155-0078-00: Difference between revisions

From TekWiki
Jump to navigation Jump to search
No edit summary
No edit summary
Line 46: Line 46:
There are several DC specs which differ with part suffix.   
There are several DC specs which differ with part suffix.   


Suffixes -10, -11, -12 and -13 were used with the redesigned M84, known as the M705.  
Suffixes -10, -11, -12 and -13 were used for the redesigned M84, known as the M705.
The M705 was necessary because the original SH2 (2-2.5GHz?) process was being discontinued.  
The M705 was necessary because the original SH2 (2-2.5GHz?) process was being discontinued.  
The M705 was designed by Hal Lilliwhite using the faster SH3 process (6-7GHz).  
The M705 was designed by Hal Lilliwhite using the faster SH3 process (6-7GHz).  
Line 52: Line 52:
The M705 was designed in late 1989 and tested in early 1990.
The M705 was designed in late 1989 and tested in early 1990.


There are two test fixtures and every M84 was subjected to all tests on each test fixture, and all tests were done manually.  
There was one high frequency test fixture and perhaps two dc test fixtures.
Every M84 was subjected to all tests.
All tests were done manually.  
I designed both test fixtures and the 25ps PPPG.
I designed both test fixtures and the 25ps PPPG.



Revision as of 07:26, 16 January 2019

The 155-0078-xx (M084) is a Tek-made broadband vertical amplifier integrated circuit in a minipac package. The input and output are differential. The 155-0078 and its test fixture were designed by John Addis for the 485 scope (Note 1), which uses no less than 18 of these chips in various grades.

The rise time of each M84 was tested using the 25ps Push-Pull Pulse Generator, and parts were binned and were given part numbers with an appropriate speed suffix. Each M84 was also tested for DC characteristics such as balance.

The last two digits of the 155-0078-xx part number are determined by a selection process after manufacture. These characteristics are summarized below:
SUFFIX SPEED DC CHARACTERISTICS REPLACEABLE BY
00 Fast Best ----
01 Fast Third best -03,-00
02 Fast Most relaxed -01,-03,-00
03 Fast Second best -00
04 Slow Same as -00 -00
05 Slow Same as -01 -03,-00,-07,-04
06 Slow Same as -02 -01,-03,-00,-05,-07,-04
07 Slow Same as -03 -00,-04
10 Fast Best ----
11 Fast Third best -13,-10
12 Fast Most relaxed -11,-13,-10
13 Fast Second best -10
Sometimes an instrument will contain a better grade of part than called for in the manual.

The manual should be consulted for the correct replacement type.

The M84 also comes in 14-pin DIP, as 155-0273-00 and 155-0274-00 (slower grade).

Regarding the various grades of M84, John Addis says:

There are several DC specs which differ with part suffix.

Suffixes -10, -11, -12 and -13 were used for the redesigned M84, known as the M705. The M705 was necessary because the original SH2 (2-2.5GHz?) process was being discontinued. The M705 was designed by Hal Lilliwhite using the faster SH3 process (6-7GHz). Additionally, the M705 apparently had the gain setting resistor laser trimmed at the wafer level. The M705 was designed in late 1989 and tested in early 1990.

There was one high frequency test fixture and perhaps two dc test fixtures. Every M84 was subjected to all tests. All tests were done manually. I designed both test fixtures and the 25ps PPPG.

Transient Response (schematic not shown) uses the 25ps Push Pull Pulse Generator (PPPG) and a sampling scope. Sampler load supplies dc current (through X2 pads and sampler input resistance. The sampling scope chassis is floating at (probably) +5V through 8.2 ohms. 25ps PPPG has transformer coupled trigger output to sampling scope trigger input.

DC test fixture:

I have attached a schematic of a dc test fixture for manufacturing. It looks like this fixture only tests for Vhump. There may have been another test fixture for the other dc tests. Those tests may have been done on a system. The Vhump test may have been so difficult to perform on an automated system that it needed this extra fixture.

Vhump is the amount of maximum dc imbalance, probably referred to the input, which occurs when the variable gain function is swept from full invert to full normal. I think that the two ends of the curve are set to the same dc output level. Ideally, there is no change in dc level under these conditions, but various imbalances in the M84 output devices result in a hump for inverted gain and a hump for the normal gain. Only the -00, -04 and -10 suffixed parts were required to pass this stringent Vhump test. The fixture was in a rubber footed metal chassis about 4” high 5” deep, and 7” wide.

The test criteria defining the various grades of M84 are as follows:
Displayed Risetime <355ps -00 and -10 -01 and -11 -02 and -12 -03 and -13
Displayed Risetime 355-410ps -04 -05 -06 -07
Normal Offset (RTO) <14mV <14mV <28mV <14mV
Invert Offset (RTO) <14mV <14mV <28mV <14mV
HFE of Q1 and Q2 70-250 70-250 50-400 70-250
Normal Gain 2.68-2.96 2.68-2.96 2.68-2.96 2.68-2.96
Invert Gain 2.68-2.96 2.68-2.96 2.68-2.96 2.68-2.96
Null Offset (Gain = 0) 10mV no spec no spec <30mV
Null Gain (Test of Q3 and Q4) 14mV no spec no spec <14mV
50% Gain 0.49-0.51 no spec no spec 0.49-0.51
Off Feedthrough <0.2mV <0.2mV <0.2mV <0.2mV
Normal Hump Voltage <2,5mV no spec no spec no spec
Invert Hump Voltage <2,5mV no spec no spec no spec

Pictures

Used in

Notes

  • Note 1: See John Addis quote in 7A11 article

Datasheet