Revox B77 Speed and Calibration: Capstan Servo, Speed Board and Tape Speed

A tape machine keeps time. Everything else it does is secondary to that, because an error in speed is not an error you can correct afterwards. Pitch and tempo are carried in the transport, and if the capstan does not turn at exactly the right rate, and hold that rate as the reels change their load from one end of a tape to the other, no amount of care in the audio chain will recover what was lost.

The Revox approach to this problem is a high torque motor driving the capstan directly, with no belt in between, corrected continuously by a servo loop that reads the motor's own rotation. This page explains how that loop works, how to set it, and how to calibrate the high speed board. It closes with a modification: setting a tape speed the machine was never built for.

Do this work in the right order. Speed measurements taken on a transport with a hardened pinch roller or a glazed brake describe the fault in the mechanism, not the setting of the circuit. If the machine is not yet mechanically sound, start with the tape transport guide. If the power supply has not been checked, start with the electronics guide. Part numbers are in the parts reference.

How the capstan motor holds speed

You may have heard an experienced engineer say of a deck, "that one is FG servo."

The capstan motor is the part of a tape machine that governs time. If it is unstable, pitch wavers, rhythm loses its edge, and the music is spoiled. Beyond the obvious division into AC and DC types, capstan motors differ in something more important: how they hold a constant speed under changing load. The FG servo is one answer to that problem, and it is the answer Revox used.

What the capstan motor actually is

An open reel or cassette deck normally carries two or three motors, and they are not equal in the demands placed on them.

The reel motors do the pulling. They wind and rewind tape, and they need force rather than precision.

The capstan motor sets the tempo. It turns the capstan, the polished metal shaft that traps the tape against the rubber pinch roller and feeds it forward at a fixed rate. Everything you hear depends on it turning at exactly the right speed and not varying.

Three main approaches to driving that shaft have been used over the years.

The three families of capstan motor

AC synchronous, including hysteresis synchronous

Found on early TEAC A-series machines and many older Japanese decks.

The motor locks to the frequency of the mains supply rather than to any internal reference. The construction is simple and the motors are durable, but the approach has two consequences. Speed depends on the local mains frequency, so a machine set up for a 50 Hz region needs a different pulley in a 60 Hz region. And because the reference is the mains itself, there is no way to offer pitch control.

DC servo

Found in most cassette decks and in later Japanese open reel machines.

A DC motor can be sped up or slowed down simply by changing the voltage applied to it, which makes pitch control straightforward and allows a small motor to deliver useful torque. The compromise is that a DC motor with no speed feedback slows down as soon as load increases, so some form of correction becomes essential rather than optional.

The Revox approach: high torque AC with direct drive

The Revox capstan is driven directly, without a belt between motor and shaft. The rotating mass sits on the outside of the motor, an outer rotor arrangement, and that mass acts as a flywheel and smooths the rotation.

Two things follow. Without a belt there is no rubber to perish, which removes one of the most common age-related faults on other machines. And the torque available from a large AC motor is enough to pull tape through a path that has grown stiff with age without the speed falling away.

What FG servo actually means

Neither a large AC motor nor a high performance DC motor holds speed on its own. Apply load and the rotation slows. The FG servo, from Frequency Generator, exists to correct that continuously.

FG: the motor generates a signal proportional to its own speed

Look at the side of a Revox capstan motor. There is a ring of fine teeth machined into it, and a small head, the tacho head, positioned close to those teeth without touching them.

As the motor turns, the teeth pass the head and induce a weak alternating pulse train. Turn faster and the frequency of those pulses rises. Turn slower and it falls. The motor is generating an electrical statement of its own speed, continuously, as a frequency. That is the FG.

Servo: the circuit corrects itself

That frequency is fed to the control board. The board does three things, over and over, far faster than a person could perceive.

It watches the incoming frequency. It compares that frequency against the reference for the selected tape speed. Where the two differ, it changes the drive to the motor to close the gap.

Watch, compare, correct. That loop is what the word servo describes.

An FG servo, then, is a system in which the motor's own rotation produces the signal by which its speed is judged and corrected. The machine holds speed at the start of a reel and at the end of it, with a full spool on one side or the other, because the correction never stops.

Why the combination matters on a Revox

The capstan control board is the brain of this loop. It takes the FG signal from the tacho head as its measure of current speed, amplifies it, compares it against a reference derived from a timer circuit, and controls torque through power transistors driving the AC motor.

What distinguishes the Revox arrangement from a typical DC servo is the torque behind the correction. A high torque AC motor under electronic FG control has the reserve to pull an old, high friction tape through the path at the correct speed rather than merely trying to. That reserve is the physical reason behind the frequently repeated remark that Revox machines sound solid and do not waver.

Motors are not simply turning

It is tempting to think of a capstan motor as a part that just goes round. In a professional machine it is not. It is one half of a conversation that never pauses, speaking in frequency to a circuit that answers in torque.

When someone says a deck is better because it is FG servo, they are not quoting a specification. They are describing the fact that its speed is measured and corrected continuously rather than merely set once and hoped for.

The next time you switch on a B77 and it reaches speed almost at once, that is what you are watching.

The capstan speed control board

In the previous article, we replaced the main electrolytic capacitors.

There are still issues with speed control; please refer to the previous article for details.

Electrolytic capacitor replacement

To isolate the problem, we will focus on servicing the Capstan Speed Control board.
Let's start by replacing the most degraded parts, the X2 capacitors, and see how it goes.

Target: Capstan Speed Control board for B77 Mk I series (1.177.325/326/327).
Purpose: Prevent age-related deterioration of RIFA capacitors (cracking, odor, bursting) and restore speed stability and safety.
Prerequisite: Use X2 safety capacitors rated for across-the-line (L-N) applications (MKP/275VAC recommended).

First, an overview of the Capstan Speed Control (1.177.325/326/327) board.

The vertical row of silver male pins on the far left of the photo (about 14 pins) is P1.

Let's look at the connections on the back.
The red line indicates P1.
The areas circled in red are the thick traces for high current (or power supply).

Image credit: Revox B77 service manual

However, when replacing parts on this board, considering the effort of identifying the fault and the replacement work, it is quicker to swap the entire board with a replacement from eBay.

Therefore, today we will focus on replacing the X2 capacitors, which are prone to frequent deterioration.

Expected benefits of X2 replacement

• Reduced risk of odor, smoke, or breaker tripping at startup
• Prevention of capstan runaway or stoppage malfunctions (cases where X2 cracking is a contributing factor)
• Stabilization of 38/19 cm/s switching and startup (indirect effect: suppression of power supply noise/pulses)

The speed control issues mentioned in the previous article seem largely unrelated to this X2 replacement.

However, the reason we are rushing to replace the X2 capacitors here is that symptoms like stopping when inserted and running at 38 cm/s when removed are highly likely caused by a board fault, so there is a good chance that swapping in a known-good board will fix it.

The X2 capacitor on that known-good board has burst, so we will replace the X2 and then try the board swap again.

