New amp build (a cautionary tale for new builders)

[Kap'n]

Well, I'm now in the troubleshooting stage of my Weber 6M40 (Super Reverb circuit in a bluesbreaker-style box) kit, and I thought I'd share some of the mistakes I've made, just to show that you shouldn't undertake a difficult amp, like a BF Fender as your first build.

While I'm not new to this, the chassis on this amp is very cramped, and it's easy to not see missed connections. This amp will also be very difficult to replace the filter caps in 2017. I'm already dreading it.

Pix will follow when I actually have the amp running.

So far, prior to putting in the tubes in, I found that I wasn't getting any bias voltage. This could have burned up a pair of output tubes, had I actually put them in. Turns out, I smoked the bias diode, because I was too lazy to heatsink it. Replacing this was a pain, as you'll see in the pix.

I put the tubes in last night, and tried to bias them. No cathode current, and the bias wasn't adjustable. I figured out that I missed the resistor between the bias pot and ground, so the voltage divider didn't work.

While I was trying to bias the tubes, I also noticed that the reverb driver tube was getting brighter than the others. Immediately, I realized that the tube was going to burn up, so I shut it down. Turns out, I missed the jumper between the grids on the tube, and as a result one of the grids wasn't referenced to ground.

I re-fired up the amp, and discovered the tubes still wouldn't bias. What was the matter? I missed connecting the screens to the power supply. I'll be fixing that tonight.

What's the point of this? I mean, other than I did a sloppy build?



It's that I made, diagnosed, and fixed all of these mistakes without even getting a single sound out of the amplifier. I can't even imagine trying to do an amp this complex without experience.

Building an amp isn't like fixing an amp. When you fix an amp, you at least know it worked at one point.

[rudutch]

hey, I has the volume pot shorted to the chassis...
tough to see in a tiny 5E3 enclosure.

I never claimed to be perfect

you are still a 'good guy' from my perspective

[Kap'n]

One end of the volume pot should be shorted to the chassis.

[rudutch]

One end of the volume pot should be shorted to the chassis.

I think the term was 'unintentionally grounded" as in physical contact
from the pot rotating whenst is was being installed

[Kap'n]

Ah...

[Kap'n]

Let's just see if I can get this working....

[JAM]

Kap'n,

I had a look at the 5A40 layout - ouch! It was tough enough for me with the relatively easy 5F6A circuit in the tweed chassis. (I'm sure you'll have it working)

[jeru]

Hell Kapn, it'll all be good. You folks don't know me all that well, but if there's worrying to be done, it's a good bet that I'm doing it.

I am NOT worried about whether you'll sort this one out. I have every bit of confidence.


(BTW, got the message. For sure. Someday, my first build, -- CHAMP)

[chuckocaster]

but we love you anyways, ya ole

[Kap'n]

OK, now I'm stuck.

I wired up the screens, and I can bias up the tubes first.

Amp passes the pop test all the way back to the grid of the second gain stage on the normal channel. Amp goes completely dead everywhere on the far side of the tone stack.

Pop test seems a little wimpy on the reverb channel. If I shake the tank I get nothing. However, if I plug my guitar in, I can get the faintest sound out.

I checked the tone stacks on both channels back to the input. Everything checks out. Same with the resistance side of the tremolo. The bug seems a little low in resistance, but I can cut that path to ground via switch, and it's still dead.

Any thoughts?

[Victorilux]

Kap'n after the secind gain stage there are the 100k resistors that all get sent back to the cap labeled "D" in the shcematic, co-incedentally the reverb and vibrato also lead back to this cap. I'm thinking that there might be something going on there? They all tie in right where your pop test stops; after the grid of the second gain stage.

You may not have made the connection of the two sets of 100k resistors, the pair after the second to the pair after the first gain stages??? I don't acutally see any connection on the Chasis layout, but it's there in the schematic...

I dunno...just looks like the area where the trouble could be, I havn't even built an amp yet, and I'm just happy to be looking at a part of a schematic for a reason.

Am I seeing this right? TR1 is your power transformer, TR2 is the choke, TR3 is the output transformer, and TR4 is the reverb transformer? If so...damn thats alot of Iron!

[marT]

I know what you are going through. I remember my first JTM45 build I went with metro and I had no idea of building amps (I built a few pedals prior so my soldering was good). So many things in the instructions threw me, like not pointing out which screws to use to attach transformers and stupid things like that. Also I thought the preamp sockets were missing because they were attached to the shield so they looked different

I basically had not idea how an amp worked so I was relying totally on the instructions, I didn't even know the different between preamp, power and rectifier tubes.

However, it worked perfectly first fire up!! After asking lots of questions on the metro forum mind you.

