Brush Soundmirror "Ribbon" Recorder (1947)

Here is the cover from the instruction book, found inside the wood cabinet. In the text below, there are links to click on to see the associated pictures. Return with your BACK button.

Brush originally did not even refer to the tape as such - it was "Magnetic Ribbon". The original tape was 3M Scotch #100. It was on a paper backing, and ripped if you looked at it sideways. The acetate-backed tapes came later -3M # 110, and #111, which had a very long and successful industry run.

Having seen many consumer and professional tape recorders over the years, I was intrigued by the mechanics. This machine has no mechanical brakes (other than two fixed, felt, friction fingers on the supply turntable to prevent spill and add back-tension); it relies instead on applying DC to the reel motors for "dynamic" braking on the rewind function. The electrical-only pushbuttons do the switching (the FORWARD button also engages the pads for the heads). So, I decided to have a little bit of fun. I had the schematic for the Brush Soundmirror tape recorder, and wanted to create a spreadsheet that modeled the transport's pushbutton selectors. It's here, along with the schematic. The spreadsheet shows under what conditions which motors get AC, DC (for dynamic braking), or nothing. By doing all this, I thought I could predict how the deck should behave, before I turned it on. See below.

Having recently restored a second wire recorder at SPARC, I noticed this machine by comparison has much better fidelity. Yet operationally, the BK-401 was awkward to thread, and tape handling was poor, compared to tape machines that followed shortly after. For example, after recording a passage, the tape has to be re-threaded to allow rewind for replay! With a rubber-surfaced capstan, the tape can't be coaxed backwards with your hands on the reels, not even with the special roller disengaged. The threading path and the transport controls are not intuitive. I can bet Soundmirror owners had lots of tape all over the floor many times by not reading the book! The Webster-Chicago wire recorder is more friendly than this.

This introduces an interesting feature seen on SPARC's example. It has a small pinch roller meant to press on the friction capstan. The wheel itself appears to be a stock roller bearing mounted on a .010 leaf-spring-loaded arm. If it had a rubber surface, there is no evidence of it now. This roller assembly is not seen in the manual pictures, or other pictures I've seen. After paper tapes, with their rough oxide, came acetate backing, then tapes with lubrication added. Tape friction was now lower. It seems that the pinch roller was a retrofit to keep the speed up - the friction alone probably was no longer enough to drive the tape if it were merely wrapped around the drive capstan.

I played the tape that was on the machine, and except for microphone hum, it sounds very good. As of writing, there is still work that could be done. The preamp is noisy, and has hum. The triode in schematic section B didn't work - R10 was open; R9 was 3.2k; R18 was 60k. CR1 needs replacing. As it stands, it plays tapes and records; rewind is weak, but the tape runs on speed!

The track format of the heads is a bit unusual. You might say the record-play track is full-track, but it is quite narrow, while the erase track is generously wider to ensure complete erasure even with the poor tape tracking. Tapes from this recorder produce output on tracks 2 & 3 of a standard 1/4-track. Playing tapes that weren't recorded on this machine had better be from a full-track machine, or at least have been initially bulk-erased. Imagine hearing three programs from a reel you thought was a two-channel recording: channel 1, channel 2, and the unerased remnant of an earlier full-track program between them!

Notice in the picture that the threading requires oxide-in on the supply, but ends up as oxide-out on the take-up. It's bizarre to see the reels turning in opposite directions. Apparently, the designers had wanted to employ standard phono motors that turn CCW. This makes it necessary to have the supply reel feed tape in a CW direction, if it is to rewind correctly.

Full Schematic (probably early production)

Return to the SPARC Broadcast News Page

Virtual Transport Instructions - for the spreadsheet

Partial schematic of the transport controls
Open this spreadsheet to play with the transport controls (my "virtual transport")

You can try out this spreadsheet simulator - no guarantees from me that it's friendly. Try to have fun.

You can simulate operations by changing only the red numbers as allowed. The STOP, REVERSE, and FORWARD are three mechanically interlocked buttons Select 1, or 2, or 3. The START button is momentary - it serves to get the tape in motion for the FORWARD functions (reproduce and record) by applying AC, while over-riding the tape-break switch S2 (while tape bounce settles down). START also invokes momentary FAST FORWARD, for which the FORWARD break switch is unthreaded. By the way, you never thread up both break switches at once.

Example 1: Change STOP to FORWARD (change 1 to 3). Nothing happens. Change START to 1. AC is applied to M2 and M3 and the tape rolls. Set S2 to 1. Now you can let go of the START button (change 1 to 0), and the tape will keep rolling. Press STOP (change 3 to 1).

Example 2: Rewinding ... Thread the tape in the rewind path by setting S1 to 1. There will be DC on the take-up motor. Press REVERSE by changing to 2. Now AC is on the supply motor, and the tape is rewinding. When the rewind is done, S1 will open (go to 0). Now there is DC braking on M1 (full spool) to slow it. On the other hand, if you press STOP before the tape runs out, there is DC braking on M3 to prevent spilling! Neat!

Example 3: Fast-forward... The booklet says to thread the tape directly from reel-hub to reel-hub. S1 and S2 are 0. Select FORWARD (3), then hold START, and fast-forward happens. When you let go of START, the spools will coast, so you MUST press REVERSE! (change 3 to 2). It's in the book. Notice that this is required to apply DC braking to M1, otherwise the tape likely will spill, especially if the supply spool is the more full.

SPREADSHEET EXPLANATION: AC is fed to the STOP switch, and DC is fed to the S1 switch. Within the switch bank, there are four interconnect busses that I named a (S2 left-common), b (S1 left-common), c (to FORWARD switch), and d (S1 right-common). The POWER column contains the logic statements for the four bus states. Each of the three motors has a logic statement as well. Doing this spreadsheet was partly an exercise to improve my skills with Excel logic functions. The results make sense to me. Heaven knows what happens if one is able to depress two or all three of the interlocking buttons at once! (Not tried).

I have verified most of this on the machine, after having freed the machine's capstan/idler bearings which were seized. Obviously, SPARC's machine is a late production version. The break-switch and push-button wiring is changed. S1 and S2 now have only SPDT wiring, just for the AC. The DC braking is present on the take-up reel with STOP pressed (makes sense).

This still is a pretty impressive design for 1946 to have all this work without any mechanical brakes whatever! More modern three-head machines are much simpler - designers realized they needed mechanical brakes (Ampex 300, Revox B-77), at least until they designed a full three-motor servo system such as in the Ampex ATR-100 or late Studer models. Relay logic helps too.

Return to the SPARC Broadcast News Page