The Anatomy of a Ratted Debate — LM308 vs OP07
Does the op-amp in a Rat actually matter? I put LM308, OP07, and the Soviet UD1408 on the scope — same circuit, swap chips, see what's really happening.
At Poland Guitar Fest I must have answered this thirty times behind the booth, and every couple of weeks somebody asks me again in an Instagram or Facebook DM, drops a comment under a post, or sends an email — the same question in one of four flavours:
"Is the LM308 worth the hype, or is the OP07 fine?"
"What's the difference between Vintage and Modern on the Bat?"
"What's the difference between West and East on the Bat Cold War?"
"What about the four op-amp options on the Heavy Bat?"
It's all the same question. And like most "which is better" questions in guitar gear, it splits the internet into two camps that talk past each other 🥊.
Camp one says you need a real vintage LM308 or it's not a real Rat. Camp two says they're functionally identical, you're imagining the difference, and here's the math to prove it. Both camps are right — from inside their own perspective. There's a measurable difference at the op-amp's output, and I'll back that up with scope traces and a bit of datasheet math. There's also a real chance your ears, your room, your mix, and your monitoring chain will compress that difference to nothing. Both can be true at the same time, and the honest answer lives in the gap between them.
That's what this article is about. I build pedals where the op-amp choice is a toggle switch on the front panel — the Bat (LM308 ↔ OP07) and the Bat Cold War (LM308 ↔ Soviet UD1408) — same circuit, same passives, same diodes, chips swapped in milliseconds. So today I want to do three things: walk you through the actual history of the chips (it's wilder than you think), put them on the scope so you can see what they're doing differently, and then tell you what I hear — which you're absolutely free to disagree with.
A Condensed (Not Boring) Backstory on These Chips
Before any of this becomes a tone debate, it helps to know what these chips actually are. None of them was designed to live inside a distortion pedal. Two of them were aerospace-grade precision parts. One of them is a Soviet-lineage cousin that's somehow still in production in Kyiv to this day. The fact that any of them ended up screaming into rail saturation behind a guitar is closer to historical accident than engineering intent — and that's part of why this comparison gets interesting.
The LM308 — Bob Widlar's precision chip that became a distortion legend by accident
The LM308 was designed in 1969 by Bob Widlar at National Semiconductor — the same guy who'd already designed the foundational µA702 and µA709 at Fairchild, followed by the LM101 and LM301 at National. It was meant for aerospace work, photodiode amplifiers, and precision integrators. The architectural trick that made it special was the use of so-called super-beta transistors on the input, which gave the chip extremely low input bias current. Great for instrumentation. Fragile under abuse.
Two practical consequences matter for guitar players. First, the chip needs a small external compensation cap added to the circuit, otherwise it oscillates. Second, that compensation makes it famously slow — it can't follow fast-changing signals the way modern op-amps can. Both of those will come back in a minute when we look at why this chip sounds the way it does in a Rat.
One more thing about Widlar himself, because it's relevant to the personality of the chip: this is a man whose colleagues at National Semi documented him taking a sledgehammer to faulty prototypes (the term they coined for it was "to widlarize"), blowing up annoying PA speakers with firecrackers, and bringing a sheep to mow the lawn outside his office because the company wouldn't pay for landscaping. The LM308 was designed by a proper maverick. I don't think that's nothing.
A few facts that almost no Rat owner knows, that genuinely matter:
- 108 / 208 / 308 are the same silicon, binned by temperature spec. The 108 was military/aerospace grade, the 208 industrial, the 308 commercial. The Rat got the cheap version of an aerospace chip — not because it was technically lesser, but because there's no business case for paying military-grade prices for a pedal that lives at room temperature on top of an amp.
- "Same part number" doesn't mean "same silicon". In the '70s and '80s, second-source manufacturing let National, Motorola, Raytheon, and SGS all sell LM308s — but they built their own dies. Ken Shirriff's die-photo teardown of the LM108/LM308 decapped one with a National Semi logo on the package and found the Motorola circuit topology inside. His blunt take: "you can't expect two chips to behave the same just because they have the same name." So when somebody on a forum tells you they only use real National Semi LM308s — they may unknowingly own a Motorola design with an NS logo silk-printed on the package. The only way to truly know what's inside is to decap the chip.
