5E3 Build - Intial Test Problem

[BluesDude]

The kit is a Weber 5e3 with a WY3GTcopper cap for the rectifier.

I got the amp together and was going through a pre-check that consists of the following steps.


1) no tubes - power up the amp and check the filament voltages - Voltages on the PT secondaries. filaments were 5v for Yellow (I might use a tube for the rectifier later). 6.3v for the Green and 680v across the red.

Ran into a problem with the pilot light... did not light. Found the threaded brass bulb holder was not connected to the soldering tab. I was able to make the connection with some solder... go that working.

2) insert the rectifier and check voltage at C12. Ran into a problem. I heard a crackle pop then a puff of smoke. I shut it down. The smoke happened quickly it was difficult to tell which cap even upon inspection... but I suspect C10. I check the connections and they all look right. So I powered it back up. Still got some crackling noise. I quickly checked the voltages B+ on C10, C11, C12 they all peg the meter (which was set to 1000v AC). I know the voltage should run high without the rest of the tubes... But that seemed extreme.


I inspected the caps. They do not physically look damaged.

I assume the high voltage was the culprit on the smoke, crackle, pop.

1) Why would the voltage be that high?

2)Do you think the caps are damaged? The power was only on for a minute or so.

3) Any thoughts? Where should I go from here?

[Jazz P Bass]

New build

You did use a light bulb limiter, right? A variac?
I would set the meter to read DC volts not AC.
With no tubes in the amp, there really should not be very much current draw.
I would go back & check the build, start to finish.

[loudthud]

Are you sure the polarity of the electrolytic caps is correct? Can you post a picture?

[Michael Allen]

In my experience, I've had similar issues that turn out to be no big deal.

1) The crackling and smoke. Upon first applying high voltage, there could be flux or solder junk that gets hot and burns up. May not be a cap at all. Or it could be a resistor that is on the B+ line that is shorting to ground and so all the current is rushing through the resistor burning it up. Check that all power resistors aren't shorting to ground.

2) The voltage ping. If you use a cheaper meter like I do, it make take a few seconds to stabilize or get a good connection with the probe. As long as nothing is burning up, leave the amp on for a second or two longer until your meter can stabilize.

[BluesDude]

Update

Jazz P Bass reminded me about the meter setting. I did measure at B+ with an AC setting. I switched to DC and got the correct reading.

Loudthud - I do have the caps installed correctly. I checked it several times.

I got it working. The cracking and popping settled down once I installed the other tubes (to take other measurements). All of the voltages were within limits so I gave it a try with the strat. It is working.

After playing for about 15 minutes at volume of 3 or 4.... I heard a noise that sounded like the cap crackling again... Then I saw another small wisp of smoke. It was in the vicinity of the power supply caps.

I shut everything down, made sure the caps were drained and took a mirror and looked at the top of C10, C11, and C12. C10 looks like it has a bit of a wet circle around the lead on the negative side. I assume it is some of the electrolyte.

Does that mean the capacitor is going bad?

Should I replace it?

Should I not play the amp until I replace it?

Thanks for everyone's help.

[Chuck H]

Is C10 the first cap after the rectifier? If electrolyte is leaking the cap should be considered bad and should be replaced before the amp is powered on again. The cap could even explode if it fails outright. The question is, why did it fail? And, will a new one fail or was this simply a bad cap? Buy two caps in case. That way if the next cap fails you'll know you have a problem to find but you will have a cap to install when it's fixed.

Chuck

[tubeswell]

'Crackling' noise can be the HT/B+ shorting to ground causing the PT to boil (especially if there is smoke). Check the continuity between your B+ and the ground with a multimeter. You should not have shorts. Do any of the parts look burnt?

Disconnect the rectifier, and measure the VAC from each side of the High Tension winding to ground. Then measure the VAC across the respective heater filaments. If it looks okay pop the rectifier back in, but leave the other tubes out. Now check that you have a highish VDC B+ at the first filter cap (It should be somewhat higher than it will be when all the tubes are in).

(Maybe when you switch the amp on next time, leave it lying face down on your bench with the back open and watch for where the smoke is coming from.)

What are the idle voltages like?

[BluesDude]

Tubeswell - Below is my initial test procedure. All voltages checked out with in a few volts of expected values (except for my the high voltage reading on C10 was a newbie mistake... didn't switch to DC on the meter. Which I rechecked).

I powered off and back up each step and followed the sage wisdom of this board on safety. TubeDepot has a guide on their site... it has a checkout procedure. I followed it. All voltages readings I took were within limits ( what I have listed below is based on my memory... I did not write them down. But they were very close to what the guide listed).

1) Power up with no tubes. checked voltages on the filament secondaries and the HV red secondary. They were correct 5v, 6.3v and 680v.
2) Added rectifier tube. Checked the B+ voltage at C12 (16 uf) and read about 460V
3) Added the preamp tubes. Checked C1 & C5 (25uf) and read about 1.5v. Checked R16 (1.5k) at the PI cathode and read about 63v.
4) Added the power tubes and plugged the speaker in. Checked C9 (25uf) and read about 22v. Then turned the volume 1 & 2 up separately. when it got around 10 I began hearing a very slight hiss (white noise).

