Hi All
Having spent most of my time in the analog world <1Mhz I recently find myself having to work in the digital world of ic comms interfaces etc. I have all the usual kit for analog work crt scope, dvm, sig gen, 6.5dig bench multimeter , trusty AVO model 8 etc. There's an array of confusing claims out there, and I am a real engineer btw and the hype blinds me. What would you guys recommend, Dig scope - what spec, logic analyser, function gen, frequency counter - whats missing
I am messing with Arduino et al micros and looking at comms, ethernet 232 485 GPIO spi, I2C, UART and particular interest in interfaces with real world devices, encoders, switches etc as I have realised how unreliable this can be and tends to be glossed over.

Any advice from old hands/links most appreciated - I find the WEB just so full of guano hype, infomercials
apologies if Im in the wrong place pls redirect
Robin

 

Sounds like you have all you need,
May be a scope with memory, a digital one, may be with 4 traces,
I2c is up to say 1 Mhz, spi typicaly the 10 of Mhz, so a 100 or 200 Mhz scope would ge good, but not needed,
I'd also suggest a magnifying glass, or microscope, so e good light and things like good tweezers, these digital things are dammed small.

 

What you have will do find. The next thing you will need is a good analytical mind.

 

Thnx 10^6, I was surprised to discover that coming into the digital world needs a different mindset, eg I ran slap bang into the debouncing contact black hole which has challenged designers for a decade. never bother me in my 4-20mA world - except years ago when I was lumbered with relay logic cabinets used for start/shutdown stuff and multi-level sump pumps - very nasty never again. Still this is OTT.
I was looking at something like the Rigol DS1054z 4ch (£354) which apparently can be hacked to do 100MHz. I looked at the Sigil but its after sales service seems iffy. Its a leaf of faith but if I want to see bounce pix I cant do that on an old crt scope, so it looks like I must pay up or remain in darkness as to the interfaces with the physical world. It seems to me that many claims are made by script kids as to the merits of their s/w work arounds - till I then follow through with user posts which show less than sufficient reliability. Analog Devices offers an i/f chip
MAX6816, MAX6817,
but its expensive $3.50 for 1 contact, also Logiswitch LS18, also $$$ - so for me, more work needed

 

You don’t need a 100MHz digital oscilloscope in order to view and understand switch bounce.

And the remedy to overcome switch bounce is well known and well practiced. It is just a real world phenomenon that seems to baffle new comers who have not encountered this before.

 

You don’t need a 100MHz digital oscilloscope in order to view and understand switch bounce.

And the remedy to overcome switch bounce is well known and well practiced. It is just a real world phenomenon that seems to baffle new comers who have not encountered this before.

Which particular remedy do you favour, RC filter, SR gate eg MC14043 - we are just now talking about mechanical switches rather than non contact types - that probably needs another thread. Of course a solution depends on the environment, noise, vibration, shock, on/off cycle time. I am looking at absolute rotary encoder apps <100rpm 8 bit resolution (but would start with 4bit. Ive looked at popular quad ABZ types and been showered with schwepps from so many illiterate wannabes and rubbish infomercials - wastes so much time on false paths.
Test rig where an encoder is motor rotated say +/- 400deg (mechanical end stop) for 24 hours with the counts being saved in register, then you see how much deviation from an average zero count has been made. Thats my goal. Then move to environment tests

 

I was surprised to discover that coming into the digital world needs a different mindset, eg I ran slap bang into the debouncing contact black hole which has challenged designers for a decade.

It's not much of a challenge if you're aware of it.


If you weren't already pretty much set for equipment, I'd've suggested a Tektronix 7704A mainframe, a couple 7A26 vertical amplifiers, an appropriate timebase (I use 7B85A which limits BW to 100MHz). The mainframe accepts a 16 channel 7D01 logic analyzer plug-in with a couple formatter options, 7D20 70Mhz DSO, 7D15 timer/counter, 7D13 DMM, 7CT1N curve tracer (wish I had one of these, I use a BK501A), etc.
https://w140.com/tekwiki/index.php/7704A

 

I do not favour any specific solution since the solution will depend on the application and implementation details.

Here are examples of solutions available.

1) Software
2) RC filters
3) RC filters with Schmitt trigger gates
4) Set-Reset flip-flops

The HW solution with mechanical switches will depend on the available switch contacts, SPST vs SPDT contacts.