I'm not saying any of this from a textbook. Over years of Bat builds I've gone through LM308s from National Semi, Motorola, SGS, Linear Technology, and ST — sometimes the same brand in three different packages (Motorola alone has shown up in plastic DIP-8, ceramic DIP-8C, and the iconic metal-can TO-99). The oldest batch I worked through was 90 pieces of NS LM308 from around 1978 — properly vintage Widlar-era silicon. My current shelf is mostly ST-branded parts out of their Italian fab. Same circuit, several die variants, and no — they don't all behave identically at the rails. The headline character is the same; the edge behaviour is where the personality lives.
Ever wonder why the LM308 needs an external compensation cap when the famous µA741 / LM741 — the chip inside the MXR Distortion+, DOD 250, and Op-Amp Big Muff — doesn't? In 1968, Dave Fullagar at Fairchild took Widlar's earlier LM101 architecture and added an on-chip 30pF capacitor to create the idiot-proof µA741. Widlar at National stubbornly refused to waste silicon area on a fixed internal cap for the LM301 and LM308, arguing that external compensation let engineers tune the chip's speed and bandwidth themselves. That 1960s Silicon Valley rivalry is the direct reason every LM308 Rat has a 30pF cap wired between pins 1 and 8.
The OP07 — George Erdi's precision successor (1975, PMI)
Six years after the LM308, George Erdi at Precision Monolithics, Inc. (PMI) designed the OP07. It was internally compensated — no external cap needed — and it used a technique called Zener-zap trimming to achieve microvolt-level offset voltage. For precision applications it was the obvious upgrade. Its slew rate sits in the same neighbourhood as the LM308, and that similarity is most of what kept the Rat sounding like a Rat after ProCo swapped chips.
Because yes — they did swap. The LM308 architecture was declared obsolete by the early 1990s. Super-beta transistors were fragile, hard to fabricate consistently, and required external compensation. Three strikes by '90s industrial-IC standards. The OP07 had quietly become the more reliable, better-trimmed, internally-compensated option for every precision application the 308 used to own. So somewhere around 1996, ProCo's engineers tested the OP07DP in the Rat circuit, decided it held up, and rolled the production line over. It wasn't a cost cut — it was an obsolescence handoff. (Worth flagging: the much-talked-about move of Rat 2 production to China was a separate, later event around 2008 — don't conflate the two.)
The two chips share enough DNA that the Rat is still unmistakably a Rat with either one. But "similar" isn't "same," and the gap shows up precisely in the rail-to-rail clipping region the Rat circuit operates in. Here are the differences that actually matter once the chip is being abused inside a distortion pedal — explained in plain English, because no guitarist needs to memorise a datasheet to enjoy this:
- How fast the chip can react. Engineers call this slew rate. Think of it as the chip's reflexes — how quickly it can change its output voltage when the input signal moves. Both chips are slow by modern standards (that's actually a feature here, not a bug — it's a big part of why the Rat sounds warm rather than icepick-bright). And here's the manufacturer-spread twist: on paper, the LM308 is rated at ~0.3 V/µs typical, STMicroelectronics specs their OP07 at ~0.17 V/µs (slower!), and Analog Devices specs their OP07 at ~0.3 V/µs (a dead tie). Factor in a real-world ±30% part-to-part spread, and any individual chip can fall on either side of the line depending on who made it.
- How much "headroom" the chip has at higher frequencies. Engineers call this gain-bandwidth product. Simpler version: how much amplification the chip can still deliver as the signal gets higher in pitch. The LM308 has a touch more than the OP07 (~1.0 MHz with the 30pF cap vs ~0.6 MHz), which means it preserves more upper harmonics through the clipping stage. In the body of a note, that translates to "more harmonic complexity in the upper mids."
- How sharp the transition from clean to clipped is. Engineers call this open-loop gain. The OP07 has slightly more of it, and the practical consequence is that when it clips, it clips more decisively — sharper waveform corners, more pronounced odd harmonics. The LM308 transitions more gradually. On the scope this looks like the difference between a hard corner and a rounded one. To the ear, hard corners read as "brighter and tighter," rounded corners as "warmer and softer."
- How the chip is stabilised. The LM308 needs an external capacitor in the circuit to keep it from oscillating, and the value of that cap actually shapes the chip's high-frequency behaviour — so the circuit designer has a knob to turn. The OP07 has that compensation built in and locked at the factory. Same job, two different solutions, and that one architectural difference alone changes the harmonic signature at the output.
- What the input stage looks like under stress. This is the one I'd ask you to remember. The LM308 uses fragile super-beta transistors — beautifully sensitive in their precision-instrument home but easily pushed out of their linear region when you hammer them with guitar signal. The OP07 uses a sturdier, modern bipolar input that's been trimmed at the factory to behave consistently. Different input architectures recover from saturation differently, and that recovery curve is what your ear hears as "bloom and sag" vs "tight and articulate."