Where are the X2 capacitors in the B77? (Capstan Speed Control area)

If you repair these machines, you will see X2 capacitors often.

First, let us build some basic knowledge.

How do you read the numbers on a part when you buy it, for example, 275 V 0.47 µF?
The voltage rating is the maximum continuous AC RMS voltage that the X2 capacitor can withstand. When in doubt, a higher voltage rating is generally better.

X2 capacitors are safety-rated components designed specifically for across-the-line (L-N) use at 100-240 V AC. The capacitors in the Revox B77 must be X2-rated.

They comply with IEC 60384-14 and are flame-retardant, self-healing, and surge-tolerant.

Also, 0.47 µF is read as "zero point four seven microfarads."
It is the capacitance of the capacitor, the ability to store charge, corresponding to the "C" in Q = C × V (charge = capacitance × voltage).

Let us dig a little deeper, even if it is a slight detour.

A thorough explanation of µF (microfarad)

• Reading: microfarad. In spoken language, some people say "uF" (from the practice of substituting u for µ).
• Symbol: µF (Greek letter mu). When you cannot type it, writing uF is acceptable.
• Physical meaning: Capacitance C represents "how much charge can be stored." Charge: Q = C × V. Energy: E = ½ C V². Reactance, which is the "resistance" in AC: Xc = 1 / (2π f C) (the lower the frequency f or the smaller C, the harder it is for current to flow).

Feeling 0.47 µF through formulas (50/60 Hz)

• Xc(50 Hz) ≈ 1 / (2π × 50 × 0.47 × 10⁻⁶) ≈ 6.8 kΩ
• Xc(60 Hz) ≈ 5.6 kΩ ⇒ No-load AC current at 100 V system: I ≈ V/Xc, about 15 mA at 50 Hz; at 230 V system, about 34 mA. This is the "hum and spike suppression" action (reactive current) of an across-the-line X2 capacitor.

Unit conversion (with mnemonics)

• 1 F (farad) = 1000 mF = 1,000,000 µF = 1,000,000,000 nF = 1,000,000,000,000 pF
• 0.47 µF = 470 nF = 470,000 pF = 0.00000047 F (4.7 × 10⁻⁷ F). Note: mF is 1000 µF, but in the market, "mF" is sometimes mistakenly used to mean µF, so be careful with parts marked mF.

Etymology (bonus)

• Farad (F): Named after Michael Faraday, a giant in electromagnetism.
• Micro (µ): The SI prefix for 10⁻⁶ (one thousandth of a thousandth). Example: 1 µF = 0.000001 F, 1 µm = 0.000001 m.

Parts to use (the most important points only)

• Value: 0.47 µF (= 470 nF) or 0.1 µF (= 100 nF) (match the marking on the board or the schematic value).
• Rating: X2 (IEC 60384-14) / 275 VAC (or 305 VAC).
• Dielectric: MKP (polypropylene) recommended (excellent self-healing and pulse withstand).
• Representative series: KEMET R46, WIMA MPX2, Vishay 3382/MKP X2, Panasonic ECQ-U(A/B), etc.
• Always verify on the body marking: "X2", "275VAC", and approval logos such as ENEC/VDE/UL.

Minimum checklist (before purchase)

• Body marking includes X2 / 275VAC / approval logos.
• Lead pitch (often 22.5 mm) and dimensions fit the board.
• Value is the same as original (do not put 0.68 µF in a 0.47 µF spot).
• MKP marking (avoid MKT = polyester).

Tools and preparation

• Soldering iron (around 350-370 °C), desoldering braid or pump, flux.
• Isopropyl alcohol, cotton swabs/brush, heat-shrink tubing.
• Screwdrivers, socket wrench, ESD gloves.
• Multimeter (DMM): AC/DC, continuity, resistance.

Actual replacement photos are in preparation.

Removal and installation (procedure)

A. Removing the existing X2 capacitor

• Alternately heat the two pins on the solder side and clear with desoldering braid.
• Clean carbonization, cracks, and flux residue with alcohol, and inspect for pad damage.

B. Installing the new X2 capacitor

• Form the leads with stress relief in mind (do not pull the legs forcibly).
• Mount so that the marking on the component side is readable, and solder so that the solder flows to the base of the leads.
• If there is a snubber resistor (R) in series nearby, reinstall it with the same value.

C. Refreshing other parts at the same time for efficiency

• Small bridge rectifier (1-2 A class) and the small electrolytic capacitor immediately after it (100-470 µF / 25-50 V): replace at the same time if significantly deteriorated.
• Re-solder the P1 connector (cracked joints are a classic fault).
• Remove and clean the IC2 (Tacho Amp) socket contacts (poor contact is common).

Calibrating the high speed capstan board

〜Open the black box and achieve perfect synchronization of 38 cm/s on the bench〜

Hello.
I'm Kotaro Asahina, a musician.

The Revox B77, the king of vintage open-reel tape decks.
The Speed Control PCB (1.177.325) governs its accurate rotation.
*This article is written with the Revox B77 MK1 in mind, but it is not just a how-to; it is a logical explanation of wisdom so that it can be applied effectively.

Many users replace aged trimmers and capacitors, then install the board in the machine and try to adjust the rotation speed (frequency or period) by feel.
But as recording engineers, let's put in a little more effort and adjust it accurately.

Now, when adjusting, you will be measuring with the board installed in the Revox unit, and there are quite a few tricks to it.

Adjustment requires a very high level of technical skill, and manual dexterity is essential, but even if you are clumsy, it is possible to compensate with various methods.

*The supplementary notes section of the article also guides you on measurement methods when connected to the actual unit.

In this article, we will disclose an "engineering approach" that goes from measuring with the board installed in the unit to extracting the capstan motor board and using an external power supply and signal generator to achieve perfect accuracy on the bench.

In particular, this time we will explain using a top-of-the-line example: a High Speed model (19 cm/38 cm specification) that has already been retrofitted with high-precision multi-turn trimmers.

Capstan Board: Essential Replacement Parts List

First, let's start with the board maintenance.

In reality, 80% of the work is done with these parts replacements.
It's quick to imagine that it's completed in two steps: parts replacement and speed adjustment.

On this capstan board, if it has never been replaced since manufacture, there are many parts lurking like "time bombs" that could stop functioning at any time.

Also, by knowing and stocking replacement parts for the capstan motor board, which is essentially the heart, you will be able to apply this knowledge elsewhere.
Of course, stock may run out domestically or internationally, or you may no longer be able to buy in Japan and have to source from overseas, but as of 2025, we have compiled a list centered on parts that can all be obtained with access from Japan.

Master the maintenance of the capstan board with this list.

Essential Replacement Parts: Defusing the Time Bombs

First, the X2 capacitor.
This is the first place to replace for now.

Almost all Revox units obtained second-hand will have this either blackened or burnt.

