normal channel fine, vibrato channel not (ab763 Hoffman TR build)

[ElliottKoch]

The normal channel has the expected output although I did need to use the trim pot to adjust the bias to the tubes to boost the output, reduce the breakup and remove the hum.

The vibrato channel has an extremely reduced signal. I am checking the inputs and rewiring the input connection to ensure this is not the issue. There also seems to be a link between the hum and reverb circuit because increasing the reverb generates more hum.

Does one need to enable the reverb/vibrato via the foot switch?

Tube biases for the pre-amp tubes (V1 - V3) look good though not entirely sure what to expect from the Hoffman vibrato build.
The B+ voltage is a bit lower than the TR schematic (around 420V) and the voltages at the four filter cap inputs are low as well although the relative differences are about the same

(attached TR schematic)

D          380V   (410V in schematic)
C          410     (450
B          418     (458
Standby 420     (460

[CraigB]

Hi Elliot - first the hum from increasing the reverb.  Yes, that is normal, although a couple things may reduce that.  Just be sure that your lead dress at the driver 12AT7 and recovery stage 12AX7 is nice and tidy, and use tube shields on those.  (Just as a for instance, with the amp on, carefully chopstick the plate lead for the reverb transformer and you may notice varying degrees of hum.)

You have some sound to the vibrato channel, that's a start.  If your voltages all seem to be in line relatively to the schematic, I'd be suspecting either a wiring error or cold solder joint, possibly a bad ground connection.  I'd focus on that portion of the schematic and go through it from input to phase inverter with a highlighter and compare it with the amp just to make sure everything's wired up right.  Check it, and then double-check it. 

Posting a more complete set of voltages (including those at the preamp, reverb, oscillator, and phase inverter) will be helpful if you're not able to track down the problem as a wiring error or bad connection.

This doesn't have anything really to do with the channel issue, but are you measuring the power tube current?  Just curious where you have it biased.

[ElliottKoch]

voltage readings on power tubes
Pin 4/5 are at about 420V
Pin 6 on 1st two tubes -52V

on preamp tubes
V1:
pin1 215V
pin3 2V (give or take)
pin6 215V
pin7 2V

V2:
pin1 215V
pin3 1V
pin6 215V
pin7 1V

V3: (these would all pop when I measured them so the readings aren't so trustworthy)
pin1 215V
pin3 <1V
pin6/7 didn't take due to pop.

I've checked the grounds, and rewired much of the vibrato channel with thicker gauge wire, which helped, but the signal is still much lower than the normal channel. Is there a chance that the vibrato circuit is bleeding off the power? As I don't intend to use this, is there a way that I can remove it and pull the associated tubes (V5 and V6)?

[CraigB]

Hi Elliot, I'm not clear on your voltage readings so I'm going to ask for some clarification.  Obviously it takes a steady hand to probe around and take these readings, so I know the feeling when you touch the meter probe to a pin and you get a nice pop out of the speaker.  It's just a little disconcerting  But this is normal, and as long as your meter is appropriate for the job and you have it set right (or it auto ranges), this is not unusual.

We won't focus right now on your power tube voltage readings, since it sounds like there's no problem there.  We will need to go back and have you measure the current you're pulling.  You really need to read (and re-read) and have a good understanding of Doug's Fender amp service info, which can be found here:

http://www.el34world.com/schematics.htm#Servicing fender amps

For now, let's go back to the vibrato channel and look at your readings.  Let's identify which tube is which.  

V1, aparently there's no problem, since that is the normal channel.  Although I note that you're measuring plate (pin 1) and cathode (pin 3), plate (pin 6), and then you go to measuring the grid (pin 7).  You want to get a reading for me at pin 8.  That is the cathode.

Same thing on V2, which is the vibrato channel preamp tube.  Incidentally, if you've got pin 7 wired as the cathode instead of grid, that would be a problem.  If not, you want to measure pins 1 & 6 (plates) and pins 3 & 8 (cathodes)

V3 is the 12AT7 reverb driver, which should have both triodes wired parallel.  So pins 1 & 6 (plate) should be wired together, 2 & 7 (grid) wired together, and 3 & 8 (cathode) wired together.  If you have it wired right you should be getting upwards of ~440v on pin 1&6, and 8-9v on 3&8 (according to the schematic).

Finally, we need to get the measurements of pin 1, pin 3, pin 6 and pin 8 on the remaining preamp tubes, V4 (reverb recovery stage), V5 (vibrato oscillator), and V6 (phase inverter).