- How the chip behaves after clipping. Engineers call this saturation recovery. The LM308 is slower and asymmetric — it lingers in the rail before un-clipping, and that lingering is itself audible. Notes "bloom" into clipping with that classic Rat sustain feel. The OP07 is faster and more symmetric — clipping starts and ends more cleanly, leaving sharper waveform edges. Same circuit, two different exit behaviours.
- How hard the chip pushes the diodes. The OP07's output stage is a bit stiffer than the LM308's — it can drive the clipping diodes with more authority. Practical consequence: the diodes themselves clip more abruptly, adding another small layer to the "tighter, more articulate" feel.
If you only remember two of those bullets, remember the input stage + saturation recovery pair. That's the mechanism behind everything you'll see on the scope later. The LM308's fragile input partially loses linearity right at the clip threshold, the chip then takes its time un-saturating, and that whole recovery curve is what gets translated into "bloom and sustain" in your hands. The OP07 clips cleaner and earlier, recovers faster, and leaves squarer waveform edges — which translates into "brighter, tighter, more articulate." In player language: the LM308 is more forgiving and singing, the OP07 is more cutting and precise. In scope language: rounder corners vs sharper corners. Two descriptions of the same mechanism.
The Soviet UD1408 — Cold War reverse engineering, alive in my pedals
The Soviet 140 series was the USSR's general-purpose linear IC family, built across multiple fabs from the late '70s onward. Inside that family, the КР140УД1408 is the LM308-compatible member — same super-beta architecture, same pinout, intended as a drop-in for LM308-class designs. Cold War reverse engineering, basically: take a Western precision op-amp the Eastern Bloc couldn't buy, build it on Soviet fab equipment.
And here's the part that surprised even me when I went digging — it's still in production to this day. Kvazar-IS in Kyiv (the same plant that's been making Soviet/Ukrainian ICs since 1962) still lists the КР140УД1408 as current product, and freshly-manufactured stock is genuinely orderable through Ukrainian distribution. So while the Western LM308 has been EOL since the early '90s, the Soviet-lineage equivalent is currently shipping from an active fab in Ukraine — meaning the "vintage Soviet chip" you're buying might actually be a recent production run. Most of the UD1408s I use are Kvazar-marked, though I've also worked through a batch from Alfa in Riga, Latvia — same chip, just a different factory stamp on the package, a small reminder that the Soviet IC program was spread across the whole Eastern Bloc rather than centralised in one place 🦾.
The Accidental Genius of the Rat Circuit — one wrong resistor
So far I've talked about the chips. But the chips aren't the whole story — a Rat is the chips plus the circuit they sit inside, and the circuit is what turns a precision aerospace op-amp into a distortion legend. Without the circuit, the LM308 is just a slow op-amp. With the circuit, it's a Rat. Worth understanding why.
The Rat was designed in 1978 by two ProCo engineers in Kalamazoo, Michigan: Scott Burnham (whose official job title at the company was the wonderfully on-brand "Hippie in Charge of Technology") and Steve Kiraly (the brilliant non-guitarist engineer the team called "Uncle Steve"). They'd been hot-rodding existing distortion pedals together since the early '70s — four years of tinkering before they sat down to build their own from scratch in ProCo's now-famous rodent-infested basement workshop.
While Burnham was experimenting with the op-amp's gain stage, he made a mistake. He fitted a 47 Ω resistor where a 470 Ω one was supposed to go — a ten-fold value swap that pushed the op-amp into a rail-saturating overdrive he'd "literally never heard before." He kept it.
That single wrong resistor is the reason the Rat sounds like the Rat 🤘.
Burnham's 47 Ω / 470 Ω swap wasn't just "more gain." The same resistor also sits inside a small filter that cuts the bass content before the op-amp clips it. That's why a Rat cuts through a mix instead of turning to mud — it's not the diodes, it's not the tone control, it's the same accidental resistor doing two audible jobs at once. One wrong part, two things nobody's been able to improve on since.
Why the chip choice matters here specifically
Here's the part that matters for the LM308 vs OP07 question: op-amp choice is audible in lots of pedals — Tube Screamers, MXR Distortion+, Big Muffs — but it matters in the Rat for a different reason, at a different magnitude.