Then I plugged in the guitar and played for about 15 minutes. That was when a second smaller discharge occurred. [I'd like to chalk up the smoke to the blazing lead guitar I was playin'... but I would just be fooling myself ]

I inspected C10 very closely with a flashlight and magnifying glass. The smoke looked like it came from that vicinity. I can see where some residue hit the brass grounding plate (just above C10) and corroded it (created a green spot about the size of a dime). I will replace that cap. C11 and C12 have no such discharge above them.


In step one listed above... I could hear a fair amount of sizzle on the cap (looks like C10) before the smoke. The weber caps are rated at 450v. Do you think the extra voltage in Step 2 (rectifier only check) that the increased voltage pushed a weak link (C10) over its limits? Once I installed the other tubes the the sizzle quieted down.

Any other thoughts?

[Chuck H]

It's worth noting that Weber has just changed all their kits by bumping up the first filter cap voltage rating to 500V "due to high wall voltages". So I expect this has happened before. I don't know what kind of filter caps they supply but the above action tells me they have seen a lot of failures. I have hit good filters with overvoltage many times without failure. If temporarily testing at 10V over the cap rating has been causing so many failures I have to suspect they are crappy caps. I would replace them all with good quality caps because the quality of the cap does effect the tone of the amp. 500V rating for the first filter (C10) and 450V for the rest of the filters. Sprague has a 500V 20uf "Atom". The Atom caps have a slipping reputation, but having found nothing better many builders are still using them. I have badly punished Atoms with overvoltage and never had a failure.

JM2C

Chuck

[tubeswell]

T
2) Added rectifier tube. Checked the B+ voltage at C12 (16 uf) and read about 460V

I inspected C10 very closely with a flashlight and magnifying glass. The smoke looked like it came from that vicinity. I can see where some residue hit the brass grounding plate (just above C10) and corroded it (created a green spot about the size of a dime). I will replace that cap.

I could hear a fair amount of sizzle on the cap (looks like C10) before the smoke. The weber caps are rated at 450v. Do you think the extra voltage in Step 2 (rectifier only check) that the increased voltage pushed a weak link (C10) over its limits?

yeah possibly. Put another filter cap in and see it its any better. As an aside, the filter caps were installed with the polarity the correct way around I presume?

(FWIW I too use Sprague Atoms as my filter caps of choice, and haven't had any problems with them yet.)

[BluesDude]

yeah possibly. Put another filter cap in and see it its any better. As an aside, the filter caps were installed with the polarity the correct way around I presume?

(FWIW I too use Sprague Atoms as my filter caps of choice, and haven't had any problems with them yet.)

I was very careful with the electrolytics... I saw a video of an explosion.


All of the electrolytics are installed (oriented) the same and the others (C11 &C12) are not having problems. I checked the voltage at C10 (with all tubes in), it is 350v. The only way the polarity could be reversed is if the manufacturer put the plastic jacket on backwards. Both sides of the cap can are physically the same... so I can not verify that way. Since I am getting the 350 v reading (110 volts lower than the reading with just the rectifier tube installed and not a constant sizzle) I am thinking it is correct... I was very careful with the wiring also. At this point I believe it is correctly installed and wired up.

I think my next move is to replace the cap and see if problem is resolved.

Do you know of another way to verify polarity (just in case there was a manufacturing problem) perhaps after I pull it out?

It's worth noting that Weber has just changed all their kits by bumping up the first filter cap voltage rating to 500V "due to high wall voltages".

Chuck

I checked with Weber. They said the cap could be bad... but as long as the voltage applied is within 450v... there should be no problem. No mention of changing the caps on their kits. As stated above the voltage measures at 350v.



Thanks for the comments and help everyone. The feedback has been very helpful.

[boojo35]

All caps should be run at least 20% under their rated voltage. That is just a rule of thumb in the electronics world. For instance, a cap that will be run at 400V should be rated for at least 480V.

[Chuck H]

Well... Since his thread is three and a half years old anyway, I'll bite... Where do I get 480V caps? Just kidding. But caps rated over 450V are pretty scarce. It might be good to mention alternative circuits that circumvent cap rating. Like a "totem pole" arrangement. With two 350V caps in series and a large value resistor across each (there's a formula for figuring the optimum value for the resistors but I don't know it so I just default to 220K) the voltage across each cap is pretty well balanced and the voltage capacity is 700V.

[boojo35]

I realized how old this thread was AFTER I dug it up. You have to remember that caps in series will do the opposite of resistors in series. The capacitance value divides rather than adds so you must use double the value in capacitance. 475V and 500V caps are available. Yes, the 480V caps might be a little rare. LOL.

[Chuck H]

I tried to be glib, but considerate. A pair of 350V/100uf caps in series with 220k resistors across the leads is good to 700V for any guitar amp circuit for a 50uf main filter capacitance. In this age of ever changing electronics, this sort of circuit is a definite "go to" for getting the job done IMHE. Since there are only a couple of MFG's for caps over 450V and they have spotty reputations!!! My latest build is using Nichicon 350V caps in series with balancing resistors. The unloaded B+ will be about 525V. So I don't want any 400V caps blowing up in a new build! and even though a lot of MFG's get away with using caps rated for the loaded B+ I just can't get past the unloaded voltage for my own personal design. Which is certainly more ideal. I just thought that the series caps were worth mentioning.