The second build (My bluesbreaker) went together about 1000x easier and better. It only took me about 4 days to build and I did a much better job.

I'm addicted now, next will be a fender amp.

[Kap'n]

Kap'n after the secind gain stage there are the 100k resistors that all get sent back to the cap labeled "D" in the shcematic, co-incedentally the reverb and vibrato also lead back to this cap. I'm thinking that there might be something going on there? They all tie in right where your pop test stops; after the grid of the second gain stage.

Yep, however there's voltage there. I also get smaller pops on the resistor side in that section, but I didn't think I'd mention them because the capacitor side is completely dead. I think something must be grounding the signal somewhere.

I also think there's something going on around the phase inverter, because the pops should be louder than they are on both channel.

Well, I was hoping to get this done before I went on vacation, but maybe a break is really what I need.

[Toneseeker]

Amp passes the pop test all the way back to the grid of the second gain stage

Scuse me gents and pardon my ignorance but "What's a pop test?"
I'm just about to start my conversion of HRDX to 5F6A. Is this something I should know about? or is that just American speak for continuity test?

Pete

[Kap'n]

A "Pop Test" is sort of like a low budget Pop Idol.....

When you test voltages in an audio circuit, it makes a small 'pop'
sound when you probe. In a properly functioning amp, the pop should get louder the closer you get to the input (more amplification stages - louder 'pop.').

If the pops disappear, it means that no signal is making it to the output, so there's a discontinuity somewhere - either an open or a ground.

[Toneseeker]

Thanks for that Kapn. Good luck with the troubleshooting.

Pete

[wingnut1]

I agree Kapn that amp would be way to much for a first time builder, especially if they have have little to no troubleshooting experience. I built the 6A20 in the blackface chassis as my third build and I still spent a lot of time troubleshooting it.

I always find it helpful to post hi resolution pictures so that I can get some extra sets of eyes on the problem. Why don't you post some before you go on vacation and you might have some suggestions by the time you get back.

[Kap'n]

IWhy don't you post some before you go on vacation and you might have some suggestions by the time you get back.

Thanks! I'll get some pix up tonight or tomorrow.

Cruising for a week. San Juan, Bonaire, Dominica, Aruba, Grenada, Guantanamo....

They said something about wakeboarding. That ought to be fun.

[JAM]

Kap'n,

With everything so tight in there, any chance something is shorting underneath the board to a mounting screw? Have you traced a signal from the inputs to see where it's getting lost? All voltages look OK?

A break can really help. I know you can look at the same mistake over and over, then get away from it and it jumps right out. I had a miswire in my 5F6A clone. Looked at it for days and then it was right in front of my face.

[Kap'n]

Voltages are OK.

There's only two board mounting screws. There's supposed to be three, but one won't really fit. One's on the far left, by the normal input. The screw is threaded onto a piece of electrical tape so the head won't risk shorting to the nearby eyelet.

The second one is mid-board. All of the wires are cloth-covered, including the ones that run under the eylet board, except a couple that are really eyelet to neighboring eyelet jumpers.

Unfortunately, I don't have a signal generator or tracer. I have a scary-old (50's?) o-scope that somebody gave me about fifteen years ago, but I've never plugged it in, because it's two-wire. Talk about scary.

[NeoFauve]

Cruising for a week. San Juan, Bonaire, Dominica, Aruba, Grenada, Guantanamo....

Life is a Beach Boys' song.

[Kap'n]

Life is a Beach Boys' song.

Sure, loosely defined.

[wingnut1]

When I first read your reply I thought you said Guantanamo and water boarding. Then I reread and saw you said wake boarding.

[Kap'n]

OK, spent a bit more time on the amp, fixed a couple of little bugs, traced a bunch of wires to make sure they're going where they should. Fixed a couple of errors on the layout.

Going by this schematic....

Pop testing from the output backwards has the amp just sort of dying right at the junction of the vibrato channel and the reverb circuit. I thought the 0.02uF capacitor was toast, but I swapped it, and it didn't fix anything.

I can get a small amount of sound out of the amp in the Vibrato channel. If I hit it hard, it gets all 'mosquito-y.' I suspect it's due to ultrasonic oscillation.

The amp also audibly oscillates when you crank it up. No, it's not positive feedback in the output stage - I swapped the wires.

"Normal" channel is completely dead.

Any thoughts?

[NTBluesGuitar]

Both tubes are sharing the cathode bias components on V1, no? Is it possible that the resistor or bypass cap is bad, some connections not so good there, or something in that area? It's the only area I can think of that affects both the inputs, and you've already successfully tested up to that area.