In a Tube Screamer or a Distortion+, the diodes and the op-amp share the clipping work — the chip's character genuinely matters (ask anyone who's swapped a JRC4558 into a TS), but the diodes still set the headline shape of the distortion.
In a Rat, that balance tips much further toward the chip. Thanks to Burnham's accident, the op-amp is being driven so hard that it becomes the primary clipper itself. It slams into its own supply rails long before the diodes ever get involved. The diodes only shape what's already been ground up at the chip stage.
That's why every parameter we walked through in the OP07 comparison list — slew rate, recovery behaviour, input-stage stress, output drive — actually matters in a Rat in a way it's softer or even absent in other pedals. The Rat is specifically built around the op-amp's worst-case behaviour. So when you swap LM308 for OP07, you're not changing a flavouring agent — you're changing the part of the circuit doing the heaviest lifting.
Six clipping modes, two chips, twelve voicings
This is also where the Bat series earns its keep as a test platform — and it changes the chip-choice question in an interesting way. The original Rat ships with one fixed clipping topology: a pair of silicon diodes to ground after the op-amp output. Hard clipping, period.
The Bat replaces that with a six-position clipping rotary — Boost (no diodes at all), soft germanium, soft MOSFET, hard silicon, hard germanium, hard LED — and pairs it with the op-amp toggle. Same circuit, same passives, different clipping topology on every position. The chip-choice difference is bigger in some modes than others:
- In Boost mode (no diodes anywhere) the op-amp is the only clipper in the path. There's nothing masking its character. This is the mode where LM308 vs OP07 is maximally audible — if you want one-shot proof that the chip matters in this circuit, it's Boost.
- In the hard-LED mode, the LEDs have a much higher forward voltage threshold (~1.8–2.0 V) than standard silicon (~0.6 V) or germanium (~0.3 V) diodes, so the op-amp gets to swing closer to its rails before they clamp. Chip character is preserved.
- In the hard-silicon mode (the OG Rat default), the diodes clamp early and aggressively. Chip differences are subtler — though the chip's timing (the recovery curve) still leaks through.
- In the soft modes (diodes in the feedback loop instead of after the output) the op-amp never reaches the rails at all — the diodes compress its gain progressively. Chip differences shrink to small-signal colour rather than saturation colour. Different mechanism entirely.
Two chips × six modes = twelve distinct voicings per Bat, and across the Bat and Bat Cold War you get the full LM308 / OP07 / UD1408 sweep against the same clipping ladder. That's why I built the platform this way — and it's also why the LM308-vs-OP07 conversation can't really be answered as a single A/B. The right answer depends on which clipping mode you're sitting in.
What I Measured on the Bench — how I set it up
Theory is theory. Let's look at what the scope actually shows when you push the two chips into a Rat circuit. I focused on the LM308 vs OP07 comparison — that's the one most readers came here for — and pulled a single confirming capture from the Bat Cold War to show where the UD1408 sits in the conversation. (Short version of the spoiler: it sits architecturally next to the LM308, and the scope confirms it.)
I wanted the test to reflect what's actually happening when you're playing, not a contrived lab condition. Here's the rig:
- Signal source and scope: Rigol MHO934. Four-channel scope with a built-in two-channel waveform generator, so the test signal and the capture come from the same instrument — no extra preamp, no DAW, no chance of something else colouring the chain.
- Input level: ~500 mVpp sine wave. That's a realistic moderate-strum level for a hot single-coil pickup — not the absolute peak you'd get from a hard pick attack, but well within "this is what your guitar actually puts into the pedal" territory. I tested at both 500 Hz (clean traces, easy to read) and 1 kHz (clearer square-wave corners for the clipping shots).
- One probe on the op-amp's output, before the diode-clipping network — so we can see what the chip is doing on its own, before the diodes shape anything. A second probe on the pedal output for the post-diode waveform.
- Knobs set deliberately, not at noon. Volume wide open (no attenuation hiding what the chip is doing). Tone all the way up — at that setting the Rat's tone control is basically transparent in the guitar range, which is the only honest way to A/B chips without a moving filter masking the differences. Distortion captured at two points: ~1 o'clock (the threshold setting where OP07 has just begun to clip and LM308 is still riding the edge of linear — the most diagnostic moment in the test), and fully maxed (both chips deep into saturation).
All captures below are off the scope's screen — no post-processing, no smoothing, what you see is what was on the screen.
Clean signal, low drive — the part where everyone gives up
The fairest place to start is a clean signal at low drive, because this is where most internet A/B tests stop. Both chips happily pass a clean sine wave, nothing exciting happens, and someone declares the LM308 hype dead.
