Topics

7A12 attenuator board clone


Martin Hodge
 

Last year I was lucky enough to find a Tek 7704A scope on CL for free! The scope came with 7A12 and 7B523A plugins. Wanting a 4 channel scope I found another 7A12 on eBay but it turned out to have a broken attenuator board on channel 1. These boards are made of some kind of transparent, rock-candy-like material and are extremely brittle. This one had cracked and separated several traces. I jumped the traces with thin wire and got the board working again. The traces were to two of the relay magnets, no signals running through them. But in the process of unplugging and re plugging the board, one of the connector mezzanines crumbled into several pieces. Something came over me at this point and I decided to waste a lot of time building a new board using modern FR4 process. The photo album linked below shows the process. I was going to make another for channel 2 out of Robertson material and run a comparison with FR4, but before I even priced a board at pcbway I realized I didn't have a signal source with a rise time fast enough to test with. In the end I grew to dislike the 7A12 for it's overly complicated internals, sticky buttons, and meek performance and got rid of them. Though I did like the front panel layout.

https://groups.io/g/TekScopes/album?id=256606
(Be sure to read the photo descriptions for more info)


 

Hi Martin,
I admire your passion. Now that I am retired I have all the time in the world to pursue quirky things that catch my attention regardless of how practical they may be.

I hope you succeed in improving your 7A12. That plugin brings back fond memories for me because it was the very first one I ever used when I was given a loaner 7603 for a weekend.
I have many 7A12s sitting in a box somewhere so I will be happy to send you a few to use for spare parts.

By now (50 years later) I know all those pushbuttons and relays are a nightmare to maintain. And, although I never thought about it at the time, this plugin was a nightmare to build. The side by side arrangement of the two channels meant that each channel's PC Board was a mirror copy of the other one. This made it annoying, to say the least, to produce this plugin.

There were many within Tek who felt the side by side layout of the 7A12 front panel made perfect sense. When Tom Rousseau proposed a much more practical layout for Tek dual trace plugins he was met with a lot of resistance. But there were so many advantages to Tom's proposal that it eventually won out. In the end Tom's 7A26, 7A24, and 7A18A designs were hugely successful.

Dennis Tillman W7pF

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Martin Hodge
Sent: Friday, November 13, 2020 7:41 AM
To: TekScopes@groups.io
Subject: [TekScopes] 7A12 attenuator board clone

Last year I was lucky enough to find a Tek 7704A scope on CL for free! The scope came with 7A12 and 7B523A plugins. Wanting a 4 channel scope I found another 7A12 on eBay but it turned out to have a broken attenuator board on channel 1. These boards are made of some kind of transparent, rock-candy-like material and are extremely brittle. This one had cracked and separated several traces. I jumped the traces with thin wire and got the board working again. The traces were to two of the relay magnets, no signals running through them. But in the process of unplugging and re plugging the board, one of the connector mezzanines crumbled into several pieces. Something came over me at this point and I decided to waste a lot of time building a new board using modern FR4 process. The photo album linked below shows the process. I was going to make another for channel 2 out of Robertson material and run a comparison with FR4, but before I even priced a board at pcbway I realized I didn't have a signal source with a rise time fast enough to test with. In the end I grew to dislike the 7A12 for it's overly complicated internals, sticky buttons, and meek performance and got rid of them. Though I did like the front panel layout.

https://groups.io/g/TekScopes/album?id=256606
(Be sure to read the photo descriptions for more info)







--
Dennis Tillman W7pF
TekScopes Moderator


Ed Breya
 

Don't bother cloning, unless you can get the right board material. I believe Tek used polysulfone for all high speed, high-Z attenuator boards. You'll also see it used in microwave stripline components, and probably HP scopes, for good RF performance. It's also good for low leakage - I have a polysulfone (I think) board saved from a junked HP3490A DMM. It's kind of a dark orange/translucent color, used to mount all the high-Z input and auto-zero circuitry directly, without using PTFE standoffs. It's also used for high-Z insulators and switches. After you see it often enough, you kind of recognize the color range, texture, and opacity, and automatically identify it. It's quite fragile, and expensive, so only used where necessary.

BTW I think the reason it's mechanically weak for board use, is that it's in fairly pure form - no reinforcing fibers or fillers allowed, which would spoil the dielectric characteristics that make it a choice in the first place. It's probably comparable to other plastics in mechanical characteristics. I have some 1/4" plates of it, that came from some big old directional couplers. One thing that seems very different is if you clink them together, they sound "glassy" rather than "plasticy," so they are somewhat hard and brittle.

