New Princeton Reverb build, reverb not working.

[MadMax]

Hey guys, I had the chassis up and running but it took a few weeks to build a cabinet so I could install and hear the amp with reverb. The amp plays fine, and the tremolo is adequate (though I will be exploring ways to deepen that). I get no reverb when I switch that on.
I have a SPST switch right on the chassis with the option of using an original floor switch. The interesting part is that whether the switch is grounded OR NOT, when I turn up the reverb pot I hear a hum that increases in response. Is it possible that the hum is the reverb return stage increasing gain in the absence of the signal that should be coming from the tank?

The tank is a new MOD: 4AB3C1B. input = 0.7 ohms, output = 206 ohms. I swapped in an Accutronics with no change.
Next, I replaced the 12AT7 tube (V2) with a new 12AU7 and still no change.

My next idea is to swap out V3 for another 12AX7, which I'll do tomorrow. Is there a better strategy to diagnosing this? Thanks for any help you can give.

-MadMax-

[sluckey]

Insure the tank cables are connected properly. Since this is a new build there is a good chance that you have a wiring error or an incorrect component value installed. Verify that first.

Still no reverb? Divide and conquer. You can mentally divide the circuit right at the reverb tank. Driver circuit on the left. Recovery circuit on the right.

Turn the amp volume up. Now turn the reverb level to 10. Smack the tank just hard enough to make the springs rattle. Hearing a crashing sound thru your speaker means the recovery circuit is working.

Testing the driver is a little more involved but not difficult. Disconnect the tank cable that connects between the amp driver output jack and the tank input jack. Now connect a speaker to the driver output jack on the chassis. Since the driver is just a small SE amplifier you will be able to hear your guitar thru the test speaker. Your regular VTB controls will affect the sound. You may need to unplug your main speaker while doing this test. Better yet, just pull the 6V6s.

Let us know the results of these tests.

[eleventeen]

Isn't the spec'ed output impedance of the (proper type of) reverb tank supposed to be more in the ~~2500 ohm range?


You say you swapped in an Accu tank with no change....but you didn't mention the impedance specs, nor did you say that you were sure it worked wherever it came from. Just sayin'. You know, I take it, that you can google "accutronics reverb codes" and get the specs on same.

http://home.comcast.net/~machrone/bjr/reverbcodes.htm


"Is it possible that the hum is the reverb return stage increasing gain in the absence of the signal that should be coming from the tank?"


Absolutely.
 
Seems like there's a fair amount of reverb issues floating around these parts, no? There is a fairly recent thread about hum in a member-built (working, but with hum) Fender reverb circuit. It couldn't hurt to read through the posts on that thread with the idea that something there could jog your thinking and maybe point you in a fruitful direction. http://el34world.com/Forum/index.php?topic=17645.0

[sluckey]

Isn't the spec'ed output impedance of the (proper type of) reverb tank supposed to be more in the ~~2500 ohm range?

That's correct. Key word is "impedance". He posted DCR values and they are fine.

[MadMax]

Sluckey you got to that point before I could. I don't have a signal generator so I don't think I can check actual impedences.

Both the MOD and the Accutronics (same PN) are new and read very similar resistances. I'm at work now, but can't wait to get home and check these suggestions. Meanwhile I'll be reading that post that eleventeen referenced in case I can find a common thread. Thanks guys.

I'll be back with results.

[MadMax]

Alright, I got the reverb to work! Now I still have the hum to deal with.
You all couldn't have known this (because it should've been irrelevant) but I had installed 1/4" jacks for the foot switches. I realized later that the foot switches come with RCA connectors. Rather than drill new (smaller) holes, I opted to insert 1/4" to RCA adapters (still an open circuit without switches plugged in right?) It's also at this point that I am discovering that an OPEN circuit is required for both vibrato and reverb to work; grounding these wires turns the effect OFF.
Anyhow, on a whim, I pulled the adapter from the Reverb switch jack and it came alive! Upon pulling the chassis to figure out the mystery, I realized that with a 1/4" male connector plugged in, the prong for the tip is moved over far enough to touch the copper ground line I'm using for the jacks (thereby shorting and turning the reverb OFF.