Rotary encoders is in a different league altogether and will require separate discussion.

Edit: Note that not all switches need to be denounced.

 

Thanx Dennis,

I'd've suggested a Tektronix 7704A mainframe, a couple 7A26 vertical amplifiers, an appropriate timebase (I use 7B85A which limits BW to 100MHz). The mainframe accepts a 16 channel 7D01 logic analyzer plug-in with a couple formatter options, 7D20 70Mhz DSO, 7D15 timer/counter, 7D13 DMM, 7CT1N curve tracer (wish I had one of these, I use a BK501A), etc.

Good one Dennis, wonder how much your kit would be in todays money £10k? I had a 500 series scope with trolley I bought of our Lab when our Chemical plant shut down £10 I think many years ago. Lots of tubes , kept my shed warm whilst I dabbled
.

 

Sounds like you have all you need,
May be a scope with memory, a digital one, may be with 4 traces,
I2c is up to say 1 Mhz, spi typicaly the 10 of Mhz, so a 100 or 200 Mhz scope would ge good, but not needed,
I'd also suggest a magnifying glass, or microscope, so e good light and things like good tweezers, these digital things are dammed small.

Yes looking into Rigol DS1054 4ch. Got a Jewellers loup 30x very useful, and also FWIW a used Bresser USB scope 102012 20x 80x 350 x (20x most useful) it takes pix £10 EB. Looks plastic mickey mouse but got me out of trouble many times with Mob Phones and RE jobs.

 

wonder how much your kit would be in todays money £10k?

I bought most of my equipment 10-15 years ago when prices were much more reasonable than it has been recently. But you could probably get the items for less than a third of that amount.

This listing has a 7D01 with formatter and 7B53A for $300 (shipping would be higher than $80 for you):


It doesn't come with the P6541 probes for the logic analyzer which would probably run another $400. For the life of me, I don't know where all of the probes went...

Unit tested for power on only. They couldn't test without a vertical amplifier.

Here's a 7D20 that fails test 6542 (one of the attenuator relays). Sometimes removing and re-installing fixes.



7D15 (only tested for power on):


This equipment is 40-50 years old, but a 7704A is still the scope I use most. The manuals are readily available and include calibration/repair sections and full schematics. The only problem is the custom semiconductors they used. I bought several mainframes/plug-ins to use for parts.

Tektronix is local to me so there's a lot of equipment in the area. I bought a 7704A with 2 vertical amps and 2 time bases on eBay from a local recycler for $16.50 in 2010. No one wanted to pay $100 for shipping. I picked it up.

 

I bought most of my equipment 10-15 years ago when prices were much more reasonable than it has been recently. But you could probably get the items for less than a third of that amount.

This listing has a 7D01 with formatter and 7B53A for $300 (shipping would be higher than $80 for you):

View attachment 303765
It doesn't come with the P6541 probes for the logic analyzer which would probably run another $400. For the life of me, I don't know where all of the probes went...

Unit tested for power on only. They couldn't test without a vertical amplifier.

Here's a 7D20 that fails test 6542 (one of the attenuator relays). Sometimes removing and re-installing fixes.
View attachment 303768
View attachment 303771

7D15 (only tested for power on):
View attachment 303769

This equipment is 40-50 years old, but a 7704A is still the scope I use most. The manuals are readily available and include calibration/repair sections and full schematics. The only problem is the custom semiconductors they used. I bought several mainframes/plug-ins to use for parts.

Tektronix is local to me so there's a lot of equipment in the area. I bought a 7704A with 2 vertical amps and 2 time bases on eBay from a local recycler for $16.50 in 2010. No one wanted to pay $100 for shipping. I picked it up.

Yup Tek was always the Rolls Royce of Lab equipment in its time with a huge price to match. I think my 500 series was ca a years wage - something you could only dream about back then. The Rigol Im looking at is ca £350, so its doable, as I cant see me trying to catch contact 5 ms bounce on a crt scope (no persistence screen) with my mobile making a vid - yuk.

 

Tek was always the Rolls Royce of Lab equipment in its time with a huge price to match.

The lab at my community college was full of Tektronix equipment because one of the instructors worked summers there. A fair number of graduates went to work there. I wanted to, but there was a recession when I graduated and they weren't hiring.

When I worked at HP, we couldn't buy Tektronix equipment. But our corporate R&D labs had just about every piece of test equipment HP made for our use.