Coincidentally, I just mentioned it in a recent thread here:
https://groups.io/g/TekScopes/message/172980

Ed


Ed Breya
 

I should mention that for lower cost, Tek used an alternative material for the 2200 series attenuator board assemblies. As I recall, it's a special type made by Oak (Industries?), that works well enough for 100 MHz-class high-Z.circuits, and is mechanically strong enough for "regular" type board applications. There may be other modern types that would do the job.

Ed


 

On Fri, Nov 13, 2020 at 06:44 PM, Ed Breya wrote:


I believe Tek used polysulfone for all high speed, high-Z attenuator boards.
ISTR that a few years ago, David Hess explained in this group that one reason to use the material in this kind of product was to avoid an effect called "hook".

Raymond


Ed Breya
 

Yes indeed, Raymond - hook was the main thing, and quite obvious. I don't recall if it was due to a non-linear capacitance effect, or a form of dielectric absorption.

Ed


Christian
 

Regarding the special attenuator substrate:

I don't believe that polyethersulfone sees very much use these days. I do
a certain amount of microwave PCB design and I can't say that I've ever
even seen it discussed in the modern literature, or listed on a fab
shop's capabilities list. Of course, with enough money some shops will
build you whatever sort of science project you want (not to say that PES
is/was a science project).

Hook is an effect caused by a change in Er across frequency. Fourier then
says that your microstip (etc) impedance will vary at different locations
in a square wave, and you can end up with some goofy-looking pulses. And
swept signals will have varying amplitudes/losses/etc.

Here in the 2000's, where we're quite enamoured of extreme
digital signalling rates, we have a lot of substrates that are designed
specifically to avoid this problem. As far as I can tell, pretty much
anything better than 'good old FR4' (or its RoHS cousic 370HR) will be fine
for any 'exotic' 7000 time domain plugin use.

I could be wrong, this is not legal advice, YMMV, etc.


Ed Breya
 

Christian, these special materials were needed ONLY for high-Z, wide band, low leakage applications, like the 1megohm front end of a scope. There are lots of options for low-Z environments.

Ed


Clark Foley
 

Ahhhh. The 7A12. I remember an unofficial accessory that was distributed partly for fun but it proved very useful. It was a pencil with erasers at both ends. After using the 7A12 for 10minutes, the joke pencil came to the rescue!


Martin Hodge
 

According to the manual, the 7A12 attenuator board is made of Polyphenylene Oxide. If one were to try to use a 7A12 with an FR4 attenuator, how would it's inferiority manifest on the screen? And what signal would one need to feed it to make it "fail"?

In case I didn't make it clear, this was a purely educational endeavor for me. I was curious to see what the difference would be, if any.


Ed Breya
 

Martin, that's pretty cool that they stated what it's made of. So, looks like PPO, not polysulfone, and probably the same goes for the other 7Ks at least. I don't know about the other model lines.

Regarding the difference between specialty materials and FR4 and such, I suppose you'd see hook, and probably more roll-off at higher frequencies, making it harder (or impossible) to tweak the attenuator compensations. And maybe more board surface leakage (DC offset or gain error) in high humidity environments. Again, I have only vague recollections of the issues, but I think the hook showed mostly at low to medium frequencies, not necessarily the higher end.

Ed


Tom Lee
 

Yes, aside from being brittle, PPO’s melting point is not that much higher than butter’s. I’m exaggerating, of course, but you do have to exercise care in soldering. If you apply any forces — lateral or vertical — during soldering, things will move around.,

Btw, in FR4, much of the hook is due to moisture wicked up by the glass fibers. I’ve seen attempts to mitigate this by using various sealants, but FR4 is just not the right material for critical instrumentation unless a decent calibration engine is available.

Cheers
Tom



Sent from my iThing, so please forgive typos and brevity.

On Nov 13, 2020, at 1:44 PM, Ed Breya via groups.io <edbreya=yahoo.com@groups.io> wrote:

Martin, that's pretty cool that they stated what it's made of. So, looks like PPO, not polysulfone, and probably the same goes for the other 7Ks at least. I don't know about the other model lines.

