Built a single channel AB763 w/o Trem and am having some PI issues!

[guitylerham]

Hey everyone,

I've recently built a Deluxe Reverb style amp using the Hoffman schem and a donor chassis/xfrmr set from a hammond organ that I purchased a while back. A few issues I'm noticing:

1) signal makes it to the speaker even with volume on zero!

2) more importantly, my amp is not very loud and has minimal headroom. It's giggable but compared to a real DR it's nowhere near as full and punchy, not to mention clean. I've traced with an audio probe from the input all the way to the Phase Inverter and have pinpointed where the signal becomes way saturated. Attached is a schem of the PI section with some color coding added. GREEN is where the signal tests strong and clear. Yellow is where the signal becomes very quiet but still reasonably clear. Red is where the signal becomes very saturated and distorted.



I have the following voltages for your reference.

My power rail supplies are:
A = 458vdc
B = 458vdc
C = 378vdc
D = 336vdc

12AT7 PI
1) 211vdc
2) 50vdc
3) 82vdc
4) 3.2vac
5) 3.2vac
6) 217vdc
7) 54vdc
8) 82vdc
9) 3.2vac

I know my supply voltages are a bit high compared to the Fender schem but that's to be expected with a donor xfrmr and I don't think it's crazy high. The odd thing to me (and it may be due to my lack of experience) is when I probed the "cross" network of 1M/470R/1M/22K, the signal was great from the input to pin 2, across that top 1M resistor, across the 470R to the pins 3/8 but as soon as I probed the other side of the 22K and 1M resistors, the signal is greatly diminished. And then the strange thing is that the output of both sides of the PI are extremely saturated if I have the volume anywhere past 1 and even that isn't clean. What gives?? Anyone have some insight into where I should be directing my attention? I'm looking forward to hearing your thoughts. Thanks!

[ac427v]

Hi Guitylerham,
The one thing I noticed is that the 100 ohm negative feedback resistor is usually the value used with 2 or 4 ohm speaker loads. Fender used a 47 ohm resistor on amps with 8 ohm loads. So maybe you are getting too much negative feedback. You could check by tack soldering another 100 ohm in parallel to the existing one.
Good luck!
Craig B

[sluckey]

Disconnect the NFB wire from the speaker jack. Does it get better or worse?

[terminalgs]

1) signal makes it to the speaker even with volume on zero!

Whatever problem is causing this could be a big factor is the amp's general disfunction. your pre-amp signal is somehow finding an alternate path to power amp, maybe via the PI.  the only path we want it to be able to take is your volume pot.  So, it is either "leaking" through (1) The B+ supply (2) the ground supply or (3) extremely close proximity of components.  It's probably #1 or #2.  While the NFB resistors values are off (and I'd disconnect that as well as advised),  with your volume pot at zero, the signal should never get as far as the PI and the NFB circuit.


for #2,  the ground:  with amp off.  make sure you have 0ohm across all grounds. check all cathode resistors's ground sides, all your filter caps ground sides, pots, everything.  make sure all your grounds are good.   don't check from the solder lug (or eyelet), check from the lead between the cap or resistor and the solder joint.


As you turn your volume pot to zero, does the volume actually disappear at maybe "1" or "2" and then come back a little at "0"?   If so,  you've got a problem with the B+ supply in your preamp (bad solder joints, bad wiring, bad filter cap).

for #1,  you can check this with amp on., but you need to be careful, of course. first amp off, solder the (-) lug of a 10uf (or 20uf or 40uf) lytic cap to your ground bus,  attach an insulated probe to the (+) side.  turn the amp on with volume @ zero, and some input to the amp (tell a buddy to grab a tele and play Lady of Spain). Check every filter cap with the probe to the (+) side of the filter caps...do the same for all your plate resistors.   If you probe on the lead between the component and solder joint, you also check for bad solder joints...  You are listening for the spot when the probe's 10uf (+) fixes the B+ problem. Of course, when you probe a plate, it should go to mute, but you are primarily interested in the B+ side of the plate resistors.

[eleventeen]

You say on your 12AT7 PI you have 82 volts on the cathodes (which is reasonable) but you have 50/54 volts on your grids. That means, from a DC standpoint that your grids are 30 volts below your cathodes. My sense is that "setup" would hard cut off the 12AT7 triodes. Nothing would get through. That does not seem right.

Measure DC volts at that "cross" node where the 2 qty 1 M R's, the 470, and the 22K come together. According to this schematic http://el34world.com/charts/Schematics/files/fender/Fender_DELUXE_REVERB_AB763.pdf there should be a small 1.5 (or so) voltage difference between the common cathodes of the 12AT7 and that node. Check to see if that's right.