Ref	Original spec	Recommended replacement	Notes
C1	0.47 µF / 150V (MP)	0.47 µF / 250V~ (Film/X2)	Polypropylene film (MKP) recommended

It is the largest capacitor at the bottom left of the board; the original is a capacitor using paper called "MP (metallized paper)," which is famous for absorbing moisture over time, causing dielectric breakdown and spectacularly emitting smoke.

Electrolytic capacitors that leak or lose capacitance

Ref	Original spec	Recommended replacement	Qty
C4	47 µF / 25V	47 µF / 35V~ 50V	1
C5	47 µF / 25V	47 µF / 35V~ 50V	1
C12	10 µF / 35V	10 µF / 35V~ 50V	1
C13	1 µF / 50V	1 µF / 50V~ 63V	1
C15	10 µF / 35V	10 µF / 35V~ 50V	1
C18	10 µF / 35V	10 µF / 35V~ 50V	1
C21	1 µF / 35V	1 µF / 50V~ 63V	1

In particular, C4 and C5 are components that should be replaced as a priority.

In the Revox community, they are sometimes called "Frako bombs." Old electrolytic capacitors used in Revox machines, such as those made by FRAKO (gold-colored cylinders), are almost 100% dried up (end of life).
They are a major cause of unstable power supply voltages and motor control problems.
Replace all of them with new, 105°C-rated, long-life types.

A higher voltage rating than the original is not a problem. Since modern capacitors are smaller, it is standard practice to increase the voltage rating (e.g., from 25V to 50V) to improve reliability.

Toshin Kogyo 1HUTSJ470M0 - High-grade electrolytic capacitor for audio, 50V

A high-quality capacitor that pursues "faithful reproduction" to the utmost limit.
This series achieves rich volume and texture, making it ideal for high-end audio equipment.
Category temperature range: -40 to +85°C
Capacitance tolerance (120Hz): ±20%
Leakage current (max.): I = 0.01CV or 3, whichever is greater (2-minute value)
Dissipation factor (max. tanθ): 0.10
Endurance (85°C, 1000 hours, rated voltage applied): Capacitance change within ±20% of initial value, dissipation factor within 200% of specified value, leakage current within specified value
Dimensions: φ(6.3+max0.5) × (11+max1.0) mm (excluding leads)

Purchase link at Sengoku Net Shopping

That one with a risk of short-circuit

There is one item in the list marked "Ta (Tantalum)."
Old tantalum capacitors have a failure mode of "short-circuit."
If they short on a power supply line, the board can burn.
For safety, it is common to replace them with electrolytic capacitors.

Ref	Original spec	Recommended replacement	Notes
C19	10 µF / 16V (Ta)	10 µF / 35V~ 50V (electrolytic)	An electrolytic can be substituted

Replace if you have the capacity

Ref	HS model spec	Material	Function
C8	1600 pF (1.6 nF)	PS (polystyrene)	Speed reference (timing)
C9	4700 pF (4.7 nF)	PC (polycarbonate)	Speed reference (timing)

This is an important area that contains values specific to the High Speed (HS) model.
The material is polystyrene (PS) or polycarbonate (PC), which are excellent parts with minimal aging degradation.
Consider replacing them only if the speed cannot be stabilized.
When replacing, you must adhere to the following "HS values" and use high-precision parts (tolerance 1% to 2%).

Timer IC replacement is also essential

Although the degradation rate is low, it is basically necessary to replace them with new ones.
Unisonic timer ICs available at Akizuki are recommended and have been confirmed to work.
Developed by Signetics (now NXP) in 1972, its ease of use has made it a general-purpose IC with second-source products announced by various companies. Equivalent to TI NE555P.

[[114051] Timer IC NE555L-D08](https://akizukidenshi.com/catalog/g/g114051/)

Important: Trimmer potentiometers and recommended replacement parts

The original open-type trimmers (where the internal mechanism is visible) can cause poor contact, leading to rotational irregularities (wow and flutter).
Replacing them with modern sealed, multi-turn cermet trimmers is strongly recommended.

While trimmers with the same form factor as the original are available, the parts I recommend offer vastly superior precision.
The originals are really only a matter of vintage aesthetics.
With 21 turns of adjustment, these are top-spec components that I highly recommend.

• Note: Depending on the board, the value may be 2.2 kΩ or 5 kΩ, so always check the markings on the actual part in your unit.

The part I have verified to work is the T93YB103KT20, available from Marutsu.

Specifications: Package: Tube; Series: T93; Resistance: 10 kOhms; Power (Watts): 0.5 W, 1/2 W; Tolerance: ±10%; Temperature Coefficient: ±100 ppm/°C; Number of Turns: 21; Adjustment Type: Top Adjustment; Resistive Material: Cermet; Mounting Type: Through Hole; Termination Style: PC Pin; Size / Dimension: Rectangular - 0.382" x 0.197" Face x 0.386" H (9.70 mm x 5.00 mm x 9.80 mm)

However, the pin size of this part is not a direct fit, so I cut the resistor pins and extended them with solder.
There are certainly many alternative trimmers on the market with sufficiently long pins.

TRIMMER 10K OHM 0.5W PC PIN TOP
Type	Description
Package	Tube
Series	T93
Resistance	10 kOhms
Power (watts)	0.5W、1/2W
Tolerance	±10%
Temperature coefficient	±100ppm/°C
Turns	21.0
Adjustment type	Top adjust
Resistive element	Cermet
Mounting type	Through-hole
Termination	PC pin
Size and dimensions	Rectangular - 0.382inch x 0.197inchface（9.70mm x 5.00mm）

With 21 turns, 100°C tolerance, and 10 kOhms resistance, the specifications are more than adequate.

T93YB103KT20

Incidentally, a similar trimmer potentiometer, the GF063P 10 kΩ, is available from Akizuki, but please note that it did not work. I mistakenly purchased it once and it was a failure.

Organizing the replacement parts order list

Before starting the repair, gather the following set:

1. Film capacitor: 0.47 µF (1 piece)
2. Electrolytic capacitor 47 µF: 2 pieces
3. Electrolytic capacitor 10 µF: 4 pieces (including C19)
4. Electrolytic capacitor 1 µF: 2 pieces
5. Multi-turn trimmer: 1 piece (verify value on actual unit)

Simply replacing these with new parts will make the B77's rotational accuracy astonishingly stable, rejuvenating it for decades to come.

