot at max. First, change the 15K to 10K. Recheck pin 5. If you still don't have approx -55v then also change the 27k to 33K. Should be getting closer to the target. Continue tweaking these two resistors until you are satisfied with the pin 5 voltage. Notice that one resistor increases the bias voltage as the resistance goes down and the other resistor increases the voltage as the resistance goes up. Should be pretty easy to hit the target. Now plug in the 6L6s and set the bias to 70% or even 60%, whatever sounds good to you.

Second... Set the PI voltage for what you want. Increasing the 1K will decrease the PI voltage. Once you are happy with the PI voltage, move on to step three.

Third... Recheck the preamp voltages. You just may be where you want to be now that you changed the PI voltage. (Changing the PI voltage also changes the preamp voltages). But if still too high for you, increase the size of the 4.7K until you are happy.

Done!

Don't get hung up on some magic voltage you have seen on some schematic. There's really no right or wrong here. Just adjust things until you like the way the amp sounds. Make it your own!

Oh, almost forgot... When you finally replace that jumper with a 100Ω resistor your NFB loop will be working. There's a 50/50 chance that the amp will squeal or howl. If so, just swap the two white wires that connect to pins 1 of the 6L6 sockets, but do the actual swap on the board. Very easy to neatly swap on the board since they are right next to each other.

BTW, how does the extra gain stage sound? This could be a place to try some different tubes for different gain levels. Try 12AT7, or 5751, or 12AY7, or even a 12AU7. All have different gains and you may find you have a preference.

Enjoy.

[PRR]

> I don't know how to calculate the 12AT7 current draw from the LTP in the schematic.

Ohms Law is the law. Obey it. Use it.

There are enough numbers on the factory plan to compute essentially "all" currents. They don't quite add-up right, because the draftsman didn't know we would be checking. But they agree to better than 20% which is golden in most tubewerk. Yours will be different again, but you can do the same figuring.

[dwinstonwood]

> I don't know how to calculate the 12AT7 current draw from the LTP in the schematic.

Ohms Law is the law. Obey it. Use it.

There are enough numbers on the factory plan to compute essentially "all" currents. They don't quite add-up right, because the draftsman didn't know we would be checking. But they agree to better than 20% which is golden in most tubewerk. Yours will be different again, but you can do the same figuring.

Thanks PRR! Up in post #117 I actually did use the voltage drop across the 1K resistor (8V) to calculate 8ma current draw at that stage. But, you're right. I'm not always approaching this amp stuff from the right vantage point.

People here have told me before that getting too zoomed in on one issue is a problem. It wasn't until I stepped away for awhile that I started thinking about Ohm's Law. Then a lightbulb went off. I was obsessing over differences in wall voltage and current-production TX's, instead.

[dwinstonwood]

OK, sluckey, here's where I'm at:

I installed a 6.8K resistor in place of the 15K. This gives me a maximum of -56.07VDC on pins 5 with no tubes in. I didn't have a 10K or 8.2K on hand. But... I know I like the sound of cooler running power tubes - 60-ish% -  so I'm thinking I'm fine with the 6.8K.

With tubes in:

I'm now at 477.4VDC on the 6L6GC plates. Still within spec.

I replaced the 1K B+ dropping resistor with an 8.2K. I left the 4.7K in place.
Here are my 9-pin voltages. Much better:

V3 (PI) pin 6 - 238VDC; pin 1 - 228VDC
V2 (gain) pin 1 - 257.9VDC
V1 (pre) pin 6 - 253.5VDC; pin 1 - 254.3VDC

I didn't have a 6.8K. I think I'm ok for now. I'll probably order some more resistors and I might try a 6.8K instead of the 4.7K. That would drop the first three triodes down a bit more.

OK. I guess the next step is to bias the power tubes. It looks like I want about 38ma to be in the 60% range, or 44ma for 70%.

Onward.

[sluckey]

I installed a 6.8K resistor in place of the 15K. This gives me a maximum of -56.07VDC on pins 5 with no tubes in. I didn't have a 10K or 8.2K on hand. But... I know I like the sound of cooler running power tubes - 60-ish% -  so I'm thinking I'm fine with the 6.8K.