Regarding the difference between specialty materials and FR4 and such, I suppose you'd see hook, and probably more roll-off at higher frequencies, making it harder (or impossible) to tweak the attenuator compensations. And maybe more board surface leakage (DC offset or gain error) in high humidity environments. Again, I have only vague recollections of the issues, but I think the hook showed mostly at low to medium frequencies, not necessarily the higher end.

Ed





Göran Krusell
 

Hi, I recall this topic quite well. Some years ago I uploaded an article from magazine EDN, I believe, which described the difficulties designing high impedance attenuators using an FR4 circuit board. The article may be found under label “hook” or my name. Interesting reading.
Göran


Christian
 

And today I've learned something new about PCB design science. I hadn't though about the effects of varying Er on low-speed, low-capacitance, hi-Z circuitry. It really varies that much over that small a frequency span? I guess that octaves are octaves no matter where you find them.

Sorry for the misleading comments that I made earlier. That's what I get for thinking that my microwave experience taught me about -all- the offbeat dielectric phenomena. Now I'm wondering just what modern substrate -does- deal with this problem...


stevenhorii
 

I don't work much at high frequencies, but from my surplus shopping days I
recall some RF circuit boards that were made of glass fiber and Teflon with
copper for the traces. I also saw some circuit boards that a scrap guy was
grinding up. They were from some military project and were ceramic
multilayer boards. No chips or discretes on these; I think they were
defective or from a contract termination (they might also have been scrap
warranty items and had to be destroyed). I asked him why he was scrapping
them and he said the traces were printed with a gold alloy. I would expect
using either substrate for a circuit board would be expensive.

On Fri, Nov 13, 2020 at 5:05 PM Christian via groups.io <cweagle=
bostondynamics.com@groups.io> wrote:

And today I've learned something new about PCB design science. I hadn't
though about the effects of varying Er on low-speed, low-capacitance, hi-Z
circuitry. It really varies that much over that small a frequency span? I
guess that octaves are octaves no matter where you find them.

Sorry for the misleading comments that I made earlier. That's what I get
for thinking that my microwave experience taught me about -all- the offbeat
dielectric phenomena. Now I'm wondering just what modern substrate -does-
deal with this problem...






Martin Hodge
 

In a previous message I had mentioned "Robertson" material when the actual name is Rogers. Specifically PCBway lists "Rogers 4003C" and "Rogers 4350B" as substitutions for FR4 in "Microwave and RF designs".
Its "hydrocarbon ceramic laminate" that is "compatible with FR4 fabrication process".

https://rogerscorp.com/Advanced%20Connectivity%20Solutions/RO4000%20Series%20Laminates

Thank you Dennis, Ed, Tom, and Göran for the article.


SCMenasian
 

Rogers Corporation makes a line of circuit board material specifically for high performance/high frequency applications. I use one of their Duriod materials and have found it to be far superior to FR4. They have a web site which gives selection charts. You can sometimes find various Duroid types on eBay.


teamlarryohio
 

Ed, ISTR the attenuator boards in a 465 (670-2434-01 I think) started out as PPO then there was a wiz article and a switch to polysulfone. As mentioned, melting point was part of the culprit. The 2 heavy wires at the rear for the gain switching contacts required enough heat to hurt the board if you needed to take it out.
-ls-


Tom Lee
 

Thanks very much for that information — I did not know about the switch to polysulfone.

Good to know!

—Cheers,
Tom

Sent from my iThing, so please forgive typos and brevity.

On Nov 13, 2020, at 3:20 PM, teamlarryohio <larrys@teamlarry.com> wrote:

Ed, ISTR the attenuator boards in a 465 (670-2434-01 I think) started out as PPO then there was a wiz article and a switch to polysulfone. As mentioned, melting point was part of the culprit. The 2 heavy wires at the rear for the gain switching contacts required enough heat to hurt the board if you needed to take it out.
-ls-





Tom Lee
 

The "hook" that people have mentioned a couple of times is the nickname for the step response artifact caused by a dispersive dielectric. If you adjust the compensation so that the rising edge peaks at the same level as the steady-state value, you find that the response sags in between. There are ways to compensate for that effect to one degree or another, but it's not trivial to do it well.

-- Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 11/13/2020 13:06, Martin Hodge wrote:
According to the manual, the 7A12 attenuator board is made of Polyphenylene Oxide. If one were to try to use a 7A12 with an FR4 attenuator, how would it's inferiority manifest on the screen? And what signal would one need to feed it to make it "fail"?

In case I didn't make it clear, this was a purely educational endeavor for me. I was curious to see what the difference would be, if any.