Other replacement parts list

Must-replace (electrolytic, tantalum, MP)

Ref	Revox part no.	Value	Spec	Notes 
C1	59.99.0450	0.47 µF	150V / MP	Hazard: this is the RIFA part. Replace with an X2 or film type without exception
C2	59.31.4104	0.1 µF	250V / MPETP	Mains noise suppression. A film type is recommended
C4	59.22.5470	47 µF	25V / El	Electrolytic. Replace with new
C5	59.22.5470	47 µF	25V / El	Electrolytic. Replace with new
C6	59.31.4104	0.1 µF	250V / MPETP	A film type is recommended
C12	59.22.6100	10 µF	35V / El	Electrolytic. Replace with new
C13	59.22.8109	1 µF	50V / El	Electrolytic. Replace with new
C15	59.22.6100	10 µF	35V / El	Electrolytic. Replace with new
C18	59.22.6100	10 µF	35V / El	Electrolytic. Replace with new
C19	59.30.4100	10 µF	16V / TA	Caution: tantalum, can fail short. Replacing with an electrolytic is recommended
C21	59.22.8109	1 µF	35V / El	Electrolytic. Replace with new
C22	59.31.1224	0.22 µF	100V / MPETP	Input coupling. Replace if degraded

Precision maintenance (generally no replacement needed)

Ref	Value	Material	Function
C3	4700 pF	Cer (ceramic)
C7	0.01 µF	PETP (polyester)
C8	1600 pF	PS (polystyrene)	Speed reference (1% tolerance precision part)
C9	4700 pF	PC (polycarbonate)	Speed reference (WIMA FKC-3 or similar)
C10	470 pF	PC (polycarbonate)
C11	470 pF	PC (polycarbonate)
C14	0.01 µF	PETP
C16	4700 pF	Cer
C17	4700 pF	Cer
C20	0.22 µF	MPETP
C23	22 pF	Cer (500V)
C24	22 pF	Cer (500V)

copy

Semiconductors

Ref	Revox part no.	Substitute and notes	Function
IC1	NE 555	General purpose timer IC	Timing generation
IC2	TBA 231	uA 739 equivalent	Op-amp, dual
Q1	BC 107 B	NPN	General purpose transistor
Q2	RCA 411	MJ 411 (NPN Power)	Motor drive, large
Q3	BC 107 B	NPN
Q4	BC 107 B	NPN
Q5	BC 178 B	PNP
D1	B250 C800	Bridge rectifier	Power supply rectification
D2-D9	1N4448	1N4148or equivalent	Switching diode
D10	15 V	Zener diode	400mW (May not be fitted on some machines)

copy

In particular, IC2 (TBA 231) is currently difficult to obtain, so care must be taken not to damage it.

Resistors

Ref	Value	Spec	Notes
R14	10 kΩ	Potentiometer	Trimmer, speed adjustment
R1	47 kΩ	5% 0.25W
R2	47 kΩ	5% 0.25W
R3	68 kΩ	5% 0.25W
R4	10 Ω	5% 0.33W
R5	2.2 kΩ	5% 0.25W
R6	1 kΩ	5% 0.25W
R7	820 Ω	5%	Listed as 820 ohms; some versions use 560 ohms
R8	10 kΩ	5%
R9	22 kΩ	5%
R10	4.7 kΩ	5%
R11	22 kΩ	5%
R12	10 kΩ	5%
R13	10 kΩ	5%
R15	86.6 kΩ	1% metal film	Precision resistor
R16	22 kΩ	5%
R17	1.5 kΩ	5%
R18	1.5 kΩ	5%
R19	10 kΩ	5%
R20	220 Ω	5%
R21	4.7 kΩ	5%
R22	3.3 kΩ	5%
R23	22 kΩ	5%
R24	22 kΩ	5%
R25	22 kΩ	5%
R26	3.3 kΩ	5%
R27	4.7 kΩ	5%
R28	4.7 kΩ	5%
R29	33 Ω	5%
R30	10 kΩ	5%
R31	6.8 kΩ	5%
R32	2.2 kΩ	5%
R33	1 MΩ	5%
R34	10 kΩ	5%
R35	2.2 kΩ	5%
R36	10 kΩ	5%

Basically, replacement is unnecessary, but as mentioned earlier, only the semi-fixed resistors (trimmers) are recommended for replacement because they cause poor contact due to aging.

Revox B77 speed control board connection pin layout (reference)

The letters shown here are the actual checkpoints.
The corresponding periodic waveforms for the letters are

The area above this line is displayed for free.

as shown, and detailed values will be published later.

Basically, if only the period of A is adjusted with the trimmer, the others should automatically align (in theory), but for example, if the tape rotation is abnormal or the running is unstable, debugging is necessary.

During this debugging work, it is important to isolate the problem at all checkpoints.

Basically, measurement of checkpoint A can be performed with the unit installed, but for example, checkpoint G cannot be measured while installed, so during debugging, the method using a stabilized power supply and generator, described later, becomes essential.

Circuit diagram and overall view

Looking at this visual, you can see that pins 7 and 8 are the input.
Pin 8 is hot and pin 7 is ground.

Measuring checkpoint A with the board installed

The issue here is whether to use pin 7 as ground when measuring with an oscilloscope, or if there is another ground point available. The pin spacing makes it somewhat risky (pin 8, the signal input, is adjacent, and there is a high risk of the clip touching it and causing a short).

This is where using P5 as ground comes in.
Tracing the route from pin 7, you can see that P5 can be used as ground.

In other words, the design makes it easy to create a workflow: use P5 as ground with the board installed, measure pin 8 (checkpoint A), and quickly adjust with the trimmer.

So it is important to connect the oscilloscope ground to P5, measure at pin 8, and adjust the period accordingly.

This is because the official service manual recommends measuring the period rather than the frequency.

The recommended values are as follows.

Manage by period T (time) rather than frequency (Hz)

In analog circuits (especially timer circuits using capacitors), the time it takes for the capacitor to charge (the period) is more physically representative than the number of oscillations (frequency).

How to read the data (target values)

1. Period (T): 625 µs (microseconds). Calculating 1 ÷ 0.000625 = 1600 Hz.

At high speed (38 cm/s), the period is 1T; at 19 cm/s, it is 2T.

1. SLOW (low speed): waveform is 2·T (longer period)
2. FAST (high speed): waveform is T (shorter period)
3. Data table: for 1.177.325, T = 625 µs
4. Voltage (U): 1 Vpp, meaning the amplitude from the top to the bottom of the wave is 1 volt.

Revox B77 speed adjustment guide

~ Using an oscilloscope to measure period T and adjust to 625 µs ~

The value in the official data for waveform (A) is "T = 625 µs".
This is the period (width of the wave) of the sine wave coming from the motor's rotation sensor (tacho head).

In other words, if you place the oscilloscope probe on input (A) and adjust the trimmer so that the wave width is exactly 625 µs, the speed will be perfect.
The waveform at A looks like this:

1. Connections (with board installed)

The most important point here is that you measure the input (pin 8), not the output (MJ411).
The signal containing the speed (frequency) is on the input side.

• Black alligator clip (GND): It is best to take ground from P5, but if your skill allows, any suitable ground point will do.

• Probe tip (hook): Connection point: the leg of a component connected to the topmost pin (pin 8 / point A). When the board is plugged in, you cannot grab the terminal. Hook onto a component on the pin 8 line, such as the leg of resistor R36 or capacitor C22 (C2) at the top of the board.

2. Oscilloscope settings (period measurement mode)

Do not just look at the waveform; adjust using the numerical value (time).

