Starting a 2204 build
- Thread starter ledvedder
- Start date
SpiderWars
Well-known member
I would lift the cap and the wire going towards the pots out and try to suck all the solder out, clean it up, and redo it. Hopefully the wires going under the board can be salvaged as is. I know that when I really take my time when modding/remodding that if I unsolder and disassemble carefully, then suck all the old solder out, then clean it up real good with Isopropyl and Qtips...you pretty much cant tell I modded it.
FourT6and2
Well-known member
Nothing jumps out at me as being off here.
FourT6and2
Well-known member
You're just getting hum/buzz? Or you're also getting HF oscillation?
ledvedder
Well-known member
Just that buzz that's in the video. And the scratching on the preamp pot.
ledvedder
Well-known member
I think I got it all sorted out. I re-did the ground connection at the 50/50 preamp cap, since it also provides ground for most of the preamp. I also reflowed several other joints that didn't look good. Everything seems to be good now. Thanks everyone for helping me get through this. Now it's time to start thinking about how to mod this thing. #34? JEL? Hmmm...
SpiderWars
Well-known member
Here's a JEL-inspired mod I did not long ago. It was a 4-holer so Hi/Lo input each with a gain control in the jack right next to it. I call it the 2203++ mod. You get 4 modes (2 per input, activated by pulling the gain pot
Plexi/2203 Lo/2203 Hi/JEL. The straight up JEL circuit is BRIGHT! Be aware of that. My version is not as bright.
Plexi/2203 Lo/2203 Hi/JEL. The straight up JEL circuit is BRIGHT! Be aware of that. My version is not as bright.
glpg80
Well-known member
I had a hunch it was still related to bad solder joints.
Now I want pictures of the input Jack wiring to the 68k grid stopper and also to the first 2.7k stage. I also want a clear picture of the wiring leaving the 2nd stage back to the board. I think your problem is not enough isolation.
Also I keep telling you this but I think you’re ignoring it. The more you add gain from here, the likelihood you’re going to have to cover the chassis to test what you’re doing since you have fluorescent lights above. More gain is only going to make the circuit even more sensitive.
I also really believe you’re going to have to get really strict with your wiring layout in addition to going all coax from input jacks to first stage. Only one side of the coax is ever grounded. Ground it on the side closest to the tube.
Before making any coax changes to the input, Add the cap back in and this time keep it. While it’s making noise, chopstick around.
Now I want pictures of the input Jack wiring to the 68k grid stopper and also to the first 2.7k stage. I also want a clear picture of the wiring leaving the 2nd stage back to the board. I think your problem is not enough isolation.
Also I keep telling you this but I think you’re ignoring it. The more you add gain from here, the likelihood you’re going to have to cover the chassis to test what you’re doing since you have fluorescent lights above. More gain is only going to make the circuit even more sensitive.
I also really believe you’re going to have to get really strict with your wiring layout in addition to going all coax from input jacks to first stage. Only one side of the coax is ever grounded. Ground it on the side closest to the tube.
Before making any coax changes to the input, Add the cap back in and this time keep it. While it’s making noise, chopstick around.
Last edited:
ledvedder
Well-known member
The buzz was louder without the bypass cap on the 10k. I changed the tube in v1 from a Tungsol to a Chinese, and the buzz significantly decreased. I've seen your recommendation for covering the chassis, but I don't have anything currently to use for that. The shielded coax going to v1 is grounded at the input jack.
ledvedder
Well-known member
That was coming from the 2 68k resistors when I was initially going to make it a switchable 2204/1987. I left it in in case I decide to add that mod in the future. It's disconnected and heatshrinked at both ends.
glpg80
Well-known member
Even though it’s disconnected it can still couple electromagnetic energy and transmit it to another part of the circuit since it’s basically a series inductor that’s sensing whatever magnetic field is nearby. If your intentions are to leave it then you must ground one part of the wire. If that causes problems for functionality then you need to completely remove it. When we suggested bone stock 2204, we really mean bone stock - no additional wiring at this stage. Anything that has no purpose has to be removed here because wiring and layout can cause instability quite easily without hassle. Yes it’s more work, there’s no shortcuts with these amps.
