[Adrian] had a Commodore computer to fix and decided to see how his latest tiny portable scope would work. He paid $57 for the tiny little test instrument although the current price seems higher. It claims to have 120MHz bandwidth along with 500 megasamples per second. There are several versions with different claimed specs, but we did find a similar device for under $60. You can see the unboxing and how it worked in the video below.
Of course, these kinds of instruments often overstate their specs, and [Adrian] was also suspicious. One odd feature of the device is it can echo its output to an NTSC video output so you can send the screen to an external monitor.
If you want to skip the scope unboxing, forward up to about 19 minutes to see the inside of the Commodore 64. The scope was easily sufficient for scanning the chips in the computer and revealed a suspicious address line. The line went to a PLA and a mux chip, neither of which were in sockets. He clipped the PLA out of the circuit, and the address line started looking normal. So the conclusion was the PLA was dead.
After that, it was straightforward to remove the chip and replace it. Well, technically, replace it with a socket to make a future repair easier. Will a $57 scope replace your big benchtop instrument? Maybe not. But it was a useful tool for troubleshooting.
Even if you don’t want a cheap scope, you can learn a lot from [Adrian’s] thoughtful troubleshooting and analysis if you are faced with any digital repair project. We do like cheap scopes around here. It is amazing how much scope $100 will buy now compared to just a few years ago.
My only complaint is sometimes you need a second probe when you’re trying to compare timing between two signals. More than two isn’t necessary – just convenient.
Logic Analyzer: Am I a joke to you?
(btw. I agree with you and I haven’t found a case where I really needed a logic analyzer and so far I fixed hundreds of arcade boards without one but I do kinda wish for a Fluke 9000something sometimes)
I have a fondness for logic analyzers, but now that most of the wide busses are internal to a chip they are less useful. However, the computer in the video would be very amenable to like a Gould Biomation K100D or any of the Tek or HP or similar LAs out there.
Apparently the same company sells a dual channel version (with slightly larger screen) for $81 USD. ZEEWEII DSO2512G . I’m curious if anyone has tried it…
https://www.aliexpress.us/item/3256802899439203.html
There exist a ~$80 2 channel SigPeak DSO2512G version of above https://www.eevblog.com/forum/testgear/new-2ch-pocket-dsosg-sigpeak-dso2512g/ using both channels cuts sampling rate to 100MHz (or overclocked 120MHz).
YT review https://www.youtube.com/watch?v=yh4s4-BT5-8
Not to be a nag, but you can’t specify sample rate in Hz, but in sps (samples per second).
Because for instance, if you have a 100MHz clock, but sample on both the rising and falling edge of that clock signal, your sample rate would be 200sps. If you sample only on the rising edge of the clock signal, it would be 100sps. But really, the sampling could even be at a higher or lower rate.
So MHz is not the metric to use. ;)
It also causes confusion with the bandwidth of the scope. You can easily have a 200MHz bandwidth scope with only a 100Msps sampling rate. Theoretically you can also have a scope with 20MHz bandwidth and 100Msps. A useless scope that would be, but theoretically… ;)
Just wanted to point this out, as it’s important. Keep bandwidth and sample rate separated
How would a person find a PLA with the same factory configuration? I would assume you would need a master and burn a copy
There is no programming in a PLA. You can put any compatible one in.
A lot of PLAs do carry “burned” configuration, especially ones used on older computers like this for address decoding etc. However they are often simple enough devices that their program is feasibly reverse engineered.
Fortunately the PLA function has been reverse engineered (it’s being a memory mapper) and there are a couple other brand PLAs pin compatible and their respective programming available online.
Being the c64 I would assume you could also just buy pre-programmed ones. The community is huge and amazing.
In the video he put a newly made replacement PLA in the 64 of the several different ones he had on hand, you can’t get original PLAs anymore but replacements, easy.
The PLA is just an address decoder. Back in the olden day, people used EPROMs as a replacement decoder. Worked with most software, but not all.
http://www.geocities.ws/sieg.peter/pla/pla.html
http://www.vic20.de/html/eprom_pla_8296_und_c64.html
http://www.cbmitapages.it/tecnico/pla.htm
Nowadays, we can alao get FPGAs and other programmable logic aa a substitute, of course. But EPROMs aren’t bad, either. Their latency is quite short (good).
Kudos to the author, also. This is one of the few Hackaday.com articles not glorifying the C64, but rather helping fixing real existing problems. I’ve gotten a C64 from a friend years ago who gave it to me just a few weeks before he passed away and it had a dead PLA inside.
Gratefully, the C64 has a fine community which documented all the symptoms of the C64 flaws in both text and screenshot form. Otherwise, I had given up fixing that mess of a home computer..
https://retrocomputerverzamelaar.nl/commodore-64-problems/
I wonder if such a scope is a decent alternative to a differential probe, it being isolated and all that.
Any firmware hacks on these kinds?
It isn’t a replacement for differential probes – at least, not one you’d want to rely on.
The probe jack isn’t insulated. If you connect it to something “hot” you’ll have hazardous voltages exposed on the metal parts of the connector, just waiting to “bite” and kill you.
At low voltage it’d probably be OK.
Even though drinking mentions that this scope has NTSC output if you actually watch the full video you’ll see that Adrian find that the composite video output is actually PAL (with no signs of a way to actually select NTSC output).
Would have been cool if he tested the actual bandwidth of the scope with an external source. When he changed to 2MHz square signal, there was already a huge amplitude drop on the signal, so I wonder what useful bandwidth you could get from this.
YT dude “TechCornerTV Electronics, Projects & Tutorials” tested both this DSO1511G and dual channel version DSO2512G, He also did a teardown.
Lattice LCMXO2, Allwinner F1C100s and one dual channel 100MHz AD9288 ADC. “120MHz” Chinese are boasting about must be them overclocking AD9288 to 120MHz, thus 240MHz sampling rate.
TLDR: around 95MHz of analog BW (pretty impressive), 200-240MHz sampling rate (sucks).