1. CH1 settings: Coupling: AC (using AC makes it easier to see the sine wave). Probe: 10X.
2. Vertical (VOLTS/DIV): 50 mV or 100 mV. The tacho signal is weak (below 1 V), so amplify it.
3. Horizontal (SEC/DIV): 100.0 µs or 250.0 µs. Since the target is 625 µs, this range is easy to view.
4. Display measurement (Measure): Press [Measure] button → [Add] → from the [Horizontal] category, select "Period" and add it. The screen will display Period: ??? µs at the bottom.

3. Adjustment procedure (target values)

This is the high-speed model (19 cm/s / 38 cm/s) of the B77, so the following values are correct.

Low Speed (19 cm/s) adjustment

When the switch is set to SLOW (small reel).
Set to "2・T" as per the chart.

• Calculation: 2 × 625 μs = 1250 μs
• Oscilloscope target value (Period): 1250 μs (1.25 ms) (Frequency equivalent: 800 Hz)

1. High Speed (38 cm/s) confirmation

When the switch is set to FAST (large reel).
It becomes "T" as per the chart.

• Calculation: 1 × 625 μs = 625 μs
• Oscilloscope target value (Period): 625 μs (Frequency equivalent: 1600 Hz)

Basically, if this is set correctly, the rest should fall into place.
As mentioned at the beginning, if there is any abnormality, it is fine to think of the following as checkpoints prepared for debugging.

By the way, depending on the trimmer part, the original has the dial facing forward from the back.
The one I use faces downward when installed, but with some ingenuity, it can be made to face forward.

My trimmer was installed by cutting the pins of a resistor purchased at the same time and extending the legs with solder; such usage is completely fine! It does not affect the sound at all.

[Debug Guide] Thorough explanation of measurement points (A to H)

This list is the story of the signal from "entering (A)" to "exiting (H)".

However, some points are quite difficult to measure when installed, so if debugging is needed for those locations, I recommend using the stable voltage and generator debugging method provided in the latter half.

Checkpoints A-D: Measurement Guide & OWON Settings

1. Point A: Tacho Signal

This is where you confirm: “Is the motor spinning? Is the sensor alive?”

• Waveform characteristic: sine wave
• Voltage level: 100-140 mVpp (a very small signal, around 0.1 V)
• Physical location: immediately after pin 8 (input)

Because the signal is extremely small, you need to “zoom in” on the oscilloscope to see it.

1. Coupling: AC

Important: Since it’s a weak signal, cut the DC component to make the wave easier to see.

1. Probe: 10X
2. VOLTS/DIV (vertical): 50 mV

If this is set to something like “5 V”, the signal will be too small and appear as a single line.

1. SEC/DIV (horizontal): 250.0 μs or 500.0 μs
2. Trigger Level: set near the center of the screen (0 V line).

Pass criteria:

• A clean “wave” is visible.
• Voltage (Vpp) is at least 100 mV.
• Period matches the specified value for the set speed (1250 μs or 625 μs).

1. Point B: Squared Signal

This is where you confirm: “Has the analog signal been converted into a digital square wave?”
From here, the voltage jumps up dramatically!

• Waveform characteristic: square wave
• Voltage level: 20 Vpp (about 200 times larger than point A!)
• Physical location: output pin of IC2, etc.

★ Be sure to change the oscilloscope settings here!
If you leave them as they were, the waveform will go off the screen.

1. Coupling: DC

From this point on, we are looking at switching from 0 V to 20 V, so switch back to DC.

1. VOLTS/DIV (vertical): 5.00 V

To see a 20 V wave, increase the scale (5 V × 4 divisions = 20 V).

1. SEC/DIV (horizontal): keep as before (250 μs or 500 μs)
2. Trigger Level: raise the arrow slightly (to around 10 V, the middle of the wave).

• The waveform is “square”.
• The voltage swings fully from 0 V to approximately 20 V (near the supply voltage).
• If point A is OK but point B shows nothing (flat), the input IC is faulty.

1. Points C & D: Shaping

This is where the wave shape is conditioned before being passed to the subsequent circuitry.
It is almost the same as point B, but the phase (timing) is inverted.

• Waveform characteristic: square wave

Looking at the image, you can see that the “bumps” are reversed (inverted) compared to point B.

• Voltage level: 20 Vpp
• Physical location: around transistors Q3, Q4

The same settings as point B are fine.

1. Coupling: DC
2. VOLTS/DIV: 5.00 V
3. SEC/DIV: keep as before

Pass criteria:

• As with point B, a healthy “20 V square wave” is present.
• If the waveform disappears here, or the voltage is low (e.g., only 5 V), there is a fault in the surrounding transistors or capacitors.

Checkpoints E-H: Measurement Guide & OWON Settings

1. Point E: Trigger Pulses

These are the sharp spikes that determine timing.

• Waveform characteristics: Downward sharp spikes. In the image, the voltage drops momentarily from a high level (about 12 V).
• Voltage level: SLOW: approx. 12 Vpp; FAST: 0.7-1 V (or smaller spikes)
• Physical location: Around NE555 (IC1)

Because the spikes are momentary, take care not to miss them.

1. SEC/DIV (horizontal): 250.0 µs or 500.0 µs
2. Trigger Level: Since the waveform sits high (around 12 V), bring the trigger arrow to the upper half of the screen for a stable display.

Pass criteria:

• Regular downward spikes are visible.
• The interval between spikes matches the period (T or 2T) of the set speed.

1. Point F: Comparator Output

This is the result of the speed comparison.

• Waveform characteristics: Rectangular wave (square wave with varying duty cycle)
• Voltage level: approx. 12 Vpp
• Physical location: Near TP1 test point

The settings can remain the same as for Point E.

1. SEC/DIV (horizontal): Keep as is (e.g., 250.0 µs)

Pass criteria:

• A clean square wave is present.
• The signal swings fully from near 0 V to near 12 V.
• When the trimmer is turned, the width (duty cycle) of the square wave changes. This confirms the control action.

1. Point G: Control Voltage

This is the "throttle position" (voltage level).
This is the quietest waveform.

• Waveform characteristics: Nearly DC (a slightly undulating horizontal line)
• Voltage level: DC 10 V to 13 V. Note the "10-13 V" annotation on the chart. It is not at 0 V but floats at a high level. The ripple (U) is small.
• Physical location: Output of the op-amp (IC2)

Settings are the same as for Point F, but the waveform will shift upward on the screen.

1. Position (vertical): If the waveform goes off the top of the screen, turn the left Position knob to lower the 0 V reference line.

Pass criteria:

• The waveform floats at a height of 10 V to 13 V.
• Turning the trimmer causes the line to move up or down.

1. Point H: Motor Drive Output

This is the power sent to the motor.
★ Pay maximum attention here!

