crazy crackle at start up

[mfreqmaster]

I have a 18 Watt build, built last year. I happened to place this particular 18W in a head cab. The deviations from the standard circuit are an EF86 preamp for channel 1, & an octal rectifier tube, currently housing a GZ34. The only issues I've had with this is that on start up, that is when taken off standby, the tube nearest the power transformer lights up pretty brightly for a moment & then goes away. The other issue is that unless I warm the tubes up by leaving it in standby for 2-3 minutes, there is a crackle sound that lasts 30 seconds or so. Once the amp is warmed up it can be played for hours without any problems. I've actually taken this to several gigs last year, I felt confident enough in it's reliability to do that, so I think overall it's in pretty good health. I'm mainly curious what that crackle sound is. I appreciate any response.

[Chuck H]

What is the "tube nearest the power transformer"? The rectifier or one of the power tubes?

The crackle sound could be related, May be a bad tube. May be a dirty contact, oxidized tube pin/s, cold solder joint or a failing component. I'm not comfortable or confident in circuit descriptions. As in "I have a 18 Watt build, built last year. I happened to place this particular 18W in a head cab. The deviations from the standard circuit are an EF86 preamp for channel 1, & an octal rectifier tube, currently housing a GZ34." It usually turns up that something else is different too. For example, you cannot have used a typical 18W build power transformer because there is no 5V winding. So that would be different too. How different? I dunno. Plate voltage could be important in this case and we can't know what that is or if there may be any other differences with your circuit.

Can you provide a schematic?

[Jazz P Bass]

Have you swapped the tubes around? (power tubes, right?)

The bright one, if it follows the swap, has a heater filament issue.

I have seen that on preamp tubes but not on power tubes.

[g1]

Yes, if it's lighting up from the standby switch, rather than the power switch, then that tube has a problem (unless it's just blue glow).

[mfreqmaster]

OK, here are some details:

Power transformer is Magnetic Components Inc 40-18035. This has a 5V winding. Here is a link, hopefully: http://www.classictone.net/40-18035.pdf

I forgot to add the amp also has a choke & a PPIMV. This was my first ever time playing through a PPIMV & I like it so much any future amp build of mine that has a MV will be a PPIMV. I really like the way it sounds compared to my Marshall 2204 type MV.

I have attached a schematic which hopefully works, but if it doesn't the major portion of the schematic would be the 6M18 available at the Ted Weber Site, modifications can be found at TubeDepot's kit section. I used the modifications found there. I also included some inside pics which reveal my untidyness & perpetual noobism, but the amp sounds great once it is passed the warm up period. You can see I have a star grounding scheme, or one close to it. The amp is reasonably quiet once warmed up. The EF86 is a great alternative to the original circuit. I now have 2 unique sounding channels in this amp, the EF86s bright crunchiness, or the mellower 12AX7. I also took RG's advice by fusing both sides of the 6.3V winding.

Hopefully these attachments will work.....







https://d1sjrnpi226dnf.cloudfront.ne...pdf?1382030428

[mfreqmaster]

Here is a reply to Jazz P Bass's question. I did swap the power tubes. Whichever EL 84 is in position V5 lights up, doesn't matter which one.

[loudthud]

I find it funny that all the threads listed in the section below which lists Similar Threads all start with the word Crazy.

This may sound crazy but have you checked for bad solder connections on the suspect tube socket or any signs of arching? Remove the tube and visually inspect the socket's contacts. Are they bent or discolored?

[Chuck H]

Since the tube lighting up off standby is following the socket and not the tube I would suspect the socket or (dare I say) output transformer. Inspect the socket for dirty or bent pins (as suggested) and also cold solder joints at the socket and at the terminations for any leads from the socket. In fact, maybe just reflow all those contacts anyway. Unplug the tubes and check the ohms for the OT from the CT to pin 7 of each socket. Please do this and report.

Not to bash any known good sounding design but I'm surprised to not see any individual screen grid resistors. Always a bad idea for any amp that will ever clip the power tubes.

EDIT: Also check for bad connections on the 470k grid load/bias resistors. Check the values for each individual resistor. Even reflow the joints for those resistors.

The acceptable value of 470k by old school standards is high IMHO for modern tubes. Maybe 330k would be better. I use 220k for these even with cathode bias. The combination of no individual screen grid resistors and high value grid load resistors is going to be very hard on the tubes when clipping.

EDIT 2: Sorry to move new opinionated issues into the thread regarding the grid loads and screen grid R's. I don't mean to get all esS Gee eM on you guys. It just seemed like a good thing to mention that I honestly hope can improve tube life for the OP.

[mfreqmaster]

Here is what I did tonight:
-Pulled the power tubes & inspected, the pins seem ok.
-Checked the solder connections to both power tube sockets, the also seemed fine. I reheated all of the connections to V5 anyway.
-Measured the output xfmr primary from CT to each tube's pin7 after unplugging the power tubes, got 337 ohms for V4 & 316 ohms for V5.

-Did voltage checks of both power tubes:
V4 pin 2 22mv pin 3 12V pin 7 346V pin 9 357 V
V5 pin 2 29 mv pin 3 12V pin 7 345V pin 9 358 V

Also, instead of the 470K ohm resistors & used a dual ganged 500K pot as my PPIMV. The signal then goes to 8.2K resistors to each power tube pin 2. I checked each resistor & they measure at or about 8.2K. Would the 8.2k resistors be the screen resistors Chuck is referring to?

Powered up, the same issue exist. I did notice that V5 seems to be running hotter than V4. Not quite red plating but getting there. I'll check the filament connections tomorrow, for some reason I assumed they were fine. I often get in trouble when I assume....

[mfreqmaster]

Sorry guys, when I said I switched tubes, I thought I had actually switched the tubes. It appears I had both power tubes out at the same time so when I 'switched' them, I just put the same tubes back in their place. I switched them tonight for real. This time the V4 power tube is the one that lights up, although it doesn't do it every time. Out of the last 5 start ups, it's done it only once(light up brightly).

The other thing I did tonight was to switch preamp tubes V2 & V3. (Couldn't swap V1 because this is now an EF86). I do not hear the crazy crackle. whew, this will be good enough for me. If V4 lights up again, I'll consider a new set of power tubes as they are nearly a year old.

Thanks for all who answered & Chuck H, I do consider your suggestions, even if my typing choice of words does not reveal that. I'd be interested in learning more about the pros & cons of 470k grid load resistors vs. 330k or 220k. I picked up a bassman chassis & there is a new build on the horizon in the next few months...... I thought at one point I had made it to the a$$hole list here to be ignored, I am glad that is not so. Otherwise I would have to join the Dark Side & become Soundguruman's Padawan!

[Chuck H]

The crackle may not have been a bad preamp tube. More likely a dirty tube socket, bad solder joint or oxidized pins on the preamp tubes... Or a bad tube. DEFINITELY get some new power tubes. I'd change out that dual 500k master for a dual 250k fo shizzle.

If you're not cranking the amp, as the master volume would imply, you may be able to get by without better current limiting on the screen grid circuit. But I'd remove the 100r shared resistor feeding the screens and put a 470r (some would go with 1k) resistor at each screen grid pin. If you're not cranking the amp anyway then you won't miss the .5W of power (or whatever) that you may lose.