EVH #12301 has been opened, photographed, and analyzed by multiple professional techs (Friedman, Sklyarov, others). Here's the consensus on what's actually in the chassis vs. what's commonly claimed:
Confirmed Tier 1 (Jose-canonical): - Extra V0 preamp tube (820Ω cathode + 0.68µF bypass per Jose canon) - Cascaded V1→V2 cold clipper (with 2.7k unbypassed cathode on V2a) - Channel-jumper switch on the front panel - Bright cap on Volume I removed entirely (recent Sklyarov teardowns ~2021 confirmed; not just reduced to 500pF). #12301 is a 1968 chassis so stock would have been 5nF; on later post-1969 chassis the equivalent mod is removing the stock 500pF. - Push-pull saturation switch — 2-position (bypassed / unbypassed), not the 3-value (0.68/1/2.2µF) ladder generalized across other Jose builds - Jose Master Volume (1MΩ audio push-pull) - Diode clipping with a Zener pair across the MV wiper. Voltage attribution is genuinely unsettled: 16V is the most-cited voltage in 1980s–2020s online sources; 15V or mismatched pair has been mentioned by Friedman in some recountings of #12301 specifically; 20V is the voltage Friedman recalls as Jose's typical part across the ~60 Jose amps he's serviced ("I got the part number once, but it was a 20-volt Zener" — Headfirst Amps Panel #11, Dec 2025). Treat any specific value for #12301 as approximate. Note: the diode pair is permanently wired across the wiper; the push-pull on the Jose MV switches between two MV paths, NOT between diodes-in and diodes-out (corrects a long-standing misconception, per Friedman, Panel #11) - NFB resistor moved to 100kΩ off the 4Ω tap (Friedman: "always" 4Ω tap, never varied)
Tier 4 (customer-specific, NOT in JMIL as toggles): - Dagnall C1999 choke (replacing the stock smaller-value choke) - *Allegedly* a Drake 1202-132 output transformer (replacing the stock Dagnall C1998) — this is forensically suggested but not 100% confirmed; the original transformer markings are difficult to read after decades of service - Bias set "cranked hot" per Eddie's preference; the actual bias point varies between repair sessions - Variac use — Eddie ran the amp from a Variac at ~90VAC during recording. This is not a mod Jose installed; it's a stage practice. JMIL surfaces this as historical context only.
Things that are NOT in #12301 despite folklore: - No PPIMV (a back-panel master was added at some point and later disabled) - No effects loop - No plate snubber caps (those are Friedman/Cameron) - No 4th gain stage via tone-stack relocation - No dual-gang push-pull MV (the dual-gang pre+post variant is documented on John Sykes's 1987 per Friedman, but #12301 is single-gang pre-TS only — see sykes-1987 for the dual-gang amp)
For JMIL, this means: chase the Brown Sound preset with the single-gang Jose MV (gang B / dual-gang OFF). The Sykes dual-gang configuration is a separate canonical Jose build with its own preset.
Pre-Jose history caveat: #12301 originally shipped to Jose already had work on it from Wayne Charvel before Jose got hold of it. Some scratch marks and component replacements people attribute to Jose are pre-Jose Charvel work. This complicates the "what did Jose do" attribution honestly — the chassis was not virgin from Marshall when Jose started.
The EVH-attribution caveat (Friedman, Dec 2025): the public narrative that "EVH used the Jose mod circuit as his recorded tone" is widely repeated, but the underlying evidence is narrower than the retelling suggests. In his December 2025 solo interview for the Marshall Modification Project, Dave Friedman notes that Eddie's documented relationship with Jose was primarily servicing Eddie's amps (re-tubing, bias adjustments, jack repair, general amp tech work) rather than installing the Jose mod circuit. Friedman, transcript [88:22–88:46]: *"He would use Jose for service and servicing his amps and taking care of his amps."*
The honest reading: #12301 has documented Jose service work on it. Multiple independent teardowns (Friedman, Sklyarov, others) confirm the Tier 1 features ARE in the chassis. What's less settled is the attribution chain — which mod, when, by whom — given Wayne Charvel had hands on the chassis before Jose, and 40 years of retelling have compressed multiple techs' contributions into one name. The Tier 1 mod list is what builders can replicate to chase the documented tonal target; the historical question of "did Eddie specifically rely on these mods in studio" is a separate, less-settled discussion. JMIL's framing is "the Brown Sound preamp signature" — more accurate than "EVH's recorded amp circuit" — and treats Jose, Charvel, and later techs as the multi-author lineage the chassis actually had, without singling out any of them as more or less responsible than the documentation supports.
For JMIL, this means: the Tier 1 mod set in this entry is what builders can replicate to chase the documented "Brown Sound preamp" tonal target. Whether Eddie himself relied on those mods in studio vs. tracked through a stock(er) Plexi pushed hard by a Variac and cab saturation is a separate, less-settled question. The lab's framing of "the Brown Sound preamp signature" is more accurate than "EVH's recorded amp circuit."
Speaker saturation as essential gain stage (Friedman, May 2025): independent of #12301's specific mod-state, the recorded EVH tone is unattainable from preamp work alone. Per Friedman, Headfirst Amp Panel #5 (~[30:41, 31:38]): *"When you do the plexi thing for like Van Halen tone, you're getting all that saturated gooey goodness… the speakers absolutely crying for mercy. If you try and do the Van Halen thing using a load box, you're driving a load box but you're not actually making speakers distort anymore. It's way cleaner. Never as gainy as it is when you do it for real."* The implication: a browser-resident chain like JMIL — or any modeler/load-box rig — can reproduce the *preamp signature* faithfully but cannot fully reproduce the *speaker-saturation contribution* that's load-bearing in the EVH tonal target. See evh-era-plexi-specs for the chassis-era voltage and filtering specs that contribute to the canonical voicing.
Why the toggles in JMIL stop at the preamp-section work: the choke + OT changes interact with the speaker, room, and microphone in ways that make a browser-based simulator misleading. Documenting them as KnowledgeEntries is honest; rendering them as user-controllable toggles would imply a fidelity the surrogate audio engine can't deliver.