• Waveform characteristics: DC with jagged peaks (ripple)
• Voltage level: Can reach around DC 20 V; the ripple amplitude itself is 1.5 V to 2 V.
• Period: 10 ms (milliseconds). Note the unit change! This is milliseconds, not microseconds. This is the 100 Hz/120 Hz component from rectifying the mains frequency (50 Hz/60 Hz).

Up to now we have been viewing in µs, but the waveform at Point H is slow. You must turn the horizontal scale to a slower setting to see the waveform.

1. SEC/DIV (horizontal): 2.000 ms or 5.000 ms. If left at "250 µs", it will appear as a flat line. Be sure to switch to the ms range.
2. VOLTS/DIV (vertical): 5.00 V
3. Coupling: DC

Pass criteria:

• Several large jagged peaks are visible on the screen.
• Turning the trimmer changes the overall height (average voltage) of the waveform.
• If the voltage changes here, the board's function is fully operational.

Especially remember to change the time base at Point H.
You have now mastered all diagnostic points on the board!

The Three Essential Tools and Summary of Settings

Professional settings for adjusting the high-speed model (38 cm/s).

① DC Regulated Power Supply: Takasago GP050-2

This supplies the lifeblood (power) to the board.
Thanks to this unit, you do not need to work inside the Revox B77.

• Set voltage: DC 21.0 V (set precisely with a multimeter)
• Current limit: 0.5 A to 1.0 A (to prevent damage in case of a short circuit)

② Signal generator: DAGATRON 8202

Sends a signal to trick the board into thinking a stationary motor is "spinning".

• Target speed: 38 cm/s (15 ips)
• Output frequency: 3200 Hz (3.2 kHz)

Important: Standard model uses 1600 Hz, but the HS model uses double, 3200 Hz.

• Waveform: Sine
• Output level: Start from minimum, gradually increase until the waveform is visible on the oscilloscope (approx. 0.5 Vpp)

③ Oscilloscope: OWON SDS1104

OWON SDS1104 digital oscilloscope (Amazon)

Visualizes the board's "thoughts" (control voltages).

• CH1 coupling: DC (essential for viewing DC components)
• Probe setting: 10X
• Vertical scale: 5.00 V/div (align 0 V with the bottom of the screen)
• Horizontal scale: 250 µs/div (a speed that makes the 3.2 kHz waveform easy to see)
• Measurement: Display Vmean (average voltage)

Takasago GP050-2 setup guide

This power supply is the "heart" for the board.
Never turn on the power abruptly; always check the knob positions before starting.

Step 1: Check physical switches and shorting bar (before power ON)

First, check two things before applying power.

1. Range selector switch (toggle switch on the right)

There is a switch for 0-50 V 1 A and 0-25 V 2 A. Set this to the lower position, "0-25 V 2 A".
Reason: We need 21 V. The 25 V range allows finer voltage adjustment and has more current headroom, making it optimal.

1. Check the shorting bar (lower terminal section)

Is the center GND terminal and the left − terminal connected by a metal plate?
In the photo they are connected, so never remove it.
Reason: This makes the negative terminal function as ground (0 V).

Step 2: Return knobs to "safe position"

1. CONST VOLT (voltage knob, lower left)

Turn this fully counterclockwise to set it to "0 V".

1. CONST CURR (current knob, lower right)

Also turn this fully counterclockwise for now.

Step 3: Power on

1. POWER switch (toggle switch on the left)

Flip the switch up to turn it ON.
The red CV (constant voltage mode) lamp on the upper left should light up; this is normal.

Step 4: Set the current limiter (safety device)

This is crucial. Even if the board is shorted, we restrict the "faucet" to prevent burning.

1. CONST CURR (current knob, lower right)

Turn this knob to roughly the "10 o'clock to 12 o'clock (center)" position and stop.
Explanation: This sets the maximum current to about 1 A to 2 A. A B77 board normally draws only about 0.2 A, but this prevents a large current from flowing in an emergency.

Step 5: Precisely adjust voltage to "21.0 V"

The analog meter needle is only a rough guide. Use your own multimeter to set it accurately.

1. Connect the multimeter

Set the multimeter to "DC voltage 200 V range".
Touch the red probe to the power supply's + terminal.
Touch the black probe to the power supply's − terminal (the one connected to GND).

1. Raise the voltage (the tuning)

While watching the multimeter display, slowly turn the CONST VOLT (voltage knob) clockwise.
Stop when the multimeter reads 21.0.
Note: The left analog meter needle should also point slightly to the right of "20".

Step 6: Standby complete

The settings are now perfect.

• Voltage: Standby with exactly 21.0 V output.
• Current: The safety device will activate in an emergency.
• Output: Caution! The terminals already have 21 V present.

In this state, proceed to connect to the board (or energize if already connected), being careful not to let the clips touch each other.

DAGATRON 8202 setup guide

Step 1: Power on and mode check

First, set the correct mode.
Most generators should work with similar settings, so adjust as needed.

1. Press the POWER switch (the red button on the far left) to turn it ON. The display showing "8.8.8.8.8.8" should light up red.
2. Counter mode switch (immediately to the right of the POWER button): there is a small square button labeled INT / EXT COUNTER. Set this to the unpressed (protruding) state. Important: if it is not set to INT (internal), the frequency you generate will not be displayed.

Step 2: Waveform selection (generate a sine wave)

Reproduce the tacho-head signal (analog wave) of the B77.

1. FUNCTION buttons (three buttons at the bottom center): from left to right, they are ~ (sine), triangle, and ⊓ (square). Press the leftmost ~ (sine) button.

Step 3: Frequency range selection

Set the "digit" to produce a frequency of 1600 Hz or 800 Hz.

1. FREQUENCY RANGE (button group at the bottom left): from left to right, they are 1, 10, 100, 1k, 10k... Press the 1k button. If in a later step the value cannot reach 3200, press the next button to the right, 10k. Try 1k first.

Step 4: Set the frequency to 1600 Hz

This is the main operation. Adjust while watching the numbers on the display.

1. FREQUENCY dial (the large knob on the far left): slowly turn this knob to set the display to 1.600 (if shown in kHz) or 1600 (if shown in Hz). Tip: if it is difficult to set exactly, a slight error is acceptable.

Step 5: Turn off unnecessary functions (most important!)

To output a clean signal, check that no extra functions are turned on. Getting this wrong will result in a distorted waveform.

1. DC OFFSET (center knob): labeled PUSH/PULL. Push this knob in until it clicks. If pulled out, the waveform will shift up or down.
2. SWEEP function (RATE / WIDTH / SYM knobs on the left): these three small knobs should all be pushed in. If pulled out, the sound will waver or distort.

Step 6: Tentative output level (volume) setting

Finally, set the signal "magnitude".

1. AMPL (small knob above the output terminal on the right): set it to approximately the 12 o'clock (center) position. Later, while observing the oscilloscope, fine-tune by turning right if the waveform is too small, or left if too large.

Setup completion checklist

The generator should now be outputting a perfect signal to fool the B77.