And honestly — at this drive level, they are basically the same. Two clean sine waves. Anyone declaring victory for either camp here isn't pushing the chip hard enough to see the differences appear.
Threshold drive — where the OP07 quietly clips first
This is the diagnostic shot. Distortion knob at ~1 o'clock. Same signal in, same circuit, same passives. The only difference is which chip is in the socket.
OP07 has clearly begun to clip — you can see the tops of the wave flattening with relatively sharp corners. LM308 at the same setting is still riding the edge of linear, with much rounder, gentler transitions on the peaks. This is exactly what the comparison list earlier predicted: OP07 clips earlier and with sharper corners (higher loop gain, faster saturation), LM308 clips later and with a softer transition (slower input-stage recovery). That's not theory — that's the same scope, same circuit, two chips, ten seconds apart.
In playing terms: at moderate gain on an OP07-equipped Rat you start to hear the dirt earlier and brighter; on an LM308 you'd push the knob a hair further to get a comparable amount of distortion, and what you'd get would be slightly warmer and rounder on the way in.
Full Rat territory — both deep into saturation
Now turn the Distortion knob all the way up. Both chips are deep into the rails. This is where the bloom-vs-bite distinction becomes properly visible.
LM308's waveform is squared off but with a softer roll into and out of the rails — that's the slow saturation recovery showing up visually. The chip lingers in the rail before un-clipping, and the curve as it comes back is what your ear processes as "bloom" or "sustain into clipping." OP07 hits the rails harder, sits there briefly, and snaps back. Sharper corners. Squarer wave. More odd harmonics in the spectrum, which translates to the "brighter, tighter, more articulate" character.
Neither one is better. They're doing different jobs. If you've been playing through a Rat for years and feel like notes "bloom" into the distortion, that's the LM308's recovery curve. If you've been playing through one and feel like the chip is cutting — articulating fast picking cleanly, every note distinct — that's the OP07.
The Soviet check — UD1408 vs LM308 on the Bat Cold War
I pulled one side-by-side capture on the Bat Cold War to settle the UD1408 question: same circuit, same settings, just flip the toggle.
The two waveforms sit right on top of each other — far more similar to each other than either is to the OP07 trace. Which is exactly what the architecture predicts: the UD1408 is a super-beta op-amp from the same family lineage as the LM308, doing the same job with the same kind of input stage and the same kind of recovery behaviour. Where the LM308 sits on the scope, the UD1408 sits next to it. Any small differences I can spot are in the same ballpark as the part-to-part variation you'd see between two LM308s from different fabs — and probably smaller than the spread between two "identical" Rat builds with different passive tolerances.
To my ears the UD1408 is a touch less articulate than the 308 — but I'm intentionally calling that a subjective impression, not a bench claim. The scope says they're cousins, and that's what the bench is for.
What changes once the diodes get involved
One more shot, because this is the most important honest moment in the bench section. So far every capture has been at the op-amp output, before the clipping diodes do their thing. That's the right place to look if you want to see what the chip is contributing. But it's not the only place that matters — the diodes get the last word on what comes out of the pedal.
The differences shrink. Not to nothing — the chip's timing (how it transitions in and out of saturation) is still visible in the way the clipped tops slope — but the diodes have done a lot of levelling work. Which is why with the original Rat's fixed hard-silicon topology, the chip difference is real but subtle. And it's also why I went on about the Boost mode earlier: in Boost mode there are no diodes in the path at all, so what you saw previously is what reaches your ears, undiluted. If you want to A/B chips by ear and not by waveform, Boost mode is where the differences are loudest.
What I Hear (and What I Think You'll Hear)
The scope tells one half of the story. Sitting down with the pedal, a guitar, and an amp tells the other 🎸. Both halves are honest — and they don't always agree on how big the difference is.
Here's what I hear in my own playing, with all the usual caveats: my room, my amp, my hands, my ears, your mileage will vary.
LM308 is the one that blooms. Notes don't just clip — they grow into the distortion. There's a small "give" in the attack that feels generous, almost vocal. Sustained chords swell rather than sitting flat. It's also the one I reach for when I want the pedal to feel forgiving under sloppy picking — the slow recovery curve smooths over inconsistencies that a tighter chip would expose.
OP07 is the one that cuts. Attack is faster, dirt comes in earlier, every note arrives with sharper edges. Fast picking stays articulate where the LM308 would start to smear notes together. It's brighter on the top end and tighter in the low end — less "give," more directness. Less classic-Rat, more modern-Rat — though that framing depends entirely on which Rat you grew up with.