Note that V4 and V6 should both have pins 3 & 8 wired together on the socket.

Also check that your filaments are lighting on all preamps tubes, which you should be able to do visually, but if that raises any suspicions, check the voltage at the socket(s)

That should give us some better info to go on.

Now, here's a mistake that I've made before and it can really be frustrating if you don't think of the possibility.  A wrong value resistor, say 100 ohm instead of 100k, just as an example.  So with the amp unplugged and voltage drained off the caps, visually inspect EACH resistor and check it against the layout and schematic.  Once you've done that, if you find nothing wrong with color coding, then confirm that they are all within the ballpark tolerance of the appropriate value with your ohms meter.

I"m not sure it's even possible for the vibrato oscillator to "bleed off power", but your statement there would suggest a big wiring error.  If you've checked (and rechecked, and rechecked) and are satisfied you don't have any wiring errors, then we move on to other possibilities.  You have to divide and conquer, as one tube amp dude I know says, in order to isolate the problem.  So take one step at a time.  That being said, you can remove V5 and the vibrato channel would still amplify if it were working correctly, just there would be no vibrato.

Another suggestion, step away from it for a day or two and come back to it fresh.  You don't want to be fatigued and frustrated when you're trying to figure these things out.  Don't ask me how I know  

[tubenit]

CraigB gave some EXCELLENT advice and perspective to follow!  Well done.

With respect, Tubenit

[ElliottKoch]

I'll get you those numbers a bit later, but I have a stupid question.

Can I drain the filter caps using the ammeter to measure the current flow from the inputs to the filter caps to ground? I know this will remove the charge, just wondering if it will cause any damage.

I didn't wire the 1 Ohm resistors to measure the current from ground on pin 8 of the power tubes, so I'll be using Ohm's law to get you current readings through the 1kOhm resistor to pin 4. Will this be acceptable?

[HotBluePlates]

Can I drain the filter caps using the ammeter to measure the current flow from the inputs to the filter caps to ground?

Do it by setting your meter to measure voltage from the B+ rail to ground.

The ammeter setting acts as a fairly good short circuit, and it's probably better for your caps (and your nerves) not to do it that way. The voltmeter setting has a fairly big impedance between the leads, which will take a few seconds/minutes to discharge the caps if your amp doesn't have bleeder resistors. Bonus is you automatically also verify the voltage has dropped to a safe value.

But if you copied a vintage Fender schematic, you probably already have bleeders (the pair of 220k resistors spanning a pair of filter caps). These will drain the caps for you, and you need only attach a voltmeter to verify their operation.

I didn't wire the 1 Ohm resistors to measure the current from ground on pin 8 of the power tubes, so I'll be using Ohm's law to get you current readings through the 1kOhm resistor to pin 4. Will this be acceptable?

This works, but only measures screen current, not plate current. Cathode current is plate current + screen current.

If you don't have the 1Ω resistors, you'll need to read up on the Transformer Shunt method of measuring plate current. I strongly suggest using meter leads that have insulated alligator clips when using this method, attaching/removing the probes with the power off, and being extremely careful not to allow the probes to short to other socket pins.

[ElliottKoch]

Below are the current voltage readings on all of the pins for each of the tubes. At one point I had voltages nearly the same as those listed on http://www.el34world.com/charts/valve/ValveData.php?e=view&f=13013 apart form pin 5 on the power tubes (usually between -50 and -60V). The current I measured passing through the 470 Ohm (miss spoke in previous post) resistor off of pin4 on the power tubes is about a milli-amp.

(not listing the heater or guitar inputs and all measurements in Volts)
V1:
P1 243
P3 1.8
P6 204
P7 0
P8 1.8

V2:
P1 238
P3 1.8
P6 240
P7 0
P8 1.8

V3:
P1 406
P2 0
P3 7.8
P6 407
P7 0
P8 7.8

V4:
P1 243
P2 (big pop and I'm too much of a wimp to push it)
P3 1.8
P6 242
P7 0
P8 1.8

V5:
P1 206
P2 0
P3 2
P6 408
P7 207
P8 214

V6:
P1 234
P2 pop and I'm a wimp
P3 84
P6 225
P7 55
P8 84

Power tubes all essentially the same.
P4 408 +/- 1
P5 -48 +/- 1

As I said at the top, I've adjusted the bias pot and these are the current readings. I have adjusted it so that the power tubes can see 420 V on pin 4, but pin 5 sees -60 V and the amp is very quiet.

