7A12

Manuals – Specifications – Links – Pictures

The Tektronix 7A12 is a 120 MHz dual-trace vertical plug-in for 7000-series scopes. This original 7000-series plugin allows large DC offset voltages, however the offset voltage is not available on the front panel to be measured externally as with the 7A11.

Switches on the front panel control miniature relays in the signal path to select input attenuators, through logic circuits.

The 7A12 was designed by Roy Hayes (who also designed the 3A3 and 3A8 before leaving Tek for HP Loveland in Jan 1971).

Key Specifications

Bandwidth	120 MHz (79xx) / 105 MHz (77xx) / 55 MHz (74xx)
Deflection	5 mV/div to 5 V/div (1–2–5), variable up to 12.5 V/div
Input impedance	1 MΩ // 24 pF
Timing	Delay difference between channels < 500 ps
Channel isolation	At least 3000:1 at 75 MHz
Maximum input	350 V DC + Peak AC at 1 kHz or less
DC offset	At least ±500 divisions

Internals

The 7A12 uses the 155-0032-00 custom IC as a gain stage and the 155-0022-00 for channel switching.

John Addis writes:

There were two mistakes, conceptual errors, that doomed the 7A12 from 'day one'. The first was in deciding to use the 1 GHz f_t Tektronix IC process (called 50/450 for resistivities of the two base layers) for the signal path. This process was Tektronix' fastest, but too slow for a plugin expected to provide 150 MHz bandwidth in the fastest original mainframe, the 7704. At the same time, HP had a 2.5 GHz process used in the 1830A, 250 MHz, dual channel 50 Ω only plugin.

The second 'day one' mistake, later pointed out by Tom Rousseau, was to arrange the two channels side by side. Roy wanted to be able to go from one sensitivity (deflection factor) to another without having to go through all the sensitivities in between as happens with a rotary switch. The vertical row of push buttons was functionally attractive. However, that would require both channels to occupy the same space if on a single, centered, vertical circuit board.

The solution was two separate, mirror image attenuator (V/div) boards, and two mirror image push button V/division selection boards. There was the essential main center board which connects signal and power to a socket in the mainframe. There were also two identical amplifier boards that took clever advantage of an asymmetrical output.

To make matters worse, the two attenuator boards were made with PPO (polyphenylene oxide), an expensive low dielectric constant material to keep the input capacitance low. There were 11 circuit boards, nine of them unique. All of this was obviously extremely expensive.

As part of the 7403 program to bring a low cost entry product into the 7K series of plugins, Tom Rousseau was assigned to design a new dual trace plugin. Realizing the 7A12's problems, Tom put one channel physically above the other. Almost all of this was done on a single circuit board. The plugin had only 80 MHz bandwidth but was much less expensive to manufacture. The design was classical, had none of the complicated offset circuitry, and the channel switch was the same venerable 155-0022-00 used in the 7A12. The expensive PPO circuit board material was still necessary, but Tom used a cam switch, less expensive than Tek relays, to select the deflection factor (sensitivity). It became the 7A18.

Peter Starič, Evaluation Engineer for the 7A12, wrote:

My task was to evaluate the plug-in, designed by that engineer R. H., of whom I have already written. At first, when we met in the new working relation, everything was looking fine, and I had ignored his scoffing at a staff meeting from over a year ago. He explained how the instrument worked, and I began evaluating it. However, as soon as I had found some faults in the operation or – God beware – a design mistake, he had taken this personally and not matter-of-factly as I did. I had to hear heavy words, like “You Yugoslavs do not know anything,” and the like. To avoid any further insults, I stopped giving any suggestions and only wrote the laconic reports of the instrument behavior, buttressed by the data of my numerous measurements.

The problem was that the plug-in amplifier R. H. had designed should be the most used in the new series. Four of five would-be customers had already ordered it, and three of them would cancel the whole order if this plug-in was not available. To accelerate the matter, management assigned design engineer Larry Biggs to assist R. H. and Jim Cavoretto to help me. Jim was a very kind fellow, and we cooperated well. The new situation had become easier for me; I let Jim to communicate with R. H. and to give him any suggestions for the necessary changes. On the other hand, Larry had found some design faults and had correspondingly simplified the plug-in.

The evaluation went on fine, except for a basic design fault, which could not be corrected without extensive mechanical and circuit layout changes, for which there was not enough time. Though Jim and I had all possibilities to prove this necessity by measuring the necessary electrical parameters and to do the thorough mathematical analysis, based on the measured data, neither of us had enough knowledge to do this. (Now forty years after, I know how to do this because I have done the necessary mathematical analysis of a similar circuit in the meantime.)

As even in the third A-phase (normally there is only one) that basic fault was not corrected, management decided to serve the soup before it was fully cooked, neglecting the B-phase entirely. So I had got the production release document — already signed by R. H. and his boss, as well as by my boss Leon Orchard — to sign. Since the plug-in was not yet ripe for the market, I refused to sign. Besides, if I would sign, I would become the convenient main scapegoat, when the customers would discover the fault. Only when the document was modified to include all my suggestions of the mandatory corrections, I had signed.

On the very next day Jim Hinze, who just became my new boss, told me that I was moved from instrument evaluation to the production — to work on the graveyard shift. I felt this as an unjust humiliation. To look for another employer would be futile in that time of general recession. Though my new boss James Baker accepted me and was very friendly, I decided to return back home to Yugoslavia because the anticipated arrival of my whole family in the USA had become impossible.

Tom Rousseau writes:

After the 7A18 project, and becoming familiar with the M84 (155-0078-xx) IC, I conjured up the 7A26 concept based on the general layout of the 7A18. I figured that with the M84, I could more than double the bandwidth of the 7A18 and provide a reasonably higher bandwidth, 1 megohm dual trace plug-in. Furthermore, I offered the possibility of making four vertical plug-ins from one basic design: 7A26, 7A24 (50 ohm), 7A16A (1 megohm, single channel) and a fourth not pursued, 7Axx, single channel, 50 Ohm. I worked on this for a while then popped the idea to management (Oliver Dalton), who agreed to three of the plug-ins.

The 7A26 became the best selling Tektronix plugin of all time. Tektronix literally gold plated a 7A26 and gave it to Tom in recognition.

See https://vintagetek.org/100000-7a26-plug-in/

Links

• 7A12 page @ amplifier.cd
• 7A12 page @ barrytech.com (waybackarchive)

Documents referencing 7A12

Patents that may apply to 7A12

Pictures

• 

  Front view

• 

  Front view

• 

  Attenuator schematic with relays

• 

  7A12 Amplifier PCB

• 

  Left side view

Components

Some Parts Used in the 7A12