As to the hum. It is increasingly loud as the  pot is turned up, whether the reverb switch is on or off. I tried a jumper from the 220k resistor to the cathode ground point on the board with no change. I did try another set of tank cables in case my beefy, home made set from an old mic cable had a problem... no change. Disconnecting the tank return cable makes the hum even louder and 'hairier'. The cathodes for V3 aren't shared on Doug's schematic, so that shouldn't be the issue. Just like in Psychonoodler's case, pulling V3 eliminates the hum.
My next move will be to replace the wires to and from the reverb pot with shielded wires (grounded shield at the pots I guess). I'll be back...

[MadMax]

Shielded wires did no good. I think I need to eliminate whatever is coming in to the grid (vs. some problem with the cathode or plate circuit). If I lift the wire from pin 2 (V3), any suggestions for what kind of signal I can substitute there to see if the tube can cleanly amplify it?

-MadMax-

[sluckey]

Rather than put a signal into the tube, put a ground on pin 2 to kill any signal (including possible hum). Does that kill the hum?

[MadMax]

First, I just jumped pin 2 (V3) to ground with the signal wire still connected (no change). Then I disconnected the signal wire from pin 2 and left the ground jumper in place (still no change). And for a brief moment I disconnected the ground from pin 2 (open grid circuit now), and it got really loud.

Lifting the incoming signal from the 100k reverb pot does eliminate the hum, so I think that fact eliminates V3-B as suspect. I swapped V3 out for a known good 12AX7 and nothing changed.
If I'm right, the problem is now isolated to V3-A plate or cathode, right? 

[sluckey]

Sounds logical to me.

[eleventeen]

Maybe try throwing a separate 20 uf / (at least) 300 volt type filter cap on that triode, close to the tube. Yes, I know the stock circuit does not have one (other than the one in the regular power supply filter) but you are more interested in the result than the parts count, I take it. Easy enough just to clip or tack solder one in. Remember to discharge it if you clip in-unclip with jumpers! You'd place it right above the 100K plate resistor. Node "D" on the (deluxe reverb) dwg I am looking at. Unquestionably a worthy experiment.

[MadMax]

Eleventeen,
I tried your suggestion and the obscene hum does go away with a capacitor in series before the 100k plate resistor. However, with a cap in place, the reverb doesn't work. I tried a 10 uF, a 15 uF, a 30 uF electrolytic, and finally a .22 uF non-electrolytic cap. The bad hum goes away each time, but the reverb won't work with the cap in place, which is odd since the D node is filtered anyhow?!

Any thoughts on how to use this information to close in on the culprit?

[eleventeen]

Are you utterly sure: The cap has to be installed (with the + terminal) ATOP the plate resistor, at the NON-TUBE end....otherwise you are pretty much just shorting out the tube in terms of its AC (eg; signal) performance.....regardless of the cap value. Just to over-clarify, the cap cannot be connected directly to the plate of the recovery tube.


Outside of confirming that your temporary cap isn't connected wrong, it begs the question, is your "D" node filtered, meaning, is the cap that should be filtering under/at node D working?


In another sense, this may or may not cure or have anything to do w/your hum if it's just shorting out the recovery tube.


 

[AZJimC]

The suggested cap should be grounded on neg, and pos to the power side of plate resistor, NOT in series with it. A series cap there would isolate the plate from high voltage DC, and thus stop it from working at all.

[eleventeen]

Yes, I missed that. You can't put such a cap "in series" with the plate resistor. The above post describes what should be there, accurately.

[MadMax]

The big cap I've been using is a multi value can cap which grounds at the body. I have a high voltage 16uF axial lead cap I'll try tomorrow. Just ground the negative lead and attach the positive lead to the power rail side of the plate resistor. I'll let you know how that goes. Thanks.