I once had a 1ohm bias measurement resistor go bad and it read something silly like 4.5M effectively shutting off the power tubes. I could crank the amp and get the faintest of sounds from it.

I'm only extrapolating that as a possible issue assuming a preamp cathode bias would be affected in the same way by a similar component failure.

[Kap'n]

I see where you're going, and yep, I've had the burned 1 ohm resistor before.

Tomorrow night, I spend more time on it. Tonight, I crank the new SED EL34's in my Bassman/Marshall conversion. They sound better than the GT KT66's I picked up in the GC bargain bin.

[Kap'n]

Progress!

I touched up a few solder joints, and now the normal channel works! Sounds pretty darn good, too.

Current signs are:

• Quiet, mosquito-y sounds out of the reverb channel. The harder I hit it, the more mosquito-y they are.
• Volume and tone controls work.
• If both the volume and treble are close to being dimed, it breaks into audible (loud) oscillation.
• No evidence of the reverb tank getting or sending signal.

That really narrows down the problem area to the vicinity of V4., I think. More work tomorrow.

[NTBluesGuitar]

Were any of the issues in the area I had guessed?

[Kap'n]

Probably. I couldn't measure any flakiness, though. I just retouched all of the solder joints on the board and sockets for V1 and V2, and the normal channel came to life, so I guess it was a bad solder joint.

This next problem is going to need a systematic approach.

[Kap'n]

Well, I found the big problem with the reverb channel.

While sorting my resistors, one needs to pay closer attention to the decimal points.

The cathode resistor on the last gain stage before the summing node was 820 kohms.

To my credit, at least, it was not only my error, but the guy who picked the kit, because there's no reason to have one of those resistors anywhere. But it was my own dumbass fault for putting it in there.

Replaced it with a resistor of the correct type, and it's working!

I still have a bit of work to do. It crackles a bit, and the tremolo thumps rather than operates properly, but those should be easy fixes.

Pix this weekend!

[Kap'n]

I also figured out that the 'pop' test isn't 100% reliable. Some of the pop travels in the B+ line, and it also must sort of hit the rest of the spots on that power supply node. You can still hear the pop, it's just nowhere near as loud as it should be. That was throwing me a bit.

[Kap'n]

And of course, I was completely remiss in not thanking NTBG for spotting the problem a mile away.

Good job!

[cdw2000]

820K. LOL, that would do it. That was probably working something like a cathode follower.

With my eyes I can hardly read the color bands on those metal film caps, so I typically use a meter to check values before I solder them in.

Congrat's Kap'n. I knew you would find the problem.

[Kap'n]

I think the tube was in cutoff - acting like an open circuit.

I've got pretty much everything going now. Except the tremolo. The oscillator works, but I'm getting a big thump-thump-thump that shakes the speakers.

I took a look. I'm getting 35V across the optoisolator. Because there's 120VDC at the summing node.

I'm going to need to think about this.

[NTBluesGuitar]

And of course, I was completely remiss in not thanking NTBG for spotting the problem a mile away.

Good job!

Yay, me! What did I win?

Seriously, I'm just glad that you found the problem.

Sometimes a fresh perspective usually yields good results...I can't tell you how many times I've spent hours on some technical issue on a server only to have a fresh set of eyes notice a comma that was misplaced, or something like that.

JAM was the one who helped me isolate a cathode resistance issue in my 5E5A converstion build, and without him, I wouldn't have that experience to draw up on in the ol' cranial library.

Glad I was able to pay something forward.

[Kap'n]

JAM was the one who helped me isolate a cathode resistance issue in my 5E5A converstion build, and without him, I wouldn't have that experience to draw up on in the ol' cranial library.

Isn't that the way it always works? Thank you so much for your help.

The amp is now DONE!

The wierdo voltage at the summing node? A leaky 0.1uF coupling cap from the reverb channel. Bizarre. I've never cooked a cap before. As I suspected, completely linked to that horrible thumping tremolo. Replaced the cap, and the amp is 100% functional.

After that, I swapped the wire on the footswitch (it doesn't come shielded), and everything is as it should be. Loud, proud, and strong.

Pix (excuse the crappy cell phone quality)





It's a little tight in there, isn't it?

Now I'm ready to fake people out with my new "Marshall-clone!"

[NTBluesGuitar]

Holy Christmas, Batman!!!!

Damn! That looks like a mess...but yet, SO COOL!



What Marshall circuit gives you the faceplate/chassis with the right control layout?

[JAM]

Wow, looks great Kap'n.

Glad you got the issues worked out.

[ziess]

Wow! So a bit of a squeeze then!

Tommy.

[NTBluesGuitar]

Glad you got the issues worked out.

Anyone who attempts what Kap'n just did MUST have issues!