As far as wiring for the first stage, make sure the grid, plate, and cathode of one side of the tube doesn’t cross with the grid, plate, and cathode of the other side of the tube. They’re different phases (and all at various voltage potentials) of amplification and if they are near, you can get buzzing and instability. It matters most in the first few stages because anything that happens there is only amplified down the circuit further and further.
ledvedder
Well-known member
OK, I'll remove that wire and the 2 unused 68k resistors. I'll also confirm none of those wires are crossing.
glpg80
Well-known member
The coax for the first stage can be grounded at the input Jack if you have a chassis ground running there as well for the 1M grid reference resistor which is OK.
The metal cover is easy. Go to lowes and go grab some duct work pot metal/tin and some tin snips. Cut what you need at home for various chassis and you’ll be able to determine if bench EMI is you’re problem or not.
I agree with others, I don’t see any problems voltage wise on the tube pins. All looks normal. Any problems here would be related to layout and amp specific wiring placement.
The metal cover is easy. Go to lowes and go grab some duct work pot metal/tin and some tin snips. Cut what you need at home for various chassis and you’ll be able to determine if bench EMI is you’re problem or not.
I agree with others, I don’t see any problems voltage wise on the tube pins. All looks normal. Any problems here would be related to layout and amp specific wiring placement.
SpiderWars
Well-known member
If you have a sheet pan in the kitchen you can use that to cover the chassis if needed.
V2a
Active member
I just came back to the last two pages of this thread. (but just appease everyone and put a baking tray over the chassis). Hang in there.
I won't reiterate other suggestions. Here are a couple different thoughts, based on your last buzz video and photos.
- I'd replace that large metal oxide resistor with a half-watt carbon film resistor (10k). Smaller might be better here.
- The buzz dropped when you put the alligator clip on the cathode side of the 10k resistor. That makes me think there is an issue with the tube itself.
- what happens if you clip the ground side of the 10k to the ground side of the 2k7 resistor (or some other ground point)?
I won't reiterate other suggestions. Here are a couple different thoughts, based on your last buzz video and photos.
- I'd replace that large metal oxide resistor with a half-watt carbon film resistor (10k). Smaller might be better here.
- The buzz dropped when you put the alligator clip on the cathode side of the 10k resistor. That makes me think there is an issue with the tube itself.
- what happens if you clip the ground side of the 10k to the ground side of the 2k7 resistor (or some other ground point)?
ledvedder
Well-known member
Yes, I'm going to try a baking try later today.
I'm also planning to change the 10k. At the time I installed it, it was the only 10k I had available.
That video was with a Tungsol reissue in v1. I replaced it with a Chinese B, and the buzzing reduced significantly.
I'll try grounding the 10k to the 2k7 later this afternoon.
sahlomonic
Well-known member
Looking at the very last pic, the turret to the right of the board standoff (V1a plate resistor lead I'm assuming here?) - it's hard to tell from the pic but I would revisit that solder joint on the base of the turret. I've had times where I would go over my wiring back and forth to make sure everything was connected, then find out I had a tacked on joint where I neglected to fully solder it, then it cleared up the issue once I flowed solder there. Maybe reorient the grid wire so it's more away from the plate lead wire? May not be the issue since you're at stock 2204 levels of gain, but if you're going to mod it for more gain down the road, it may create more noise if it isn't addressed. But then again I'm no pro here . . .
FourT6and2
Well-known member
The vast majority of your solder joints are cold. Don't mean to be harsh. But... gotta work on your soldering skills a bit more. When there's a hard transition, for example, from a blob of solder on a turret, to the turret itself, that's a cold joint. It hasn't "wetted" properly. Not to toot my own horn, but this is what joints need to look like.
If you fix all the cold joints, it will help your noise floor. Might not cure it 100%, but it will help.
If you fix all the cold joints, it will help your noise floor. Might not cure it 100%, but it will help.
Last edited: