It is a common 'tweak' to replace the 15V BC filament lamps with LEDs, and many photos of such can be found on the internet. A caveat to which some people succumb here is to use only 4 or 5 LEDs. The result is then 4 or 5 bright spots and uneven illumination, which doesn't look becomming. To avoid this, I made a board for twelve 3mm LEDs (below left photo). The illumination is very uniform.
I then discovered that the LEDs themslves are reflected in the top edge of the drawer window - ugly! This is cured by cutting a thin strip of opaque plastic and glueing it in the inside of the fascia panel, so it protrudes just 1mm or 2mm into the drawer aperture, hiding the LEDs from the top edge of the window. You can see this if you look in the clearance gap above the window in my far right photo, below.

After running the player for a while like this, I became aware that the flickering of the well lit spinning disc is extremely annoying! Fortnately, it is very easy to slot a carefully cut slip of opaque plastic behind the top section of the window. The plastic slips easily into the groove between the window and the vertical metal trim strips (photo far right)

The camera doesn't accurately show the extreme green of the 575nm LEDs, very pretty to my eyes. But it does show up every speck of dust and fluff on the window pane!

ReVox B225 new LED ciruit board. ⌘

B225 with LED lamps. ⌘

B225 - new LED illumination, but the sight of the rotating CD is annoying! ⌘

Thin opaque plastic masks the LEDs themselves, and hides the irritating flicker of the spinning CD. ⌘

The B215 was ReVox's third ever cassette tape recorder. The first, the D88 was a top-loading language laboratory unit, not for home or Hi-Fi use. Around 1982 came the B710 - the first home hi-fi machine, also sold in the Studer skin as the A710 to broadcasters and studios. The later B215 inherited its tape transport, almost unchanged, from the B710. Four direct drive motors, no belts, no clutches, no mechanical brakes, the pivoting headblock being the only moving part. Twin, independent capstan motors, each with its own PLL system are locked to a common quartz crystal. The Portescap reel motors are directly coupled to the cassette hubs. Braking is entirely electronic, the whole transport being under the control of THREE Philips microprocesors. A novel feature for 1984 was - and still is - the real time tape counter, which displays the elapsed time even when a partially wound cassette is inserted.

The B215 is not quite a solidly built as its B710 predecessor, but its coup-d'grace is the alignment computer. It takes 30 seconds to measure the tape formulation parameters, and then sets itself for optimum bias, equalisation and Dolby level. Although it doesn't specifically recognise IEC Type-3 tapes (ferro-chrome) it does offer "CrO₂ bias at 120uS".

Still more amazing technology hides within the microprocessors, although never fully exploited by ReVox. This machine has a serial data port on the back, which as an input, offers access to all the user controls, play, rew, ff, rec, pause, etc. Plus things like store counter in memory, record cue marker, search for cue marker, goto time mm:ss
As an output, it provides data such as transport status and mode, and counter reading. When linked to the B250 amplifier in a multi-room configuration, all the status data of the cassette deck can be shown on the amplifier's alphanumeric display - even from another room!

The infra-red emitter that detects the end-of-leader/start-of-oxide was faulty - very low output, so I substituted one chopped from an old 5¼" floppy disk drive. However, the wavelength may be unsuitable, since Panasonic cassettes' leader tape is now undetectable, and some of the TDKs (with striped leader) cause it to stop 5 times between the beginning of the leader and the start of the oxide. I believe ReVox changed to visible red LEDs in those role in later production - possibly to cure the same oddities I am discovering.

B215 Cassette tape recorder ⌘
B215 tape drive ⌘	Tape transport ⌘	Rec & Play heads ⌘	Rec & Play heads ⌘

B215 internals. From L to R: microprocessor board, capstan PLL (perpendicular), transformer + tape drive. ⌘

B215 internals 2. From L to R: microprocessor board, capstan PLL (perpendicular), transformer + tape drive, record amplifier, input/output/playback amp (horizontal board), Dolby NR board. ⌘

The B215 has no obvious indicator that it is in record mode; only a small word "RECORD" in the LCD display. The Studer-skinned A721 variant has an LED above the REC button, (shown on the right, the Nov 1987 brochure) but I did not want to drill holes in mine, and nowhere else to put an indicator. I replaced the four incandescent lamps that illuminate the LCD display with 10 green LEDs. I wanted to make things so REC mode would be indicated by a change in the display colour from green to red. A logic signal is present on the record amplifier board (R̅E̅C̅). The display illumination board is quite tiny (15cm x 1cm) and fitting 20 LEDs, resistors and switching logic in that space was too tight a fit for a quick job. I'll leave this for a future endeavour,

New LED & old incandescent lamp circuit boards ⌘

LED Illumination of B215 VU meters. ⌘

The B285 receiver is an integrated amplifier and AM/FM tuner combination. A variant was available without the AM section.
The manual for the power amplifier section claims 2x110W sinewave power, and 2x220W music power.

The AM section on mine was sonically dead, but auto-tuning, centre frequency indication and S-meter functions all worked: clearly a fault in the AM audio section. Most of the AM work is done by a TDA1072 IC, but happily this turned out to be working fine. A Wima polystyrene coupling cap had gone open circuit (quite a rare failure I think).

Because of the way the circuit cards are slotted in side-by-side, it is impossible to measure anything or adjust any trimmers while the B285 is in normal operation. An old computer ATA/IDE ribbon makes a perfect extender for the various modules, allowing them to be fully exposed for testing (see photo below).

B285 Receiver with AM tuner exposed; linked to the interconnection board by an ATA/IDE cable. ⌘

B285 transformer - totally invisible interior; it's cast in rubberised polymer. That Wima class X capacitor (across the 220VAC) on the voltage selector board sometimes goes up in smoke - so I've replaced it. ⌘

ReVox B285 receiver chassis, all modules removed for cleaning. ⌘	B285 FM front end and IF strip. No Mickey-Mouse tin plated screening boxes here - this is 1.5mm steel! ⌘	B285 FM front end (L) and IF strip (R). Each PCB is secured by ELEVEN screws. ⌘	B285 power amplifier stage, with the main heatsink removed. Some overheated TO-220 transistors are to get extra heatsinks. ⌘
New power supply capacitors and aqua-green (575nm) LEDs to replace the incandescent display lamps in all the ReVox components. ⌘	Original display illumination board - two 15V 3W festoon lamps, and new d.i.y LED circuit board - 12 LEDs and 3 dropping resistors to be installed. ⌘	New LED display illumination board installed. ⌘
Displplay with new LED illumination. ⌘	LED display illumination - nightime view. ⌘

Heatsink modification for the B285

The original production run had no heatsinks on the AF driver transistors. Q110=2SC-2238, TO-220 NPN. Q111=2SA-968, TO-220 PNP.
An update changed the NPN device to a TO-126 type device, and put these on heatsinks (see left photo below). However, the PNP TO-220 device was still left free-standing, and it gets much too hot, having a reputation for ultimately failing. Curiously, there are holes punched in the heatsink for transistors in these locations, but the two transistors are bent aside to stand free of the heatsink (yellow arrows). I figured thee must be a reason for this, so I did not try to attach then to the existing heatsink

My 2SA968s looked cooked, although they still worked. I bought new ones and made a free-standing heatsink for them. To get them up high, away from the board, and make clearance above the other components, I extended to transistor legs with the crimp pins from a molex plug (centre-left photo below).

Installed (centre-right photo) they still lean, but should run cool.

While I was in there, I decided that Q109 & Q209 should be heatsinked as well. They also run hot, and my preference is that NO component exept valves should ever be run too hot to keep your finger on (right photo below). On these two I had enough space to use "M"-type heatsink clips, which I prefer, rather than screws and insulating washers through the transistor tag.

Original factory B285 heatsink modification. ⌘

My B285 heatsink modification 1. ⌘

My B285 heatsink modification 2. ⌘

My B285 heatsink modification 3 - Q109 & Q209. ⌘