[Jukebox-list] Need "Expert" Seeburg Tormat Info
jay at west.net
Sun Mar 4 12:26:18 PST 2001
On Sun, 4 Mar 2001, Jukin' Geo wrote:
> Hello, I had just aquired a Seeburg 220 and this is the first Tormat machine
> that I am doing a full restoration on. I have the manual for the unit and I
> am basically trying to follow how the tormat actually works. I fully
> understand the principle behind the tormat and have basic knowledge. So
> this question is geared more to an advanced techie. Looking at the basic
> schematics for the read in and read out circuits, I can see four different
> current paths through the toroids.
> I figured out that the four paths are as follows. One group selection, One
> single selection within the group, read out, and trip path which runs the
> entire circuit.
Not exactly, but close.
> I understand that the Tormat works by storing induced current in a magnetic
> field around each selected toroid. Then the read out circuit will find
> only the toroids that have a stored charge. Upon readout, the trip circuit
> sences a "disturbance" in the field and activates the trip coil.
> This is all I understand. But what I don't understand is the write in
> circuit and the voltages involved. Does it read in with a negative voltage?
> Does it read out with a reverse voltage? Why isn't the readout voltage
> stored? How is it that two conditions have to be met to write in, but only
> one to read out?
> As you can see these questions are geared to those who knows Tormats inside
> and out.
There's a publication, "The Seeburg Engineering Manual" edited by Mike
Zuccaro, that gives the most succinct and clear description I've seen.
I believe it's available form one or more of the repro and manual shops
for about $15.
Think of the toroid cotes as two-state magnetic devices, similar to the
reversing switch on a mech in principle. They have two stable states, and
require a "push" of a certain magnetic flux to change states, or "flip".
Also they have the following characteristics:
* Flux less than the "critical" push won't have an effect.
* Once "flipped" to a state, another pulse in the same direction has no
* When a core "flips" to the opposite state, the "flipping" action generates
a magnetic pulse that can cause an induced current in a wire passing
through the core.
Each toroid has four wires running through it. Two of the wires are used
for write-in. The selector has a regulated supply with a charge pump that
provides slightly more than half of the required current to flip a toroid.
When a selection is made, one wire through the toroids is connected to
the letter buttons, and another to the number buttons in a matrix. When
a selection is made, all of the toroids corresponding to the selected
letter get a pulse of just more than half enough current via the wire
matching the selected wire. The same pulse also goes through all of the
toroids corresponding to the selected number. All of the toroids which
match that number also get a pulse.
One and only one toroid will have both the letter wire and the number wire.
As the flux is equal to the amperes multiplied by the number of turns, this
toroid alone will receive enough flux to "flip"state, as it has two turns
which are energized, the "letter" wire and the "number" wire.
In summary, a single pulse of half the energy to flip a toroid is applied
through the letter wires and also through the number wires. The one
toroid in the matrix where these wires intersect will receive this pulse
twice, once via each turn, and "flip" state to the "selected" direction
The 100-select machines may use a 20 x 5 matrix instead of the 10 x 10
of letter-number, but the principle is the same. It's letter-number on the
160 and 200 select units.
A third wire, used for read-out, goes from ground through a single toroid
and to a contact located on the bottom of the Tormat unit. There is one
such wire for each toroid, and hence for each unique selection. When the
mechanism scans, a star wheel discharges a capacitor as the mechanism
passes each button. The reversing switch moves the discharge pulse to
another row of contacts for the opposite side of the record. This read-out
pulse is of sufficient energy to "un-flip" (or flip to the opposite state)
any toroids which have been "flipped" by the write-in process. So, every
selection toroid receives a "read-out" pulse per scan, whether it has
been written to or not. A complete scan will re-set the polarity of
magnetization to "de-selected", as every toroid is pulsed through the
contact on the bottom when the read-out capacitor discharges through the
star wheel and the contact.
In other words, applying the read-out pulse destroys or "un-flips" the
selected state of every "selected" toroid, but has no effect on those
which were not "selected".
A fourth wire runs through every toroid in the entire matrix, and is
connected to the RCA plug via a shielded wire. When a toroid changes state
from "selected" to "de-selected", an electical pulse is induced into this
wire due to the change of magnetization of that toroid. This is amplified
and used to trip the mechanism. The cores that are already "de-selected"
don't change state and thus don't induce a pulse. As this pulse occurs
only during read-out, and at the time when the carriage is aligned with
the button contact of a "selected" toroid, the trip solenoid gets pulsed
at the correct instant to play that selection.
This wire also detects the pulses that occur when a selection is made and
a toroid flips from de-selected to selected, but these pulses are of the
opposite polarity and ignored by the amplifier.
The "battery" trick:
Because the detection wire runs through every toroid, even though its
purpose in normal operation is to detect the small voltage generated when
a toroid flips from selected to unselected, it can be used in another
A 1.5 volt "D" battery can deliver enough energy to "flip" the toroids
to the opposite state. So, tapping the tip of the RCA plug from the
Tormat to one end of a D cell while grounding the other will flip all
toroids from selected to unselected or vice-versa, depending on the
polarity of the battery. This is a quick way to test the read-out and
detection circuitry. "Flip" all of the toroids to selected with the
battery, plug the RCA plug back into the selection receiver, and the
box should play every selection, once. The act of reading out clears
> As of now, the only circuit I fully understand is how the trip circuit
> works. I have enough knowledge to fully rebuild this circuit and get it
> going. I just would like to understand more completely how the Tormat
> works. Unlike the older Seeburg pinbank Select-O-Matics, the 220 manual
> doesn't go really into depth on how the Tormat fully functions.
There were numerous refinements and changes as the Tormat evolved. Some
early ones used thyratrons both for write-in and read-out, and the polarity
was changed at one point. Some jukes select with positive battery on the
RCA plug, others with negative battery. There's also quite a bit of pulse
shaping and noise and transient filtering. The 222 is about in the middle
of the development of Tormats, so the early bugs have been more than worked
out. I'd suggest not trying to re-design it, just troubleshoot and fix
what's broken. As with most electronics of this vintage, I'd look at the
capacitors first. Values are more likely to be critical here as the
Tormat is fussy about the shape and energy of the write-in and read-out
Jay Hennigan - Network Administration - jay at west.net
NetLojix Communications, Inc. NASDAQ: NETX - http://www.netlojix.com/
WestNet: Connecting you to the planet. 805 884-6323
More information about the Jukebox-list