UD1408 leans LM308-ward, which is exactly what the bench predicted. The bloom is there. The "forgiving" quality is there. To my ears it's a touch less articulate at the very top end than a good vintage 308 — but we're talking about a difference you'd need a quiet room and a good monitoring chain to confirm. In a band mix it's a Rat, full stop.
Now the honest part: how much of this actually translates outside my workshop?
- In a band mix, the gap shrinks. Other instruments cover the top-end nuance, the room masks the recovery-curve detail, and what's left is "it's a Rat" — slightly tighter or slightly more sung depending on the chip, but unmistakably the same family of sound.
- In a recording, it's bigger than the mix but smaller than solo through the amp. Microphones flatten some of the dynamic feel; the rest survives. I've A/B-tested chip swaps in DI recordings and the difference is real but you have to be listening for it.
- With diodes engaged on the OG Rat topology, it's subtle. Block 15 is the visual proof — the diodes do most of the levelling, and the chip's signature lives in the timing of the transitions rather than the amplitude. You'll feel it more than you'll point at it in a frequency analyser.
- In Boost mode (no diodes), it's loud. This is where the chip choice is unmistakable, and it's also where I'd send anyone who wants to prove to themselves whether they care about the difference.
The honest one-line summary: it mostly shows up in how the pedal feels under your fingers, not how it sounds in a recording. Which is a more important thing than the internet usually admits — feel is what makes you play more, and "the pedal you play more" beats "the pedal that measures slightly cleaner" every time 😎.
The Mythology Problem — the NOS LM308 you bought online might not be an LM308
A few internet myths deserve quick honesty before this article gets used as ammunition in someone's forum argument.
The NOS market for vintage LM308s is heavily counterfeited. I've been burned more than once over the years of building Bats — it's part of the job at this price point.
The most common fake isn't the dramatic one (a relabeled TL071 — much faster, sounds nothing like a Rat). It's the boring one: a relabeled LM741. Same pinout, dirt cheap, easy to sand and re-mark. In a Rat circuit, a 741 sounds like a slightly grainy clean boost — no bloom, no Rat character at all. Which is the silver lining: even if a fake gets past visual inspection, it's immediately obvious by ear 🕵️♂️.
Quick visual-screen checklist. If any of these fire, treat the chip as suspect until proven otherwise:
- Marking method vs era mismatch. Vintage NS-era parts were ink-pad-printed. Laser engraving on something sold as vintage NS is the biggest red flag. (Modern ST/LT reissues are legitimately laser-marked — this only flags fakes sold as vintage.)
- Repeated date codes across a whole "vintage" batch. Counterfeiters re-mark whole donor reels at once, so a batch where every chip shares the same date stamp is suspect.
- Lead-plating mismatch. Vintage '70s/'80s parts have hot-solder-dipped legs that have oxidised dull-grey by now. Modern matte-tin legs on a chip carrying a '70s date code is a contradiction.
- Confirm by ear. A fake 741 sounds like a flat grainy clean boost with no Rat character at all. If a "NOS LM308" doesn't bloom, it isn't an LM308.
Fakes in finished pedals, Woodcutter legends, and the "passives matter more" argument
And here's the uncomfortable part. These fakes don't just hit DIY builders sourcing one-off chips from eBay — they ship inside finished pedals at scale. The cheap end of the market is the obvious case: a lot of the sub-$80 Chinese "LM308 Rat clones" almost certainly contain relabeled 741s, because that's what's in the bins their assemblers buy from. What's more uncomfortable is that I've seen internals photos of boutique Rat-style pedals — names you'd recognise, sold at three-figure prices on the explicit promise of vintage LM308 inside — where the chip in the socket shows three counterfeit tells at once. I'm not going to name anyone without bench-confirming a unit myself, and I genuinely can't tell whether it's intentional fraud, a builder who got burned by their own supplier, or a builder who just doesn't know what to look for. But it's worth knowing: "real LM308 inside" on a spec sheet doesn't always mean what you think it means, even at the boutique end.
The Woodcutter Rat (and similar legends). A short one. Forum lore claims there's a special early ProCo Rat called the "Woodcutter," built by one specific employee using golden-ear components, sonically superior to the regular production run. Same parts, same line, just a worker's nickname that escaped into the collective memory and grew a legend. Not a real variant. Don't pay vintage-Strat money for one.