• Is the display around 3200 (or 3.2k)?
• Is the waveform button set to ~ (sine)?
• Are all knobs (OFFSET, etc.) pushed in?
• Is the cable plugged into OUTPUT 50Ω on the bottom right?

If all are OK, the generator preparation is complete!

Simulation start method: connection guide

Stabilized power supply (Takasago GP050-2) connection

Target: electrolytic capacitor "C4" on the board, slightly left of center connect directly to the FRAKO capacitor.

• Red clip (+21V): connection point: upper leg (positive) of C4
• Black clip (GND): connection point: lower leg (negative) of C4

Stabilized power supply connection: why "C4" is correct, not "C1"

When testing the board alone and supplying external power (DC 21V), connecting to the wrong point will not only prevent the circuit from operating correctly but also risks damage.
This board (1.177.327, etc.) has a layout that can visually mislead.

In the lower left corner of the board, there is a very conspicuous large capacitor C1 (0.47µF / X2 capacitor, etc.).

Because it is near the power pins and large in size, at first glance it looks like the "main filter capacitor at the power input."

However, connecting a stabilized power supply here during simulation checks is incorrect.

Reading the schematic, C1 is not between the power line (+21V) and ground.

• Role: Snubber circuit. C1 is connected to the power transistor (Q2) for motor drive and the diode bridge. It is a protective part that absorbs "back EMF (noise and sparks)" generated when the motor rotates. It is, so to speak, the circuit's "shock absorber."
• Why you must not connect power here: This point is part of the motor drive current path (or escape route), not a "reservoir" that distributes stable voltage to the entire circuit (control ICs, etc.).

Following the +21V input line on the schematic, you reach the electrolytic capacitor C4 (47µF) near the center of the board.

• Role: Bypass capacitor (decoupling). The positive terminal of C4 receives the incoming +21V, and the negative terminal goes to GND. Here the voltage is smoothed (stabilized) and clean power is supplied to control ICs like the NE555 and op-amps.
• Why you should connect to C4: C4 is the "power dam (reservoir)" on this board. By connecting the stabilized power supply directly here, you can supply the designed voltage to the heart of the circuit via the shortest path.

When performing board-level testing, do not be fooled by appearances; connect as follows.

• Connection target: Electrolytic capacitor "C4" slightly left of center on the board.
• Connection method: +21V (red): to the positive leg (upper side) of C4. GND (black): to the negative leg (lower side) of C4.

This correctly distributes power to the entire circuit, enabling safe and accurate adjustment.

Of course, ground (GND) can also be taken at P5.
P5 is a very easy-to-use and safe ground point. Continue using it with confidence.

Generator (DAGATRON 8202) connection

Inject the signal from the "topmost pin."

• Red clip (signal input) connection point: Pin 8 (topmost pin)
• Black clip (GND)

For ground, use P5 as in measurements with the unit installed.
Of course, if you find a better point, that is fine.

1. Stabilized power supply: Confirm voltage is 21.0V. Still OFF.
2. Generator: Frequency: 1600Hz (correct to set using the period explained earlier). Waveform: sine wave.
3. Oscilloscope: Vertical: 5V/div. Horizontal: 2.000ms (to observe output waveform H). Input: DC coupling.

Step 2: Power on

1. Turn the stabilized power supply ON.
2. Briefly check that no smoke comes from the board (if connected to C1, it should be fine as long as polarity is correct).

Step 3: Signal injection and observation

1. Observe the oscilloscope waveform. If a jagged wave (or varying DC) appears around 20V on the screen, the circuit is alive.
2. Adjustment: Turn the blue trimmer on the front of the board. If the waveform height (average voltage) on the oscilloscope moves up and down, the board function is normal. Set it to the specified point (the boundary where the voltage switches) and you are done.

Supplement: Explanation of P3, P4, P5 roles

• P3 (y-REFEXT): External reference signal input. Normally the speed reference is generated by the on-board NE555 (IC4), but by inputting an external signal here, the speed can be controlled from an external device. This is a terminal for connecting a varispeed unit, etc.
• P4 (y-SUPPLY): +21V power input. This is the "main power input" in the actual unit. 21V is supplied here via the harness from the power supply unit. *Earlier I said "pin number 2 (second from bottom)" is also a power input, but actually P4 and Pin 2 are connected on the board. Both are power entry points.
• P5 (GND): Ground. The 0V reference point for the entire circuit. In the actual unit, a GND wire is also connected here, commoning the potential with the chassis and other boards.

Why connect via a harness?

These terminals (P3, P4, P5) are provided independently at the top center of the board, separate from the board's edge connector (8 pins).

This is to physically separate the wiring for "optional functions (varispeed)" and "stable power supply" from the main signal lines.

During operation in the actual unit, power is supplied from here (P4, P5) to run the motor.
In this standalone test, we are using the "leg of C4" instead of P4 (or Pin 2) to supply power, but the electrical meaning is exactly the same.

Bug-specific solution list

From here, let's organize and summarize the correction methods for each bug.

Error explanation: Not getting a clean sine wave

A common error during simulation checks is not obtaining a clean sine wave.

This is due to the voltage level.

As evidence, even if a trapezoidal waveform appears in simulation, connecting to the actual unit often results in a clean sine wave.

And that difference originates from the strength (amplitude level) of the input signal.

The appearance of a trapezoidal waveform does not indicate a board failure; it is highly likely caused by the signal input from the generator being too large (strong) compared to the signal from the actual unit's tacho head (rotation sensor).

To explain the reason technically, the input stage of this Speed Control board has an operational amplifier (or transistor) for amplifying weak signals.

This trapezoidal waveform occurs when the output voltage (amplitude) from the function generator is too large relative to the gain of this amplifier.

The amplifier tries to further increase the input signal, but hits the limit of the power supply voltage (e.g., 21V), clipping the top and bottom of the waveform.

This is the trapezoidal wave (clipped waveform).

An analogy: it is the same as shouting too loudly into a microphone, causing the sound to distort.

Then, why was it a sine wave in previous actual unit measurements?

The tacho head (rotation detection sensor) attached to the capstan motor of the actual unit generates electricity through magnetic induction, so the output voltage is extremely weak (in the range of a few millivolts to tens of millivolts).

As a result, because the input is weak, even when amplified by the amplifier, it does not reach the ceiling (limit) of the power supply voltage, and is output as a clean sine wave in its original form.

If you are concerned and want to check whether this board is normal: If your function generator has a knob to adjust the output level (Amplitude), try turning it down to reduce the signal.

On a normal board, as you weaken the input signal, the corners of the trapezoid should disappear, gradually returning to a clean sine wave.

If this can be confirmed, it proves that the amplifier is operating with normal gain and exhibits the same behavior as the actual unit.

38cm/s and 19cm/s cannot be switched

A common symptom in the Revox B77 (Mk I) capstan speed control circuit (Speed Control PCB 1.177.325.00 etc.) is that the speed does not change when the speed selector switch is operated (it remains fixed).