I think you're good to go on the bias. But before you put the tubes back in rotate the bias pot to the other extreme. What is the voltage on pin 5 now?

Then return the bias pot to -56v, plug in the tubes, and set the bias where you want it.

[dwinstonwood]

sluckey, I forgot to measure the sweep with the tubes out.

But, I'm there!

If you picture the bias pot as a clock face it's turned to between 11 and noon, just below halfway.

That gives me 38ma on each grid.

New plate measurements are 468.2VDC and 468.1VDC

That's right on 60%. 17.79 watts from the outer 6L6 and 17.78 watts on the inner 6L6. Is that good enough?

Thanks so much sluckey for your interest and help with this project! And, thanks to everyone else here who helped me through this. This forum is great. 

I'm basically the monkey who can soldier in resistors and use a multimeter, but I really don't know how to make electricity perform the acts that I want it to.

[sluckey]

So, what do you think about the extra gain stage? Is it controllable?

That gives me 38ma on each grid.

Surely you mean cathodes?

[dwinstonwood]

Yes, I meant 38 millivolts across the 1 Ohm cathode resistors! The adrenaline was up when I posted.

As for the gain control. I haven't cranked things up too far yet (Sunday morning is not the best time where I live). But, with the volume on 4 and the gain on 7 it's not unlike having a built-in overdrive. In fact, I'm getting sounds that are very similar to the BYOC TS-808 clone pedal kit I built. I need to play with it more to hear all the various settings. Once I get a good idea of the gain range I'll decide whether or not it needs a resistor instead of the jumper. Also, I installed the 100 Ohm NFB resistor, so that's hooked up, too.

With gain down around 2-3 and the volume up it's nice and clean. This amp is definitely not too bright. I'm setting treble at 6 and bass at 3 with bright off and it's very balanced and full sounding. But, it can be made bright with the switch pulled out and using the bridge PU. The push/pull bright switch is like the normal and bright inputs on my 6G3. But, I obviously don't have to unplug and plug back in to change it. Thanks for that addition. It's convenient.

So far, I find having the gain around 4-5 and volume around 3-4 is nice. A hint of overdrive that's controllable by string attack.

I'll report more after playing some more. Thanks!

[dwinstonwood]

BTW, how does the extra gain stage sound? This could be a place to try some different tubes for different gain levels. Try 12AT7, or 5751, or 12AY7, or even a 12AU7. All have different gains and you may find you have a preference.

After I play it more I may want to try different gain stage tubes. I was just reading the JJ 5751 description:
https://www.jj-electronic.com/en/5751
At $13.95 it's a cheap experiment to see how less gain/more headroom would sound. Though, I don't really need maximum clean headroom volume from this amp. Someone else might, so it's definitely a tweakable build.

It's a versatile amp from what I can tell so far. I'm only speaking of bedroom volumes as of now, but with the gain at 2 and the volume on 4 it has a very clear, smooth tone. Really nice, not hard or harsh, but plenty of treble. As I mentioned, reverse those two settings (say volume on 2-3 and gain on 4-5) and even at that low volume there's some overdrive in the sound.

Initial impressions are that it's more and better than I was expecting. Though, it could all fall apart and sound horrible when really cranked. I guess a lot of that will depend on the speaker(s).

[shooter]

gain at 2 and the volume on 4

I'm home alone; both on 7, guitar 6-10, you play longer when it's louder 

[dwinstonwood]

I'll be in that situation tomorrow, when the neighbors aren't home.

[dwinstonwood]

A morning of tweaks and fiddling.

One last change to the B+ dropping resistor: preamp plates range from 245VDC to 249VDC. If there is an audible difference in 10 to 15 volts either way, my ears aren't going to hear it. Done.
I added the tube clamps on the octal bases, tubes are nice and firmly in place now.
I'm not sure I'm in love with the push/pull volume pot. It does nothing until around 2 or so, and if I wiggle it I get some scratchy noise. I double checked all solder joints, and they're shiny, solid and thorough. I think the mechanical design of this pot has some play/wiggle in it. I might just go to an audio taper and ditch the bright feature. My Tele and my ears don't really need it anyway. I'd like to have the most precise control below 50%, especially for playing at night.

I've been poking around in my 6G3, too. I added two more shielded wires, bent some signal and power wires around, and redid both grounding points. This amp is now noticeably quieter than it's ever been. I'm not 100% sure exactly what I did, but I'm not messing with it anymore.

Finally, I swapped the 100K grid resistor in my AC4 for a 56K (sluckey's AC15 schematic shows a 33K, but I didn't have one). It sounds better/brighter. Probably like going from a low input to a medium-high input. I think it sort of woke up the EF86 from being a bit drowsy. I also replaced the tan grill cloth with Vox diamond cloth, so it now resembles a Vox.

I built these three tube guitar amps (my first ever) over the last seven months with the help and guidance of you all here. All three amps and great, each with it's own unique character.

Thanks!

[sluckey]

One last change to the B+ dropping resistor

So what values do you have for the two B+ dropping resistors?

And what voltages do you measure for nodes A, B, C, and D?

And what is your input line AC voltage?

And which OT and PT are you using?

[dwinstonwood]

Thanks sluckey, here's the data:

1st dropping resistor - 10K 3W
2nd dropping resistor - 10K 3W

Node A - 471.8VDC
Node B - 471.1VDC
Node C - 395.4VDC
Node D - 359.2VDC

AC Line In - 123.2VAC

PT - Classic Tone 40-18005 (330-0-330 @ 250ma; 6.3V @ 4A; 50V bias)
OPT - Classic Tone 40-18006 (4K Ohms; 4,8,16 secondaries)
Choke - Classic Tone 40-18003 (4H @ 90ma; D.C.R. - 105 Ohms)

David

[sluckey]

Here's the final as built drawings. They include your changes and voltage readings.

[dwinstonwood]

That is really generous of you sluckey. Not just the documents, but your oversight/corrections, guidance, and shared knowledge in this project are much appreciated. I learned a lot with this build. And, it was great fun.

I have an amp that looks good, sounds great, and can go from "AB763 bright & clean" to dirty with no pedals.

 

[sluckey]

FYI, the AB763 Tremolux and Vibrolux (same amp but different speakers) use 10Ks for B+ droppers. But B+ is a bit lower because of smaller PT and tube rectifier.

[dwinstonwood]

Cool. That's good to know that my choices are not unheard of values in Fender amps.
[With the 22uF caps that makes two low pass filters with a cutoff of 0.7237984945 Hz. Merlin's website says we want the cutoff below 1Hz if possible for B+ filtering, so in that respect the 10K's look to be a good value.)

Not to bemoan current production transformers again, but the schematic indicates a PT with 305-0-305 secondaries. The one I bought is sold as a 125P7D replacement, yet it's 330-0-330.

But, all of my plate voltages are now good, so I'm not really complaining!

[dwinstonwood]

I spent three hours at a local recording studio this morning with the amp. I played my Tele, a Gibson Les Paul, and even a Martin acoustic with a clip-on Fishman.

With the gain on 10 and the volume around 6 it is very loud through that Marshall cabinet, and it's a big room. To my ears it really mates well with the Fender. I have a 5751 in the gain location now. It's crunchy but not what you'd call real distortion (I didn't dime the volume). I need to go back and try a 12AX7 next weekend. Though, it sounds great clean with a bit of reverb. And, I'm glad I installed the speaker selector switch; the cabinet was 16 Ohms.

[dwinstonwood]

Done.

I ending up using grounded RG174 for both the send and return leads to the gain pot. I'm using a 5751 for V1 (something tubenit suggested in another thread). And, I repositioned the OPT primaries up away from the heater wires.

It's now the quietest it's ever been. There is a low hum when you lean close to the amp, but not any more than my other amps (I might check the PT bolts for tightness).

So, thanks to all the help here - in fact, sluckey more or less built this amp remotely - I totally consider this project a success.

And, it's relatively compact: 24" high, 18" wide, and 10" deep with a 12" speaker.

Since it didn't cost me anything (used the laser engraver at work) I made a back panel, too.

[dwinstonwood]

Sorry to keep resurrecting this thread.
I took the amp to a studio again tonight to wring it out. I'm really happy with the way it sounds. Adding the mid control was worth the challenge it caused me - a simple mod for most here. I can really get the most out of a range of different guitars now.
I couldn't be happier with it.