"But the passives matter more than the chip, right?" You'll see this argument in every Rat thread, and it's half-true in a way that gets used dishonestly. Yes, passive tolerance is real — two "identical" Rat builds do measure slightly differently, which is part of why mojo-hunting between specific vintage units is mostly chasing tolerance ghosts. But tolerance jitters the baseline by a few percent in either direction. A chip swap changes the clipping mechanism itself — different input architecture, different recovery curve, different envelope. They're not the same kind of variance, and the bench shots earlier in this article are what that looks like with everything else held constant.
Worth being honest about why the "passives swamp the chip" argument gets repeated so loudly in some corners: there's a commercial flavour of it. A few well-known builders ship "many-Rats-in-one-box" pedals that emulate the entire Rat lineage using a single modern op-amp plus a switchable network of passives and clipping diodes. The whole marketing premise of those pedals depends on the chip being interchangeable, and only the surrounding parts mattering. Those pedals genuinely sound great — but it's a framing with a financial incentive baked in. The Bat literally puts two different chips on a toggle, so my incentive is the opposite: I want the chip difference to be real. Read either side knowing which way each bias points.
The takeaway. Chasing vintage NOS LM308 is worth it — for me, anyway. It's a deliberate production choice: I go through the sourcing pain, run the visual screen, and put verified vintage stock into the Bat and Bat Cold War, because the chip earns its reputation on the bench and because a meaningful share of customers specifically want that vintage silicon and tell me it matters to them. The point of this whole section isn't "don't bother, just use OP07" — it's what to look for so you don't get burned if you want to do the sourcing yourself. If you don't want to do the filtering yourself, fair enough — that's also why people buy a Bat. I've already done it.
One honest note for the long view: the LM308 supply is finite. The chip's been discontinued for decades, the NOS pool shrinks every year, and at some point the well runs dry even for builders who source carefully. I've thought about that, and I have a Plan B mapped out for when the vintage stock stops being viable in production volume. It's a different story for a different article — but it exists, and the Bat platform isn't going anywhere when the LM308 finally does 🛡️.
The 30pF Capacitor — one technical footnote worth knowing
This one comes up in every forum thread the moment someone mentions swapping chips, so it's worth thirty seconds to settle it.
The LM308 needs an external compensation capacitor — a small 30pF part wired to pins 1 and 8 — or it oscillates. That's not optional; the chip is unstable without it. And as I mentioned earlier, the value of that cap actually shapes the chip's high-frequency behaviour, so in an LM308 Rat it's a genuine part of the voicing, not just a stability band-aid.
Here's the twist that trips people up: on the OP07, pins 1 and 8 do something completely different. The OP07 is internally compensated — it doesn't need or want an external cap for stability — and those same two pins are instead the DC-offset trim. Same physical pins, entirely different job.
The practical consequence for anyone modding a real Rat: if you socket an OP07 into a board built for an LM308 and leave the 30pF cap in place, nothing bad happens — the chip just ignores it. It's not adding warmth, it's not "de-tuning" the pedal, it's electrically along for the ride. So if you read a forum post insisting you must pull the cap when you swap to OP07, or that leaving it in "darkens" the sound — you can let that one go. It does neither, which makes swapping chips in a socketed Rat completely plug-and-play.
So Which IC Should You Actually Care About?
If you read this far, you've seen the history and the scope traces — so let's make it practical.
- If you want the OG bloom-and-sag character — the classic Rat that sings and sustains and forgives your picking — that's the LM308. Or the UD1408, which as we saw sits right next to it on the scope and is a sane, verifiable, still-in-production way to get the same super-beta voicing.
- If you want a slightly tighter, brighter, more articulate Rat — faster attack, dirt that arrives earlier, notes that stay separated under fast picking — that's the OP07. And I want to be clear: that's not a downgrade. It's a flavour. Plenty of the Rat records you love were made on OP07-era units.
- If you want to try all of this yourself, on the same circuit, same passives, same diodes, and decide with your own ears — that's literally what I built the Bat series for. The Bat puts LM308 and OP07 on a front-panel toggle; the Bat Cold War swaps that second slot for the Soviet UD1408. Flip the switch mid-riff and the only thing that changes is the chip 🎚️.
And if you're the kind of player who wants to go further down the rabbit hole — a quick honest map, because the LM308/OP07/UD1408 trio is the faithful-to-the-original family, but it isn't the only chip that'll physically drop into a Rat socket:
- LM301 — the closest cousin of all. Same Widlar-era National lineage, also externally compensated, just a slightly different input architecture. "Still a Rat, slightly different bloom." ProCo reportedly used 301s in some early units.