[sluckey]

You say on your 12AT7 PI you have 82 volts on the cathodes (which is reasonable) but you have 50/54 volts on your grids. That means, from a DC standpoint that your grids are 30 volts below your cathodes. My sense is that "setup" would hard cut off the 12AT7 triodes. Nothing would get through. That does not seem right.

Nothing wrong with that. You can't accurately measure the dc voltage on the very high impedance of the grids on that LTP because even a 10MΩ DMM will significantly load the circuit.

Measure DC volts at that "cross" node where the 2 qty 1 M R's, the 470, and the 22K come together. According to this schematic http://el34world.com/charts/Schematics/files/fender/Fender_DELUXE_REVERB_AB763.pdf there should be a small 1.5 (or so) voltage difference between the common cathodes of the 12AT7 and that node. Check to see if that's right.

That's better. The voltage at that node is the same as the voltage that's really on the grids.

[guitylerham]

Ok, I just got back today and have been dying to try some of the tests you guys have suggested!

First thing I did, was disconnect the NFB and wouldn't you know, the amp got much louder, fuller, and cleaner! It sounded great, actually. What does that mean? I connected it back up and there was the anemic, fizzy sound as before.

I measured the dc voltage at the "cross" and it was 80.1vdc, whereas the common cathode was 81.8, as you suggested. The voltages stayed the same even when I reconnected the NFB. I also went through and checked all the grounds. Everything read 0.1R so that's acceptable, right? I didn't notice that the middle lug of my volume pot never got below 1.8R when turned to zero. So that means faulty pot?

[Willabe]

First thing I did, was disconnect the NFB and wouldn't you know, the amp got much louder, fuller, and cleaner! It sounded great, actually. What does that mean?

It means the feed back that was applied was positive (PFB) and not negative (NFB).

Happens quite often and you never always know if the color code on the OT is correct as far as the primaries start/finish of it's wind and the color code of it's secondaries start/finish wind.

If either 1 is backwards, start to finish of it's wind (flip/flopped) then instead of applying negative FB you apply positive FB and it cancels.

So, many guys like to leave the primary fly leads a little long just in case their backwards so they can fire up the build and have enough fly lead length left to flip/flop them at the power tube sockets.



                     Brad     

   

[guitylerham]

First thing I did, was disconnect the NFB and wouldn't you know, the amp got much louder, fuller, and cleaner! It sounded great, actually. What does that mean?

It means the feed back that was applied was positive (PFB) and not negative (NFB).

Happens quite often and you never always know if the color code on the OT is correct as far as the primaries start/finish of it's wind and the color code of it's secondaries start/finish wind.

If either 1 is backwards, start to finish of it's wind (flip/flopped) then instead of applying negative FB you apply positive FB and it cancels.

 

So I should swap the OT secondaries at the speaker jack?

[Willabe]

So I should swap the OT secondaries at the speaker jack?

(I added a few lines to my post above.)

Yes either will work, but most guys flip/flop the primaries.


                Brad   

[sluckey]

So I should swap the OT secondaries at the speaker jack?

Yes, IF you only have one secondary tap. But if you have multiple secondary impedance taps available ( 4, 8, 16Ω, etc.) it would be better to swap the OT primary plate leads.

[guitylerham]

whoa, I swapped the two secondary wires on the jack and now the amp squeals insanely loudly!

[guitylerham]

Swapping the primaries too fixed that right up.

[guitylerham]

I still prefer the amp without NFB. It sounds fine with it, but it's louder and bolder without. I might put it on a switch.

[sluckey]

whoa, I swapped the two secondary wires on the jack and now the amp squeals insanely loudly!

Swapping the primaries too fixed that right up.

That all means the NFB phasing was correct to begin with.

[Willabe]

Swapping the primaries too fixed that right up.

Something was not right, even though it's working now.

You should only need to swap either the primary or the secondary leads (if you have a single secondary speaker tap, like Sluckey posted).


                Brad        

Edit; posted this as Sluckey posted.

[Willabe]

That all means the NFB phasing was correct to begin with.

So....... bad solder joint or tube socket, that was wiggled when he disconnected the NFB wire?



              Brad     

[sluckey]

So....... bad solder joint or tube socket, that was wiggled when he disconnected the NFB wire?

I don't think anything has changed. I think he just prefers the sound of no NFB.

There may still be a wiring error or wrong value component in the NFB circuit that causes the amp to sound bad with the NFB connected. And none of this addresses the issue of having sound even with the volume control set to zero. I don't think he's outta the woods yet.