[CraigB]

Sorry Elliot, got busy with work and other stuff.  Just checking in to see how it's going.  Those voltages all look fine to me.  We've narrowed the field down to a couple possibilities.  You can try the obvious and swap different 12AX7s that are known to be good ones into the V2 and V4 positions.  Try swapping one at a time and see if there's any improvement.  Maybe you've already gone through this exercise, I don't know.  (As long as you can hear that your reverb is working and you can dial it up to a really wet surf tone, then we know the driver is working, so you can leave V3 alone.)



Take your meter and set it to continuity and make sure you've got bulletproof solid ground connections, & touch up any iffy looking solder joints.  A dull looking cold joint can really mess with things, and it's hard to see.

Did you happen to go through and measure all of your resistors in the vibrato channel preamp section?  Just make sure all your values are right and/or you don't have any that measure way off tolerance.

Also check and make absolutely positive that you have your potentiometers wired up correctly with solid ground connections.

You said you got a big pop when probing V4 pin 2.  I can pull my DR chassis and see if that happens.  Don't have the schematic in front of me, so I don't know which half of that 12AX7 that grid belongs to.

Isolating a bad signal cap may be the last avenue to take, so get back to us on your progress and we'll go from there.

Maybe since we haven't heard back, you got her running nicely and you've been busy playing guitar through it?  Hope so!

[ElliottKoch]

Not to worry, I totally understand and got a bit swamped myself. Already swapped the tubes, no effect.

I was going to check the current through the tubes via the transformer shunt method, but as the Normal channel is playing beautifully (essentially no hum and very loud), I think the bias setting is well tuned.

Like your previous post, I think that the its a solder joint, the input, and/or the pots (most notably the volume pot on the Vibrato channel). I'll check the pots tonight and re-solder some of the joints to make sure its all good.

Spent about 3 hours pouring over the circuit, hoping to find an error with no such luck. Pulled the vibrato tube, no change.

Regarding playing the amp...
I've found a work around. As the Vibrato channel is attenuated, I've jumped Normal input 2 to Vibrato input 1, with an AMAZING result. I understand the two channels are supposed to be out of phase, but with the offset in the signal, its still loud, no hum and the reverb sings. If all else fails, I will be very happy with this result, though not ideal.

(CraigB) Your help (and experience) is deeply appreciated. You are a legend to which us mere mortals will sing praise to, especially once my TR fills the air with the sound that only Fender products can.

[CraigB]

Hey now, I haven't come up with any solution for you yet.  I'm running out of ideas to help.  Yes, I think I've just been bumped back down to mere mortal status    But I'm OK with that!

I'd still like to see you get this thing working right, so don't give up. 

I agree, as long as you've got no redplating and the amp sounds fine, probably no need to worry about the power tube bias.  It is a good thing to know, so you can come back to that.  Be real careful when you do the transformer shunt style of measuring as HotBluePlates mentions.  You might want to consider the minor surgery that installing bias tip jacks and the 1 ohm resistors entails.

Is there a specific reason you suspect that volume pot?  When you plug into vibrato channel and sweep that pot, does the volume increase and decrease, or is it all just one attenuated volume?

Do your input jacks look like the attached photo?

Speaking of photos, those can often help the eagle eyes around here spot a problem area.  I'd be interested to see the area of jacks, pots, board and tube sockets associated with the vibrato channel.  Can you post some?

[ElliottKoch]

though not as pretty, yes the input jacks are electrically the same.

photos are below.
In the first one V2 is upper right and much of the circuit is in view. In the second V5 is in the upper right (blurry, sorry about that). The big black wire going over the top is the input from the jack. The other black wires are ground. The red wires fanning out across the top go to the tubes and the white wire to filter caps (point B in fender schematic - I think, but already checked that).

Wasn't the pots, they're all the correct values and made sure all solder joints are kosher.

For gits and shiggles I tried turning up the Normal channel to see how loud it would get. Although the volume at around 2 is what I expected from a TR, taking it up to 5 didn't quite get as loud as expected. So I suspect that both channels have the issue, only more prevalent in the vibrato channel.

If the tubes are under biased, would the vibrato channel be more affected than the normal channel? Also, whilst checking for red plating, I noticed that not all the heaters in th 6L6's were glowing with the same intensity (oh yeah, no red plating that I could see).

Lastly, when I got the tubes, I notice a small glass bead loose in one of the 6L6's. Since the all appear to be working, I thought nothing of it, however, I'm now a bit concerned.