- µA741 / LM741 — yes, the same chip counterfeiters relabel as an LM308 is also a legitimate mod in its own right. Darker, grainier, more compressed. Some players love it for doom and sludge; others find it muddy. Not subtle.
- TL071 / NE5534 / OPA134 and friends — the fast crowd. These never slew-limit in the audio band, so the Rat's signature "warm collapse" simply isn't there. You get a brighter, harder, more aggressive distortion with hard-stamped clip corners. Interesting territory — but it stops sounding like a Rat and starts sounding like its own thing.
- OP27 / OP37 — Erdi-lineage cousins of the OP07, faster, for builders who want "OP07 family but without the slew-rate collapse."
The honest framing: inside the LM308 / OP07 / UD1408 family, the chip is a flavour knob on the same pedal. Step outside that family and you're building a different distortion pedal that happens to share the Rat's topology. That's genuinely fun territory — just don't expect Rat character from it.
One production detail that matters if any of this tempts you: I socket every IC in the Bat series. Both chips on the toggle sit in proper machined DIP-8 sockets, not soldered down — so if you ever want to pull one and drop in an LM301, a 741, an OP27, whatever, you can do it with a chip puller and zero soldering. That's not standard at this price point (soldering ICs down saves a few cents), and it's a deliberate choice: the Bat is meant to be a platform for exactly the kind of curiosity this article is feeding.
If Soviet silicon is the part that caught your ear, the Bat Cold War is the obvious start — but it's not my only home for Eastern Bloc chips. The Disambiguation 2 runs Soviet-era ICs in a different circuit for a different goal, if you want to hear what that flavour does outside the Rat topology 🦾.
And the honest sign-off on the whole debate: the people who don't hear a difference aren't wrong. The people who do aren't wrong either. It's a taste call, sitting on top of a real-but-small measurable difference. Pick the one that makes you play more — that's the only benchmark that's ever mattered.
Want to Go Deeper? (Further Reading)
I leaned on a lot of sources putting this together. If you want to keep pulling the thread, here are the ones worth your time, sorted roughly by curiosity level:
- ElectroSmash — ProCo Rat Analysis — the gold-standard technical teardown of the circuit: the gain math, the filter poles, the slew-rate derivation. (Their live site has been down since 2025, so this is the Wayback capture.) If you want the engineering under everything I described, start here.
- JHS Pedals — ProCo Rat Myths, History, and Timeline — a decade of research into the human story: Burnham, Kiraly, the Kalamazoo basement, the 47Ω mistake, the Woodcutter debunk.
- righto.com — Superbeta transistors inside: die photos of the LM108 — Ken Shirriff decaps the chip and shows you what "super-beta" and "same name, different silicon" actually look like under a microscope.
- Rochester Electronics — The Birth of the Monolithic Op Amp: LM108 — the history-meets-engineering bridge on Widlar's design.
- Analog Devices — OP07 datasheet — the official primary source for the OP07 numbers I quoted.
- Analog Footsteps — Guru #3, George Erdi — the human side of the OP07's designer, if the Widlar stories left you wanting the other half.
- PedalPCB forum — LM308 and OP07, the plot thickens — community peer perspective on the exact swap this article is about.
- Kvazar-IS — КР140УД1408 datasheet — the manufacturer's own datasheet for the UD1408, straight from the Kyiv fab that still makes it, for the Cold War-curious. (Ukrainian/Russian-language.)
The Bottom Line
So — does the op-amp in a Rat actually matter?
Yes, and no, and that's the honest answer. There's a real, measurable difference at the chip's output — you saw it on the scope: the OP07 clips earlier with sharper corners, the LM308 blooms in slower and rounder, and the UD1408 sits right next to the 308 like the cousin it is. That difference is real. It's also small, and it shrinks the moment you engage the diodes, put the pedal in a mix, or record it. Where it lives most is in how the pedal feels under your fingers — which, for my money, is exactly where it counts.
Both camps on the internet are right. The measurement people are right that it's subtle. The tone people are right that it's there. The mistake is thinking those two statements contradict each other — they don't. They're the same truth, seen from two sides.
If you want to hear it for yourself, the easiest way is a switch: the Bat (LM308 ↔ OP07) or the Bat Cold War (LM308 ↔ Soviet UD1408). Same circuit, same passives, same diodes — flip the toggle and the only variable that changes is the chip. That's the whole reason I built them the way I did.
Now go plug in and make some noise 🦫🍺
