HP vs Tektronics design philosophies....

[tkamiya]

I've been collecting (and sometimes making useful measurements  ) with mainly HP/Agilent/? and Tektronics test equipment.  It just occur to me lately that HP equipment tends to be based on card/module construction and easy to get to just about everywhere; however, Tektronics stuff tends to be mechanical engineers dream come true.  Its scopes have this 3D constructions and I often can't get to something without going through 3 layers of other stuff.  Just yesterday, I had an obvious case of power supply failure.  To get to it, I have to remove the main board, and to do that I had to remove various ribbon cables and odd card edge PCB.  Then there were another board, then aluminum shield.  Then finally get to the power supply board.  I can't really test anything because load has been removed.  Just a nightmare to service.

I'm mainly dealing with 80s and 90s equipment.  Is there some kind of design philosophies anyone might be aware of? 

[0culus]

Tek was largely focused on one thing...oscilloscopes and instruments for aligning them...for most of their 'golden years'. They eventually expanded, but analog CROs were their bread and butter.  And they mainly designed instruments for use on the bench (or on a rolling cart). HP, in contrast, ended up making literally everything under the sun and a large fraction of it was optimized for rackmounting first and bench use second. The rackmount form factor is considerably more restrictive in terms of internal layout while maintaining serviceability. Whereas not being constrained by rackmounting allowed Tek to be more adventurous with form factors. That's my theory anyway.

[james_s]

I suspect a lot of it comes from the Tektronix focus on building the best of the best analog scopes. Many of them were designed from the ground up to be the highest performance scopes on the market, and high speed analog design doesn't lend itself well to modular construction with plug in cards. The earlier vacuum tube based scopes were mostly hand built and were works of art.

HP built a lot of excellent gear but up into the 80s at least their scopes couldn't quite match many of the Tek offerings. A lot of the non-scope gear Tek sold was not nearly as impressive, much of it not bad, but it wasn't exquisite like the scopes.

[tkamiya]

My first scope was a rack mount tektronics.  RM15.  I still miss it.  While it was 15MHz scope, I used it at 2 meters band (144MHz) to align a transmitter.  It was awfully off on voltage but all I needed to see was a relative strength.  It worked well.  Inside was warm glow of BIG tubes!

You maybe right on analog scopes.  It would be awfully hard to do a card cage construction, although modular build may very well be possible.  THAT, however, is not suited to making it compact.

Just from hobbyist repair stand point, I find tek scopes are pain.

[james_s]

My first scope was a huge old Tek 531A, I actually still had it up until a couple years ago when I gave it to another member here. I loved that thing but it was just way too big and I hadn't used it in years. It was a masterpiece of engineering though, absolutely beautiful construction.

Tek scopes were never hobbyist instruments, they were top of the line and cost a fortune, almost all of them cost as much as a new car when they came out, sometimes a very nice new car. They were optimized for performance above all else, including serviceability. A high end state of the art sports car is usually gonna be a pain to work on too.

[0culus]

My 585A isn't hard to work on, until you have to move it to lay on the other side.  Tek in this era was unparalleled IMO. Everything about it oozes quality, and the documentation in the manuals is top drawer.

[Cubdriver]

Back in the day, there were three names in Oscilloscopes - Tektronix, Tektronix and Tektronix.  For everything else, there was HP.  Sean may well be on to something about the differences due to form factor - the Tek chassis gave them a bit more freedom design-wise whereas sticking everything into a 19" rack prompted HP to make other choices.  HP definitely leaned more towards removable cards in many things.

This should be an interesting thread - thanks for starting it!

-Pat

[tggzzz]

I've been collecting (and sometimes making useful measurements  ) with mainly HP/Agilent/? and Tektronics test equipment.  It just occur to me lately that HP equipment tends to be based on card/module construction and easy to get to just about everywhere; however, Tektronics stuff tends to be mechanical engineers dream come true.  Its scopes have this 3D constructions and I often can't get to something without going through 3 layers of other stuff.  Just yesterday, I had an obvious case of power supply failure.  To get to it, I have to remove the main board, and to do that I had to remove various ribbon cables and odd card edge PCB.  Then there were another board, then aluminum shield.  Then finally get to the power supply board.  I can't really test anything because load has been removed.  Just a nightmare to service.

I'm mainly dealing with 80s and 90s equipment.  Is there some kind of design philosophies anyone might be aware of?

There are contra-examples (click to expand the pictures)...

Have a look at some contemporary spectrum analysers:

• Tektronix 492 which have easily extracted PCBs, except for the PSU in the rear end cap assembly which is amazingly difficult to access
• HP8562B's heroic internal assembly, with the PSU buried under the other PCBs in the middle of the chassis

The Tek492 is mechanically surprisingly simple internally...



For the HP8562B, it is surprisingly difficult to capture just how well all the circuit boards are packed into a confined space. This is one time where a stereoscopic picture would really be helpful, or a video showing the gentle gradual unfolding. But in the absence of those...

Here's a service manual diagram illustrating how they can fold out.



Naturally there is no way these could be held in that position, nor would you want to, since you would need access to both sides of the boards while testing them. So there is a wonderful four segment hinge that allows each board to be folded out on its own, and/or rested against either neighbouring board in either position.

Here it is with boards A2,A3, A4, A5 all folded out, allowing access to the PSU A6 buried under A5 in the middle of the instrument.