[MadMax]

Well, when connected as described above, the 16 uF cap does not change the hum (reverb works though). I did check the main filter cap with my capacitance meter. It's a 20-20-20-20 uF can cap. Installed, the readings are:
A: 39 uF
B: 38 uF
C: 25 uF
D: 26 uF
Seems a little odd to me that the first two are almost double the rating. I guess that with no resistor between node A and node B (I'm using a choke instead), it is acting like a parallel connection between the first two sections.
I have mentioned that the hum goes away with the reverb pot zeroed out, and it builds to intolerable as the pot is turned up, whether the reverb is switched on or off. However, to be perfectly clear, when the switch is on, it IS a bit worse (in fact the hum feeds back with the reverb switch on and begins to howl).
God bless y'all for sticking with me on this, I'm about tapped out of ideas.
Only four components are involved with V3-A plate and cathode circuits. All test fine (in place). I replaced the 100k grid resistor already since I had it out for these tests. That just leaves the 3.3 nF cap, and the cathode RC parts. Unless you all can suggest a better course of action, I'm about to replace all three just for giggles.

-MadMax-

[sluckey]

Your hum may be due to layout. Show us some hi rez pics.

[eleventeen]

Agree w/Sluckey, layout issues may be it.


Your power supply filter cap is dead on the money normal.



"That just leaves the 3.3 nF cap"


Eh?


You mean 3.3 M resistor in parallel w/the 10 pf cap....?

[sluckey]

"That just leaves the 3.3 nF cap"

Eh?

You mean 3.3 M resistor in parallel w/the 10 pf cap....?

He means the 3.3nF coupling cap that's connected to the reverb pot.

[eleventeen]

Are you saying that a .003 uf cap can be expressed as 3.3 nf? Never seen that used.

[sluckey]

Are you saying that a .003 uf cap can be expressed as 3.3 nf? Never seen that used.

Yes. It's a new generation thing. There were no nano nano before Mork and Mindy. 

1 microfarad = 1 farad*10^-6
1 nanofarad = 1 farad*10^-9
1 picofarad = 1 farad*10^-12

[MadMax]

Fellas, after seeing some of your work, I'm almost embarrassed for you to see mine. Just bear in mind that this is my first build from scratch. It'll probably be easy to condemn the dress of the entire thing, just remember that only the V3 seems to be involved.

Sorry for the crappy resolution, I can't seem to get better ones with my little Kodak.
The preamp tubes from LEFT to RIGHT are V1, V2, V3, and V4.
 

[eleventeen]

I'm not prepared to tell you "this is it" but I am also not sure you can get away with the reverb transformer inside the chassis, esp so close to your V1/2. The rev trans gets very high B+, only separated from output tube B+ (eg; highest in the amp) by the choke and much higher voltage (and thus unfiltered ripple) than the preamp tubes. And yeah, the rev trans is metal enclosed and thus shielded, so maybe I'm speculating out of thin air. But that's a VERY sensitive zone as far as hum inducing components. Seems to me that you've left those leads long enough to unscrew the rev trans and swing it out to the outside of the chassis as they are and hold it there with a little "C"clamp. Or if necessary, splice some extensions onto the rev trans wires and try same....be careful, there's BIG volts on the red & blue wires. Or...just go for mounting it on the outside and lead the wires through a hole w/a grommet.

[2deaf]

Don't do anything until Sluckey tells you to move the brown grid wire off of the green heater wires.

[MadMax]

I have grabbed  that brown wire with a small pair of needle-nose pliers and moved it away while the amp was humming and it didn't seem to change anything. I tried to keep the heater wires reasonably short to minimize their impact on things, but maybe I have to redo them long enough to push out of the way? The red wire for the reverb tranny does cross the heater wires at about 90 degrees, could it be picking up a 60 Hz hum from them?

That red wire, the mounting location of the rev. tranny (and the brown signal wire) would only be responsible for the hum getting worse with the reverb switched on. But I will address them when I get a few minutes.

[2deaf]

I noticed you had the heater wires flat against the chassis on 10/12/2014 and now they are up off the chassis.  Why?