[guitylerham]

I agree, something is still not 100% yet. Still need to figure out the volume issue. I will check the power supply with that large elec probe as mentioned before. My audio probe setup needs to be fixed before I can recheck the signal quality at the PI like I did before (that'll happen tomorrow). There is still some intermittent fizziness going on. Could be speaker but we'll see. The whole nfb thing is still a mystery.

[guitylerham]

I haven't had the time to work on my amp yet but I just wanted to thank everyone for taking the time to help me out. I'm gradually learning and it's with your input that I continue to make more sense of this stuff. I'll keep you posted when I get some new info to share in the next day. Thanks again!

[guitylerham]

Ok, I spent some time today and didn't get very far. I attached the negative end of a 10uf/500v elec to my ground buss, and probed various points with the positive end and a wire. With signal running through the amp, I noticed no changes when touching the positive sides of any of the filter caps. And when probing the plate resistors of the preamp tubes, nothing seemed to happen. When I probed the tube side of V1A and then V1B (typo in the schematic calls this one also V1A) plate resistor, the signal muted as expected but it didn't do that when I did the same thing to either V3A or V3B plate resistors. Is that abnormal?

When audio probing, I noticed that at the first gain stage, I could dime the volume and the signal was clean but when I probed V1B's output, the same volume setting would yield overdriven signal and even more so at V3B output. Is this standard cascaded gain stage results?

I replaced my volume pot, which read over 1ohm at "0" with a new one that read 0.1ohms at "0". I still get signal to the speaker. When turning the volume down, the signal sounds full range until I get to about 1-1.5 and from there to 0 it gets thin and tinny but doesn't get much quieter if at all. So it's like 1 on the dial is as quiet as it gets but the quality changes below that point. A funny thing I noticed was that when using  a dummy load resistor in place of the speaker and when turning up the volume with signal fed to the amp, I could hear it emanating from the OT area! I didn't know that happend. I noticed that when audio probing, with volume set to 0, at V3B pin 7 I couldn't really hear any signal but at V3B pin 6, I could definitely hear the louder thin signal. Is this because this stage amplifies the input and even a quiet "0 volume setting" signal at the input would be amplified a bit and be audible at the output? Once again, the inputs to the PI stay clean until maybe 75% volume but at the output of the PI, It only stays clean until about 25% volume. I realize that when my input signal is, say itunes songs from my computer, it may not be the same amplitude as when my guitar is the input. When I play guitar, I really can't hear it at the speaker when volume is at 0. Forgive me but I'm afraid I'm not being too terribly scientific with my process and am including variables that could throw off the diagnosis. It's possible this amp is functioning as properly as it will given my building abilities and component selection and that it just doesn't sound good (not terribly loud, buzzy distortion beneath the clean guitar note). If that's true, I'd consider gutting the whole thing and starting new. 

[tubeswell]

Going back to what  terminalgs said, check the DC continuity of the ground return lead for the pre-amp filter cap, at the supply node where the volume control is. It might not actually be connected to the ground return

[guitylerham]

Going back to what  terminalgs said, check the DC continuity of the ground return lead for the pre-amp filter cap, at the supply node where the volume control is. It might not actually be connected to the ground return

I just realized I haven't posted the entire schematic in this thread yet, so I apologize!



I went back and clipped my negative lead to the ground buss and the positive lead probed the negative lead of that last 22uf filter cap to make sure it is continuous and I read 0.1ohms, same as the other filter caps as well. And from pin 7 of V1B to ground is 0.1ohms when volume is at "0".

Also noticed another anomoly: when the reverb mix knob is turned up to about 90% all signal cuts out. This is also interconnected with the reverb tone knob setting. The higher the tone knob is, the lower the setting on the reverb mix has to be before all signal cuts out. Very strange. The cutout is very abrupt too. So I can adjust reverb tone and it varies the point on the reverb mix knob where it cuts out. Something is amiss.

[terminalgs]

If you as you turn the volume down,, you eventually get to no instrument sound at something around  vol='1-2',, but then instrument sound comes back as you continue to '0',  what you've got is phase cancellation between preamp stages as the signal sneaks across plate resistors (filter cap is not doing its job).  the amplitude of that sneaking signal is attenuated by plate resistors forming voltage dividers, so when it finally comes back to the "real" signal path via some plate, that sneak signal is (1) out of phase and (2) small enough that at vol='1-2', you get complete phase cancellation,, then as you rotate to zero, the "real" signal isn't there, and you've got only the stray sneaking signal. 


I don't know if thats really what you've got, but it could be. 


the reverb at '9' muting all sound is odd.  Does the reverb at "9" issue persist if your reverb pedal os set to "off"  (i.e. the recovery grid is grounded)


you've got a lot of triodes supplied by B+ node "D",  but Leo did it that way too, so it should work.

[guitylerham]

I get basically gradual volume attenuation until about 2 on the dial at which point, it stops decreasing, becomes thin and stays at that audible whisper. I set up my guitar looper and played loop into the amp in order to more closely simulate real world signal.

I hard wired the reverb on, so no footswitch. After moving some wires around with a stick, I noticed that my weird reverb cutout was due to either a cold solder or something else everytime I'd tweak the reverb pan output wire. I replaced that wire from the reverb pan out to the recovery tube pin 2 with a shielded one just to be safe and now the reverb works fine (albeit when full reverb, there are some high-pitched harmonic overtones arppegiating in the reverb wash... that's an odd bunch of words together). So at least that issue is gone. And you know, I checked all grounds and they all read practically 0 ohms. The amp sounds better to my ears the last I played it. The guitar signal doesn't really overdrive majorly until the last quarter of the volume's rotation. Fair enough. It has much more umph to it now, though I can't quite recall what I actually fixed in order to get here. It's possible that my earlier complaint about the PI outputs saturating could be that the output from the PI was overdriving my headphone listening amplifier input.

I guess I'm not sure how I feel about the amp still. I'd want to gig with it a few more times before making the call. I still have that strange volume issue. Can't figure that out. Should I be able to short pin 7 of V1B to ground and kill the signal? Doing so doesn't alleviate the problem.

Also, I figure it's worth mentioning but I'm running a 5y3GT recto in order to get the B+ down to a manageable level. No problem with that, right? I understand it could effect the feel of the amp though.

[guitylerham]

Further observation,


With volume at "0" and I jumper from ground to anywhere along the signal path up to and including the green dot, I can silence the pesky signal. At any point after that green dot, there is no effect shorting the signal path to ground.

[tubeswell]

You've got a lot of stages running from the 'D' node filter cap - you might not have enough filtering there. You could either try soldering another cap in parallel with the D node cap, or putting an additional coupling cap in the signal path between the 1st (V1a) stage and the tone-stack.

[guitylerham]

I clipped in a 33uf cap in parallel with that last filter cap and didn't notice any change. Also tried inserting a 0.1uf cap in series with the V1a pin 1 and that 100k/250pf intersection and didn't get any signal so I must not have implemented your suggestion correctly. From all my research, I'm lead to believe maybe this problem just isn't so much to get worked up over. Perhaps I should just let it be for a while and see if it really bothers me.

[Toxophilite]

This might be a bit pedantic/obvious
and I don't know if you've done this so please forgive me if you have
But have you slowly gone through your whole amp using the schematic , starting at the power amp and heading towards the preamp( or visa versa) to ensure that everything is wired properly, is the right value, and all the solder joints are good?
If not do it about 3 times...It's what S. Luckey got me to do when trouble shooting a few builds.
Even simple easy things like swapping preamp tubes with different ones to eliminate them as possibilities.
You can use a chopstick or something similar to push solder joints /components to see if the joints are bad


The reverb squealing at the top end of the range sounds like a feedback problem (I could be wrong though) and can be fixed by changing the proximity of the tank to the speaker and or putting it in a fabric/canvass/naugahyde/vinyl bag


I like negative feedback switches, I don't have them in my current deluxe builds (both made in donor hammond chassis)
Try a pot instead of a switch for fun, you can dial in exactly how much you want


 I am not really qualified to offer advice (having only built a few amps) I've found that a very methodical approach to my trouble-shooting produced the best results.and I've had to trouble shoot each one!!
Best of Luck!

[tubeswell]

I clipped in a 33uf cap in parallel with that last filter cap and didn't notice any change. Also tried inserting a 0.1uf cap in series with the V1a pin 1 and that 100k/250pf intersection and didn't get any signal so I must not have implemented your suggestion correctly. ..

Inserting a coupling cap between the plate of V1a and the tone stack shouldn't cause loss of signal, so yes, you must have done something wrong, like maybe shorted the a/c side of the coupling cap to ground (maybe contacting the chassis through the back of the eyelet board? - can't really tell what went wrong without a picture).

[terminalgs]

+1 for Toxophilite's suggestion to trace the entire circuit comparing to schematic.


I would not evaluate the amp until it's fixed.  It's working, but not working correctly.  The volume problem, even though you can still play the amp, is a symptom of a serious malfunction.  The NFB circuit needs work too, before you judge it (your 820R is the wrong value for the speaker tap you are using...).


