2205 Repair Questions

[stephenl]

A friend asked me to take a look at his bandmate's 2205. He said he put in new tubes and it started blowing fuses...
The schematic says that USA units came with 6550's - he had put EL34's in it. I'm not sure what was in it previously - need to ask. Not sure if it was re-biased.
The speaker impedence selector was set to 16 ohms - I need to ask him what he had connected.
Looks like some carbon tracking between pins 1 & 2?
Screen Resistors are 1K, not sure of the wattage.
Looks like someone added grid stoppers to the power tubes - good idea, poorly executed. Attached to flying leads, I'd at least put some heat shrink over them.
Power supply e-caps don't have any bulges - look like they're original to the amp though.
I haven't powered it up yet, I'll do so through the light bulb limiter.

Any common issues you folks have run into with these? I guessing, initially, that he didn't check the bias and it red-plated...

Socket.jpg

Screenshot 2021-06-21 220757.jpg

Power Sockets Inside.jpg

E-Caps.jpg

ECap Outside.jpg

[maxkracht]

Those grid stoppers might be stock. Marshall did/does? some stuff that I find questionable but rarely seems to be an issue. Bad tubes/not re biased would be my guess. Check the char mark to see if it is conductive. Even if it isn't all the way through, might be worth replacing the socket anyway. If it's just a bit on the surface you can scrape it off or ignore. Might be worth replacing the filter caps. Not sure about that model, but Marshall tends to hide one below the board where it's a pain to get to. I would definitely replace the bias supply caps just to be safe.

[roberto]

Turn it on and report the voltages: you can immediately find if bias has been adjusted (I'd guess no) and which tubes were in.

[Stevem]

That mode off failure ( internal shorting ) with EL34s is common, and yes that pin 2 heater to cathode ground short will blow the fuse and so can produce the resulting arc trace on the tube sockets.

The standing up grid snubber resistors is how they are made, and the screen resistors shoul be 5 watt.

6550 outputs need a bunch more bias voltage then the 34s do so I don't think those outputs failed from running too hot, I just don't know if you are able to get the 34s to sound good ( cold clipping ) or put out much power if you can't set the bias voltage without a resistor change down to around -32 on pins 5.

See if you can scrape away the arc trace with a sharp long finishing nail or better yet a scribe.

If those 34s are biased cold ( still set up with 6550 bias voltage settings ) the amp will sound like crap at most any volume level!

[stephenl]

Just looking on Tube Depot - a lot of backordered tubes...
What do you guys think of the JJ EL34's and KT77's ? I won't have a chance to power this up til tonight, but I think these have relatively high B+, >500vdc?

[TUBEDUDE]

I'd grind the socket surface lightly with a Dremel, or replace the socket, and load a pair of Gold Lion KT-88's in there.

[Stevem]

If the owner of the amp plays metal with the amp set for a good level of volume then 6550s would be a better fit

A KT 77 is a Euro version of a 6L6GC basically.

AES has a matched and burned in pair of JJ 34s for 35 bucks.

Euro tubes on the West coast has a matched pair of the same for 38 bucks.

[Leo_Gnardo]

Tube socket discoloration looks to be between pins 2 & 3 - a much more likely spot for an arc. Pin 3 = plate typically 450-500V and likely spiking higher when amp is driven into clipping, which is how Marshall players often get their signature overdrive tone. Pin 2 = filament, practically at ground potential.

Note burnt looking chassis at ground connections near filter caps. Could be due to a resistive ground connection. I would remove bolts holding down the solder tabs, then sand, grind or scrape chassis surface underneath to make sure there's a good connection on reassembly. Would be a good idea to replace solder tabs with fresh clean bright new ones, also replace weak brass bolts with steel. Many times I've had those brass bolts snap either while removing or reassembling so why take the chance?

As it's been pointed out, if the bias voltage hasn't been adjusted for EL34, it's likely the very high bias set for 6550's would leave the EL34's at or near cutoff - hardly conducting at all - sounding like a fly fart in a bottle. Then again, some people might like that. It won't harm the EL34's OTOH you won't get the rip-roarin' CHUNK you expect out of your Marshall. Without modifying anything, measure the bias voltage range you can set with the circuit as-is & let us know. Typically EL34's like to see bias voltage say -37 to -42 volts.

[stephenl]

Powered up with light bulb limiter w/o power tubes -ok
Normal power up w/o power tubes - didn't clear fuses.
Unloaded measurements:
Va = 502vdc
Vb = 500 vdc
Vbias = -41 vdc at pin
Bias range = -35 to -46 vdc

I cleaned the carbon tracks off the socket.

So maybe just a bad tube/tubes?

I haven't done any testing on the OT - hopefully that's not bad.

I don't have any EL34's so I'll have to order some. As a next step - should I put new EL34's in with the bias at max? through the light bulb limiter?

[martin manning]

As a next step - should I put new EL34's in with the bias at max? through the light bulb limiter?

You can start that way. If you can power up with the limiter (tubes in, bias set at max negative), then you should be ok to take the limiter out. The 500V B+ will drop some with load, and you might be close, but I think you might want to get more negative voltage by reducing the resistor feeding the bias circuit from the HT.

[stephenl]

The stock resistor value for the bias feed is 220K for UK/EL34's. It's 150K for US/6550's. This amp has the 220K resistor in that position. Strange, the QC/Tester sticker on the end of the amp indicates it's for the US but its set up for EL34's. The resistor is identical to other 220K's used in the amp...so it doesn't look like it was modded.

[Stevem]

6550s where much harder to get for awhile then 34s in Europe.
One can assume that to get amps out the door they had to revert back to the more shipping sensitive 34s after exhausting there supplies of 6550s.

Your OT is likely fine since a tube that fails that way can't red plate and possibly over load it, also you did not mention if the HT fuse blew also?

[martin manning]

The stock resistor value for the bias feed is 220K for UK/EL34's. It's 150K for US/6550's. This amp has the 220K resistor in that position. Strange, the QC/Tester sticker on the end of the amp indicates it's for the US but its set up for EL34's. The resistor is identical to other 220K's used in the amp...so it doesn't look like it was modded.

What is the date on the quality check tag?

[Ten Over]

There is the other 2205 that clipped using a bridge rectifier and a discrete diode. They always seem to have EL34's in the states.

If you post a picture of the circuit board, I can tell you what you have.

[stephenl]

Thanks Ten Over, that is the correct version w/ the rectifier + single diode for the clipping circuit.

I'm currently waiting for a parts order to continue...

Has anyone tried replacing the stock clipping circuit w/ 24v zeners ala Jose Mods?