[MadMax]

Good attention to detail 2deaf. Just tried a little different position hoping for some change in the hum, though none was forthcoming. It does sound like 60Hz, and as near as I can tell, only the heaters have AC over that way.

[2deaf]

Is it possible that the heater is getting to the cathode either through the tube internally or through the socket/crud/wiring?

[MadMax]

I've ruled out the tube by swapping it out but I'm not against redoing the wires. Maybe this weekend  I'll pull up a picture of something wired by Sluckey or EL34 and try to duplicate the dress (although I already feel like the socket terminals are getting tired of seeing my soldering iron!).

As you can see, the salvaged Frontman 65R chassis is angled in at the back making access tricky. I already took eleventeen's suggestion and moved the reverb tranny outside of the chassis, I just haven't had a chance to solder the wires back and try it out yet.

[eleventeen]

"moved the reverb tranny outside of the chassis,"


Eager to hear the results, if any. We're rooting for ya!


That's rooting, not rotting. 

[2deaf]

While the amp is off and cold, see what the resistance is from V3 pin 3 to ground.

[MadMax]

Alright, for those who've been following this hum hunt... First to 2deaf, the R from V3(3) to ground was 1470, and since there's a 1.5k ohm resistor in that path, it seemed good to me.

Anyhow, I rarely make more than one change at a time while on a hunt like this because I want to know exactly what fixed it. But in this case, the socket wiring was much easier to get to while the transformer wires were not in place. So I moved the reverb tranny outside of the chassis, and I did a repair to the heater wires. While looking carefully at them I realized that V3 and V2 both had inverted connections, in other words the wire from 4/5 of V4 connected to pin 9 on V3. I tried to keep them all the same, but you know, they're both green and they're twisted together.
After both changes, I fired the chassis up and... Still howled when the reverb is on and wide open... but the reverb was also NOT working. Trouble shooting with an ohmmeter, I discovered that when I unsoldered the reverb tranny red wired, the wire underneath the board had come unsoldered and dropped out. I twisted them together, and... The reverb is working, and although there is a noticeable hum when it's turned all the way up, there is no howl when it is switched on. I'm okay with it this way, not sure which move was the coup de grace, but the chassis is going back in tomorrow so I can enjoy the amp. Thank you to everyone that helped me through this, I hope I can help you one day.

-MadMax-

[sluckey]

Trouble shooting with an ohmmeter, I discovered that when I unsoldered the reverb tranny red wired, the wire underneath the board had come unsoldered and dropped out.

I don't like using underboard jumpers but sometimes it's unavoidable with a complicated board. When I do use them I want to know they will stay in place securely. So, I always push the end of the jumper completely through the turret and 'hook' the end back over the outside of the turret. This little detail allows you to see the jumper even after soldering, but more important, it prevents the jumper from falling out of the turret whenever you do any later soldering.

[MadMax]

Thanks Sluckey, I sure will do it that way next time!

[PRR]

> I don't like using underboard jumpers but...

I think it is begging for trouble. Run 'em on top before you stuff the parts.

A key rule in electronics soldering is "mechanically secure". The part/wire "should" stay in place before it is soldered (and re-re-re-soldered as happens in new-work). Yes, I often ignore that rule.... but I like to keep the joints/parts/wires out where I can see what I am knocking-off.

The J-leg at the top is "mechanically secure". I suspect in a tight chassis I could knock it off in re-re-work, but it aint easy.

Still hard to see the jumper after soldering.

But what are those extra holes? Aren't they for off-board jumpers? Can't you use the same idea (and some of the same holes) for really-secure on-board jumpers?

If you use insulated wire (reduce under-board shorts) in several colors, you can also see from the half-inch exposed if the leg-bone connects to the hip-bone, or the neck-bone (oops).

[PRR]

It's a new generation thing. There were no nano nano before Mork and Mindy. 
1 microfarad = 1 farad*10^-6
1 nanofarad = 1 farad*10^-9
1 picofarad = 1 farad*10^-12