The amp sounds fantastic, just more like a DR, not a TR regarding the output and diminished output on vibrato. I've not yet tried swapping out V4. Everyone is asleep at the moment, so it'll have to wait until tomorrow.

Regarding biasing, if pin 3/4 (as well as B+) have voltages that are the same, but about 10V lower than "expected" (i.e. I measure around 410V and one of the Hoffman pages suggest that I should measure about 420V), should I be concerned?

[CraigB]

Hi Elliott - I only see one picture attached - I'd like to be able to see more, like the wiring of the pots associated with the normal and vibrato channels and a larger overall view over on the preamp side.

Not sure I'm understanding why you have your preamp tube sockets mounted with pin 1 towards the back of the chassis, at least that's what it looks like in the photo.  Your board appears to be a Hoffman style, so just makes sense to orient the sockets the way his layout drawing shows.  The orientation of the sockets as you have them here causes wires that shouldn't need to cross over each other do just that.  That would not be an explanation necessarily for the problem we're trying to figure out.

What brand are you running for 6L6s?  NOS?  Current production?  A little flake of glass rolling around in there isn't likely to be a problem you should be worried about.

You should try swapping out V4 with another known good tube, it couldn't hurt to try.

Your power tube bias, if too cold, will give you a weaker output overall, but you've still got the vibrato channel that's super weak compared to normal channel, so the concentration needs to be at the preamp for now, I think.  If that can be ironed out, then we can move on to getting the power tube bias measured/dialed in. 

But the bottom line from the limited scope of what I can see:  some solder joints on top of the board look iffy, so the other big concern would be about any under-board connections also being not quite up to par.  That's a real bummer, and I speak from experience on that.  And then I see a plate resistor with solder splatter on it, which you really don't want rolling around in there if it falls off.  Plus, the lead dress really needs to be cleaned up - immensely helpful to me in that regard has been taking a real close look at other folk's builds here on the forum.  The lead dress and attention to detail is inspiring.

Give an update on your progress - it's a pretty complicated circuit, so a little harder to troubleshoot from here.

[ElliottKoch]

I was limited by file size and photo quality, but will provide a bit more later.

The orientation of the sockets was due to budge limitations (i.e. the school workshop built it for free), so indeed pin 1 is towards the back.

I was going to address the connections on the underside (as this is one of the few points of contention left). I will also clean up my soldering, which I've already done to a large extent.

I admit, that my inexperience in laying the wire is mildly embarrassing, especially when compared to the beauty and skill that others use. Though I've been bitty by the bug and in spite of the troubles I'm having, I will build another one and am very happy thus far (i.e. expected worse).

I agree that the problem lies somewhere in the preamp, and if not in the wiring of the "undercarriage" is likely V4 (easy test that will be done tonight). I'll keep you posted.

Again, I appreciate you help and insight.

[CraigB]

Sure, post more pics and progress as time allots!

As far as the sockets, if you plan to rewire, all you need do is remove the bolts holding the sockets, swivel into the intended position, mark where the new holes need to be drilled, drill, rebolt to chassis. 

Lead dress doesn't necessarily have to look real pretty in order for the amp to sound good.  It just makes it easier to work on, and in some cases quieter.  However, the thing you absolutely do want to look pretty (which translates into bulletproof) is your solder joints.  The last thing you want is to have to get at the bottom of the board after it's all wired up.

If this is your first amp build, you're doing MILES better than I did.  I think my first AB763 type amp would have been my third build, and I had all sorts of problems before I finally got it working right.  I'm glad you're encouraged to build more, it is quite an addicting hobby!!!

[ElliottKoch]

Sorry about how long its been, I've been inundated with the "day job". I have to say that my life is getting in the way of living ;-(


That said, I have found a couple of things...
In two separate places I, (like an idiot - or boss) had swapped two neighbouring capacitors.
I also found that one of the capacitors was bad, and very leaking (that's what I get for going with cheap parts, so I ordered some orange drops from Doug H.).

Won't post again until the new caps come, at which point, I'm hoping for the best.

Vibrato input is below - yeah, I know, its not as pretty as some of the others, but not convinced that this is the issue, especially in lieu of my capacitor issue. I was so focused on the resistors and other wiring issues, that the caps completely slipped my mind.

Elliott

[proaudioguy]

Regarding playing the amp...
I've found a work around. As the Vibrato channel is attenuated, I've jumped Normal input 2 to Vibrato input 1, with an AMAZING result. I understand the two channels are supposed to be out of phase, but with the offset in the signal, its still loud, no hum and the reverb sings. If all else fails, I will be very happy with this result, though not ideal.