This circuit switches speed by controlling Q3/Q4 based on the voltage input to S-SPEED (Pin 5), thereby changing the characteristics of the trigger signal to IC1 (NE555).

In the case of this fault, it is highly likely that this part is not functioning.

Therefore, the first solution is likely to be replacing Q3, Q4 (BC107B) and the timer IC.

BC107B is often out of stock, but Sengoku Trading has a relatively high stock rate.

Sengoku Trading BC107B

There is also a possibility that the board itself is not receiving the command to switch.

Check whether the voltage at board terminal Pin 5 (S-SPEED) changes (normally, one will be around 0V and the other around +21V).

If the voltage does not change, the cause can be considered to be dirty switch contacts on the main unit side, broken wiring, or missing supply voltage (+21V) to the switch, rather than the board itself.

Remote control dummy plug contact failure: Also pay attention to the "REMOTE CONTROL CAPSTAN SPEED" connector at the bottom of the schematic.
On the Revox B77, if a remote control is not connected here, a dummy plug must be inserted.

Part of the circuit is completed via this (especially GND and control signals), so if the dummy plug pins are dirty or there is a solder crack inside, speed control may stop working or become fixed at a specific speed.

Countermeasures include unplugging and replugging the plug inserted in the rear remote terminal, cleaning with contact cleaner, and checking continuity inside the plug.

Also suspect capacitance loss in capacitors C16 / C17.

These are the 4.7nF capacitors (C16, C17) connected to the collectors of Q3 and Q4.

C16 and C17 are coupling capacitors that shape the rising/falling edges of the tacho signal into pulses to trigger the NE555.

If these deteriorate and their capacitance changes drastically or they leak, switching operation will not be performed correctly.

Try replacing them with new film capacitors, etc.

Also suspect an open circuit in resistor R8 (10kΩ).
However, since this is a mandatory replacement part, it is assumed to have been replaced, but for example, suspect malfunction.

Capstan motor does not turn

The capstan motor of the Revox B77 is a critical component for maintaining stable tape speed. If the motor fails to rotate, first verify the power supply voltages on the motor control board. Check the mains fuse and the motor run capacitor. Inspect the capstan shaft for seized bearings and ensure the motor windings have continuity. If the motor still does not turn, the drive transistors on the motor control board may need replacement. Always consult the service manual for detailed troubleshooting steps.

Setting a non-standard tape speed

Tape speed is normally 19 cm/s or, on high-speed models, 38 cm/s.
However, Akihiko Kaneda, who has always broken conventions with his DC recording method, has written about the possibility of 27 cm/s.

This is described in Chapter 8, "Open Reel Recording Amplifier," of his book *Audio DC Amplifier System* (Vol. 2), which pursues music reproduction that transcends time and space.

This is extremely interesting, so let's immediately try setting the speed to 27 cm/s on a B77.

Reference values

The capstan motor control of the B77 detects the motor rotation with a tacho head and compares its frequency with a reference signal to keep the speed constant.

The manufacturer's reference values for the 38 cm/s (15 ips) high-speed setting are as follows:

• Measurement point: Capstan motor board, check point A (or tacho head signal input)
• Waveform: Sine wave
• Reference period (T): 625 µs (equivalent to 1600 Hz)

Calculation for 27 cm/s

Tape speed and the tachometer signal period (time) are inversely proportional.

If the speed is reduced, the motor rotation slows down, and the time for one cycle becomes longer.

Of course, when changing to other speeds, please use the calculation program at the beginning.

The target period when reducing the speed from 38 cm/s to 27 cm/s

TtargetTtarget​

can be derived from the following formula:

Formula
Target period = reference period x (reference speed / target speed)

With the numbers substituted
Target_T = 625μs × (38cm/s ÷ 27cm/s)
Target_T = 625 × 1.4074...
Target_T ≒ 879.63 μs

In other words, if you aim for a reading of approximately 880 µs on the oscilloscope, you will theoretically obtain a speed of 27 cm/s.

Practice: How to adjust?

Although the calculation gave a target of "880 µs," there is one obstacle.
The question is: "Can the speed be reduced by as much as 30% within the standard adjustment range of the B77?"

In conclusion, it is highly likely that the variable range is insufficient with the standard circuit.
So let's try an approach of changing the resistance value of the semi-fixed trimmer.
The original uses 10 kΩ, but we will try replacing it with a 20 kΩ (or 22 kΩ/25 kΩ) trimmer.

Reason (basis of calculation)

1. Required resistance multiplication factor

To reduce the speed from 38 cm/s to 27 cm/s, the resistance value needs to be increased by a factor of about 1.4.

1. Current simulation

Assume that the fixed resistor on the board is 4.7 kΩ and the current trimmer (10 kΩ) is used at around the middle, 5 kΩ, to achieve 38 cm/s.

• Current total resistance ≈ 9.7 kΩ (4.7k + 5k)

To change this to 27 cm/s (multiply by 1.4):

• Target total resistance ≈ 13.6 kΩ (9.7k × 1.4)

Why 20 kΩ?

13.6kΩ13.6kΩ can be made from a fixed resistor (

• Required trimmer resistance = 13.6k - 4.7k = 8.9 kΩ

Although it seems that the original 10 kΩ trimmer could barely reach this, there are the following risks:

• If the current setting uses a higher resistance value, 10 kΩ will not be enough (it will max out).
• Using the trimmer near its end (near maximum value) makes it prone to instability due to temperature changes, etc.

If you replace it with a 20 kΩ trimmer:

• The variable range expands to 0-20 kΩ.
• The required 8.9 kΩ will be near the center of the trimmer, allowing you to use the area that is easiest to adjust and most stable.

• Recommended replacement: 20 kΩ or higher
• Type: 21-turn (multi-turn) type

When ordering parts, look for "20 kΩ 21-turn cermet trimmer" (made by Copal, Bourns, etc.).

The author will add further notes after experimentation.

Summary

When making a large speed change to 27 cm/s on a Revox B77, the following two points are important:

1. Since the standard trimmer is out of range, it is essential to replace it with a multi-turn trimmer with a higher resistance value.
2. Aim for a target period of approximately 880 µs.

Because the Revox B77's speed control is an analog servo system, it is possible to create any tape speed by adjusting the period time on an oscilloscope according to the calculation.

This time, the calculation resulted in "880 µs for 27 cm/s," but if you want other speeds, you can calculate them by applying the formula in the code block above.

Please use this as a reference when archiving special tapes or making experimental recordings.

Measuring rather than trusting

Every adjustment on this page ends the same way: with a measurement. A test tape and a frequency counter, or a recorded reference tone played back and measured, will tell you in seconds what an hour of listening will not. Speed error of a fraction of a percent is inaudible on a single machine and obvious the moment a tape recorded on it is played somewhere else.

That is the reason to calibrate against a standard rather than against your own ears. A machine set up by ear is set up to sound right to you on the day. A machine set up to a reference will still be correct in ten years, and tapes recorded on it will play correctly on someone else's deck.