And here's the hinge with boards A5 (hidden) and A4 still unfurled, and boards A3 and A2 folded out. That gives access to the top of A4 (boring since it is fully shielded) and the bottom of A3. EDIT: swap A2 and A3, doh!

[wn1fju]

Here is another difference between Tektronix and HP that I have noticed over the years...  I am referring to circuit boards in the 1970 - 1980 time frame.  HP seems to construct boards with all of the traces running vertically and/or horizontally.  The components end up where they end up and can be diagonally placed.  Tektronix, on the other hand, mounts all the components vertically and/or horizontally and then runs curved traces to get where they need to be.  Very curious indeed.

I know this is a generalization, but I have seen it a lot.

[vk6zgo]

When they got away from Oscilloscopes, Tektronix indeed didn't always do well.

Their "650" series picture monitors were beautifully crafted, but had numerous problems------ the file of modifications sheets was like a phone book.
Try as you might, it seemed impossible to get the promised performance out of them.

Sony, as a "next generation" replacement a few years later, made the BVM1301, which was identical in outside dimensions.
The Sony interior wasn't near as pretty as the 650 series, but they met specs "out of the box", & still did, 20 years down the track, unlike the Teks, which were often sitting unused in the test equipment store with various problems 10 years or less after purchase.

If they did fail, they were easy to fix, &  didn't require any "unobtainium" parts.

I think Tek suffered a bit from hubris, didn't really research their market properly, & thought "Everybody would love it, because it came from Tektronix".

People did buy them, & "really tried to love them", but after becoming used to 'scopes that "just kept on working", it was a bit of a shock to get something with continuing problems from the same company.

[AndersJ]

Possibly a bit off topic here,
but I’ve always found processor based Tektronix instruments a bit sluggish and unresponsive.
I often get the feeling that the used microprocessors and/or software
should be an order of magnitude faster and more alert.

I have no hands on experience with current Tektronix instruments.

[tkamiya]

I think it's a result of attempting to do what they did with analog scopes with new digital.  My older digital scopes are all like that.  Microprocessors were not at today's standard back then. 

It's amazing they did spectrum analyzers with chips that was originally intended for calculators.  In B series, they did upgrade it to Motorola 68K processors.  I believe "retrofit kit" was like $6000US? 

[0culus]

When they got away from Oscilloscopes, Tektronix indeed didn't always do well.

Their "650" series picture monitors were beautifully crafted, but had numerous problems------ the file of modifications sheets was like a phone book.
Try as you might, it seemed impossible to get the promised performance out of them.

Sony, as a "next generation" replacement a few years later, made the BVM1301, which was identical in outside dimensions.
The Sony interior wasn't near as pretty as the 650 series, but they met specs "out of the box", & still did, 20 years down the track, unlike the Teks, which were often sitting unused in the test equipment store with various problems 10 years or less after purchase.

If they did fail, they were easy to fix, &  didn't require any "unobtainium" parts.

I think Tek suffered a bit from hubris, didn't really research their market properly, & thought "Everybody would love it, because it came from Tektronix".

People did buy them, & "really tried to love them", but after becoming used to 'scopes that "just kept on working", it was a bit of a shock to get something with continuing problems from the same company.

In later years, the company leadership changed from engineers to businesspeople/bean counters. They coasted downhill on decades of goodwill for a long time. The stuff they offer today is overpriced garbage IMO. Would never even consider today's Tek for a work purchase.

[David Hess]

I'm mainly dealing with 80s and 90s equipment.  Is there some kind of design philosophies anyone might be aware of?

Keep in mind that Tektronix had separate design teams for different series of instruments of the same type.  For instance the 4 channel 22xx series was designed by a different team than the 4 channel 24xx series.  The 5000 and 7000 series of oscilloscope mainframes used independent design teams.

They did make mechanical improvements with new models of the same series; the 2232 allows the storage board to be rotated out of the chassis giving much better access while the 2230 it is based on is horrible to work on.

But one thing I have noticed is that Tektronix hardly ever redesigned an assembly for commonality between instruments.  So for instance something like a vertical amplifier might be improved for each new oscilloscope in the same series, but the new ones would never be retrofitted to the older models even though it would seem to simplify production.  Once an assembly was working, they left it alone if possible.

A rare exception was the readout board used in the 7000 series which had to be redesigned when some of the original custom ICs became unavailable.  The power supplies and some other circuits of the 2 channel 22xx series were upgraded continuously as new models came out but the improvements were never applied to older existing models which were still in production.  But another exception might be the 2235, 2236, and 2232 which became the 2235A, 2236A, and late model 2232 with more commonality at the end of availability of that series.

[anotherlin]

I cannot comment on the design philosophy of the circuits, yet they've probably use some common concepts and ideas. The laws of physics are the same in the universe, after all !
However, for the physical implementation, or rather the organization, that is not so much how the circuits are layout on the PCBs, but the way PCBs and various electro-mechanicals (potentiometers, switches, etc) are connected .
Seems that HP is a bit more organized, and have some equivalent of a company wide "mechanical coding style" in place.
Of the HP instruments I have, all have neatly aligned PCBs connected with edge connectors.
Whereas, the 7603 scope I have, features cables dangling everywhere. Same goes for the plug-ins.
There may be an explanation for that: Tektronix makes heavy use of "plug-ins", that is standardized boxes with standardized connections.
So, any design team can do whatever it wants inside these plug-ins, as long as they meet specification outside.