Why does my 1977 2103 sound much better than my friend's 2203x?

[greengriff]

This is a serious question to which I'm hoping to discover the answer, not an 'aren't vintage amps great' thread.

I have a stock 1977 2103. The original McKenzie speakers aren't great, so we'll forget about them. I had a chance to take it along to a friend's. He has a 2203x, a 1960TV with UK Greenbacks, and an oversized Blackstar cab with Vintage 30s, so we thought it would be fun to do a back to back comparison. The main difference is that the reissue has, at least at the volume levels we tried it at, a harsh treble response that you can almost feel rather than hear. It's also a little thin sounding. The 77 has a nice treble and a bit more punch and body to the tone. These differences were apparent through both cabs.

The only difference in the circuit as far as I am aware is that the 2103 has an extra cap across the mid control to help tune the tone for use in the open back combo format. If mine has this cap then it's not on the pot so it must be on the PCB somewhere. Could that be responsible? If not, what is it? Maybe Marshall chose to voice the amp like that for some reason? They claim the 2203x was patterned after the best sounding 2203 they could find.

Both amps are biased to circa 33ma. The test was done using the same guitars and even the same cables to minimize differences. I'm really curious about this.

[nickb]

You don't say whether you tried the amps through the same speakers. It rather sounds like you didn't. The speakers have a massive impact on the tone so you should repeat the test and then tell us the results.

[greengriff]

Both amps were tried back to back in the same room, using the same guitars, the same cables, and the same speaker cabs.

[pdf64]

Maybe consider swapping the valves over too (and re-checking the EL34 bias)? Perhaps your amp has an old Mullard ECC83, which can have a softer response, see a certain person's page on this https://bmamps.com/v01/home/techie-c...parison-tests/
Does your JMP still have a 1nF bright cap wired to the gain pot?
Another factor to consider is that old components can exhibit value drift; old ecaps might develop higher ESR; may need to check everything to get to the bottom of this

[Helmholtz]

Could be many reasons. Different tubes, different PT, different OT, different (value?) filter caps, maybe different voltages. Hardly something as simple as some resistor or cap value difference of 20% or so.
Actually I never played a reissue amp that sounded exactly like a good original. OTOH originals varied considerably too.

[greengriff]

Maybe consider swapping the valves over too (and re-checking the EL34 bias)? Perhaps your amp has an old Mullard ECC83, which can have a softer response, see a certain person's page on this https://bmamps.com/v01/home/techie-c...parison-tests/
Does your JMP still have a 1nF bright cap wired to the gain pot?
Another factor to consider is that old components can exhibit value drift; old ecaps might develop higher ESR; may need to check everything to get to the bottom of this

I'll read that article, thank you. My JMP still has the stock bright cap. Preamp valves are new unbranded (Chinese I presume). I've had them sitting around for a while so I don't recall where I got them. Power valves are new JJ. His amp was all Marshall branded valves - white logo in v1, gold logo in 2 and 3, then 4 white logo el34s. Can remember the code but could get him to send a photo. Great idea about swapping all the valves too. I'll have to arrange another meet to do that. As it happened I checked every component on my amp that I could check (without desoldering) and everything was to spec, except some filter caps which I replaced. As far as I can reasonably tell my amp sounds how it's supposed to sound, and his does too. One difference I did notice is that the Presence control on the JMP had less of a sweep than the 2003x.

[Enzo]

Even if all the tubes were the exact same make and model in each amp, tubes vary quite a bit. SO the JJs in one amp can sound different from the JJs in the other amp.

[greengriff]

Even if all the tubes were the exact same make and model in each amp, tubes vary quite a bit. SO the JJs in one amp can sound different from the JJs in the other amp.

I intend to meet up again and do a full tube swap ASAP and compare again. I'll report back once I've done it.

[Pedro Vecino]

Countless elements. In addition to the ones mentioned above, the shielded cable lines in the modern 2203 make a difference in themselves.
But the most important difference must be the high voltage. I have had a pair of 2103 recently and its high voltage (with 230VAC in Spain) was:
335V adjusted to 240VAC
370V adjusted to 220VAC (6.42V in the heaters).
I selected the latter in both.

2203 reissues usually have 455V.

[greengriff]

Countless elements. In addition to the ones mentioned above, the shielded cable lines in the modern 2203 make a difference in themselves.
But the most important difference must be the high voltage. I have had a pair of 2103 recently and its high voltage (with 230VAC in Spain) was:
335V adjusted to 240VAC
370V adjusted to 220VAC (6.42V in the heaters).
I selected the latter in both.

2203 reissues usually have 455V.

My 2103 had an anode voltage 475v when I got it. Once I'd piggybacked the bias resistor (covered in another thread) and was able to bias the amp properly, the anode voltage came down to 445v.

[greengriff]

OK, so I went over for lunch and we swapped valves. Firstly we swapped the power valves only. That made no audible difference. Then we swapped preamp valves and that did make a difference - it moved the tones of the amps slightly towards each other in that it took a bit of the high harshness from his amp, and added it to mine. However both amps' general character differences remained, in that his 2203x sounded thinner, harder and less dynamic than my 2103. Interestingly the difference was magnified rather than masked when we added a boost pedal (TS808, level max, drive off): The difference in dynamics became more pronounced. Anyway, then I ran out of time and had to leave. Maybe another time we can exchange individual preamp valves, but I'm not expecting much. It would seem to be that the differences are 'baked in' to the circuit design or hardware somehow. I was hoping that somebody might chip in and say 'ah, that's because of x' but it doesn't look like that will be the case. The bottom line is that the old amp has a more dynamic, 3d sound, with a sweeter treble response, whereas the reissue was harsh(er) and more 2 dimensional. One caveat is that we only got the master on each amp to around 2-3. That's still effing loud, but doesn't bring the power amp into play. Maybe once the master is would up the 2203x will come to life and the 2103 will fall to pieces in a mushy mess.

[Pedro Vecino]

My 2103 had an anode voltage 475v when I got it. Once I'd piggybacked the bias resistor (covered in another thread) and was able to bias the amp properly, the anode voltage came down to 445v.

Sorry for the confusion. I just checked that mine was 2104 (50 watt), not 2103.

[greengriff]

No problem.

[Enzo]

Even if the amps had identical circuits, individual components vary. Things are tighter nowdays than they were back in the 20% resistor days of old Fender amps, but there still are variables.

[nickb]

Thanks for clarifying the speaker question.

The circuit math means I really don't believe that tubes in similar condition affect the tone very much at all. Yes I found differences in the tests I did (linked above) but you have to understand that they did have differences in specs and had to be overdriven to really hear a difference. For you to have such an apparently marked difference between the two strongly suggests significant circuit differences. Have you looked inside and thoroughly checked for changes from stock? Sometimes it's not obvious, a component may have been changed or a wire moved. Since I don't have the amp in front of me I'm not 100% sure what the schematics are, perhaps you could post what you have so eliminate any doubt.

If these were on my bench I would at the very least measure the frequency response of both. Ears can be very fickle.