If you pull both reverb tubes, does volume@zero problem persist?

[guitylerham]

I agree on running through the whole amp again and comparing to the schematic. I'll do that later tonight.

As far as the NFB resistor, both the Hoffman and Fender's original schem call for that 820R resistor and the single 8-ohm speaker setup. Do you guys still think I should change it? I did put the NFB lift on an auxiliary slide switch that I have on the front panel so I can easily adjust and defeat it!

I pulled both reverb tubes and all signal went away completely. Only when I dime the volume and Dwell do I hear a whisper of trebley signal. So either that is normal behavior, especially since you can't hear anything at moderate volume settings, or it means signal is still bleeding somewhere. And maybe it's due to capacitance between leads or components.

I did some more research and it was brought to my attention that the 100R resistor is also part of the NFB loop! This should be a 47R for a DR. I will adjust that and see how it fares too. I'm thinking of installing a presence control on this amp in place of the Reverb Tone knob that I currently have, seems to make little audible difference to me to justify the real estate!

[Toxophilite]

I take the 'reverb tone" control is an addition to the Hoffman schematic posted above?
Just curious as I couldn't find it on the schematic

[guitylerham]

I take the 'reverb tone" control is an addition to the Hoffman schematic posted above?
Just curious as I couldn't find it on the schematic

Yeah it is an addition. Sorry to spring a surprise like that! I'll post back after I remove it and back to stock to see if that clears something up. Boy, wouldn't that feel silly.

[guitylerham]

Clipped out the Reverb Tone with no issues.

I shielded a wire going to the Treble pot because it was making noise when I bumped it. Seemed to minimize the zero volume signal a bit, possibly just attenuated the highs is all. Now my reverb is howling. I just feel like gutting the whole thing at times. My build quality is just a mess on this particular amp (first BF style build). I still need to trace the circuit and I'll prolly find time tomorrow for that. Stay tuned!

[Toxophilite]

I was just curious about the reverb tone, it's a cool idea. one of the neatest features of the standalone unit.
I just asked about the schematic because some of the more knowledgable folks on the forum (which doesn't include me!!) like to see a complete schematic of what you actually have, to help them, help you.


From my own experiences of unhooking things left right and center when trying to trouble shoot an amp....I can certainly advocate for a methodical check through the build, section by section using the schematic and layout. If I'd always done that right off the bat instead of theorizing, I could've saved myself some grief and messing up what often was decent work to begin with.   I would do it at least twice..really!

My father, a mechanical engineer and all around smart fix-it guy says two things
"check all the easy stuff first" (swapping out V1 in the stage where you might be having a problem?)
and "speculation is useless"
Too often he has been proved right

It's quite startling what I sometimes discovered despite feeling assured that I was very thorough in my build.

That's no comment on your build as I've not seen it nor your abilities as I really have no idea, Just my own experience.

[guitylerham]

Thanks for sharing your experiences, Toxophilite. I can use all the insight I can get.

I traced the schematic and my amp and didn't see anything out of the ordinary. I do have a question regarding the use of shielded wire. In the case of a pot control that is located far away from its components on the board, is it ok to run a 2 or 3 conductor shielded wire from the board to the pot, connect to the lugs, and say one of the lugs is grounded, to connect that lug to the shield and tie that shield to ground back at the grounding buss located near the board/components? I'm getting some annoying oscillation when either cranking the volume sometimes or the reverb. Two seperate issues but related. So far have about 3 or 4 shielded runs of wire from the jack/V1A, preamp to volume pot, board to treble pot. I know shielded wire is more of a bandaid.

I did fix the volume issue by installing another power supply filter cap/resistor and fed both triods of the first preamp tube so it get its own supply now. No more signal at zero volume. I guess that means it was leaking through the power supply side.

The more I tinker on this amp the less I'm sure it sounds good or not. I'm half tempted to rip everything out and start again. Can't tell what's vibration, distortion, etc. Sitting in back of the amp, it is painfully loud so I think I'm getting full power, it just doesn't punch much until I'm at 60% or more on the volume dial (could be the taper of my pot).

And while I'm thinking on it, with volume cranked and decent reverb dialed in, when I smack the cab, it's more like ghostly moans and low freq wash than the normal "whip cracks" and snaps that I'm used to hearing when smacking an amp with reverb. Just an observation.

Anyway, I truly appreciate everything that you all have suggested and the time you've taken to help me out. I've learned a lot with this build. Not everyone can be a glowing success though!