(CraigB) Your help (and experience) is deeply appreciated. You are a legend to which us mere mortals will sing praise to, especially once my TR fills the air with the sound that only Fender products can.

I haven't read the rest of the thread yet. As an audio engineer with decades of trouble shooting behind me I would say your issue is between the point where the reverb taps off where there is a 3.3meg resister and where it returns to mix in with the dry. Perhaps the 3.3 meg resister itself is bad or open circuit. I say this because you are getting that nie lush reverb but little to no dry tone if I understand you.

[proaudioguy]

Just a suggestion but when I was about 13 I started soldering. When I was 18 I went to tech school and they had a mil spec soldering class. After 5 years I sucked at soldering. I ended up doing  well and got a B in the class. It was only 2 hours a week I think for just 8 weeks. Worth looking into but by no means necessary.  Practice under a critical eye helps.  I've spent most of my soldering career on xlrs trs and multipins. I've built at least 30 172 pin mass connectors and thousands of the others. My ab763 twin was my first build but my soldering was not an issue although I am highly critical of it. The only mistake I found in the circuit was the neg feedback wire was in the wrong side causing positive feedback. These guys helped me find it. I also have a bad mid pot or a bad mid ground connection I haven't had a chance to check out. When I turn the mid almost all the way down it disconnects intermittently and the gain goes through the roof the way I imagine lifting the tone stack would do.  I'll get around to pulling the chassis back out but for now I'm jus enjoying the amp.

[ElliottKoch]

(proaudioguy)

that's pretty much the nail on the head. I'll check those connections and the ones just up stream.

not going to pretend that my soldering is beautiful, but don't have the time for a class nor do I think that they're available (or if they are reasonably priced) out here in the land of OZ. the "day job" is as a research astrophysicist, so the fact that I can even solder is nothing short of a miracle ;-).

as I did find a bad cap, I'm still waiting on the replacement and will post the results once I have them.

I must say that its a bit of a tease with the Normal channel sounding so sweat, the reverb so wet, but quiet.

[proaudioguy]

Any updates?

[ElliottKoch]

I just can't catch a break!!!!

Replaced the caps - no change.

Added shunt resistors to power tubes V7-10 and now even when running the Normal channel, once the amp has been on for about 5 minutes, V7 develops a "pretty" blue glow which is quickly followed by red plating.

When the amp was first turned on, there was about 10mA (i.e. 10 mV drop across 1Ohm resistors on pin 8 of power tubes). There was some variation from tube to tube, but less than 10%.

I will most likely remove the shunt resistors (though I'm sure they aren't the issue, the red-plating came only after they were put in and I'm reaching an illogical state of mind).

The major change that was done was replacing the majority of the caps with the orange drops. This served two purposes. (1) I found a couple of bad caps and (2) forced me to make sure all cap values are correct, as I had found a couple that were wrong earlier.

So, I'd like to say that things are better. However, I've now developed a red-platting issue on V7 in addition to the Vibrato channel very quite. The reverb is doing its job, just the dry signal isn't making it out.

I tested the 3MOhm resistor as suggested by proaudioguy, but it is working properly.

Any help would be deeply appreciated. Even if its a lesson in what causes red platting in tubes. One more piece of information regarding this last bit. After I noticed the red-plating, I moved the tubes around and dropped the bias to ensure that I wasn't over biasing the tubes. This did not help. The red-platting is not caused by a failed(ing) tube nor over biasing.

[sluckey]

Added shunt resistors to power tubes V7-10

Please explain. Exactly how are the resistors connected?

[ElliottKoch]

just 1 Ohm resistor from pin 8 to chassis ground (ground peg attached to bolt, but all ground lead to the same place as measure with the multimeter).

[ElliottKoch]

Well, its been a while, so I feel I need to provide an update...

I have narrowed it down to the input and pots on the vibrato channel. Out of frustration (and not wanting to deal with the spaghetti that is my poor wiring) I jumped the Normal channel (after the 0.047uf cap) to the 3MOhm / 10pf connection to add reverb to that channel. I also removed the 220 kOhm resistor that leads to the output tubes from the Normal channel.

The results you may wonder...
A sweet sounding, loud fender twin.

Thanks to all who helped.

At some point I'll rewire the Vibrato channel, but for now, I will enjoy the sound I have. Besides, I was going to mod it along these lines at some point anyway.