Oscilloscope for beginners
- Thread starter big10p
- Start date
I'm considering this one:
https://www.telonic.co.uk/Rigol-DS1054Z-4ch-50MHz-Digital-Oscilloscope-p/ds1054z.htm
Is 4 channels useful in our hobby though? And would the 50MHz bandwidth be a limiting factor (I can't think of anything that would need that much bandwidth otoh)?
I'm quite new to this but this one has some decent reviews. It's at the top end of what I want to pay for an entry level scope, but seems to have a lot of features.
agent41252020-06-24 13:56:53
https://www.telonic.co.uk/Rigol-DS1054Z-4ch-50MHz-Digital-Oscilloscope-p/ds1054z.htm
Is 4 channels useful in our hobby though? And would the 50MHz bandwidth be a limiting factor (I can't think of anything that would need that much bandwidth otoh)?
I'm quite new to this but this one has some decent reviews. It's at the top end of what I want to pay for an entry level scope, but seems to have a lot of features.
agent41252020-06-24 13:56:53
John Bennett
Senior Member
As I said earlier, I don't know how people manage without a scope and lots of channels. Definitely 4 is a good start. Mine's got 20 and it's often not enough.
My brother uses a full-blown logic analyser on some of his repairs (must be 40+ channels and masses of RAM).
Suppose it depends on the fault. Many things you can see by probing around suspect areas, looking for poorly looking signals.
But 4 channels is a good start, as with many chips, you'd want to have the clock, the chip select (or output enable), an input and an output to see if it's reacting when it should do (the output on ICs is often nonsensical unless they've been chip selected onto the bus).
John Bennett2020-06-24 14:03:53
My brother uses a full-blown logic analyser on some of his repairs (must be 40+ channels and masses of RAM).
Suppose it depends on the fault. Many things you can see by probing around suspect areas, looking for poorly looking signals.
But 4 channels is a good start, as with many chips, you'd want to have the clock, the chip select (or output enable), an input and an output to see if it's reacting when it should do (the output on ICs is often nonsensical unless they've been chip selected onto the bus).
John Bennett2020-06-24 14:03:53
agent4125 said:
as I said earlier I don't go that deep into repairs any more BUT I WOULD SAY AT FIRST GLANCE AND AROUND £300 ITS PROBABLY GOT MOST IF NOT ALL YOU SHOULD NEED AT HOBBY REPAIR LEVEL ?
maybe someone else knows more about the model ( I would be happy to own that piece of kit) and its not only practical but pleasing on the eye ………. IF ITS EASY ENOUGH TO SET AND READ ?
could be a good find?
John Bennett
Senior Member
Yeah, those Rigol's are decent enough.
I think the emergence of scopes like that forced the 'big boys' - Tektronix and Agilent to look at their prices, as you could get something just as useful as their kit for 1/5 of the price.
50MHz is probably enough for most 80's and early 90's stuff, I think. If you were analysing a 48MHz clock signal, it wouldn't work due to the scope limitations (you'd get sinusoidal nonsense), but most signals are clocked down to 10MHz and below on PCBs, which is should be alright with.
You need the bandwidth of the scope to be 2x the max signal speed you want to look at, as a minimum.
12 Million data points capture is great - means you can 'stop' and zoom in/rewind on triggered stuff. It'll decode serial data on the screen too, which is sometimes useful.
John Bennett2020-06-24 14:19:29
I think the emergence of scopes like that forced the 'big boys' - Tektronix and Agilent to look at their prices, as you could get something just as useful as their kit for 1/5 of the price.
50MHz is probably enough for most 80's and early 90's stuff, I think. If you were analysing a 48MHz clock signal, it wouldn't work due to the scope limitations (you'd get sinusoidal nonsense), but most signals are clocked down to 10MHz and below on PCBs, which is should be alright with.
You need the bandwidth of the scope to be 2x the max signal speed you want to look at, as a minimum.
12 Million data points capture is great - means you can 'stop' and zoom in/rewind on triggered stuff. It'll decode serial data on the screen too, which is sometimes useful.
John Bennett2020-06-24 14:19:29
It's looking good then for that Rigol... thanks for your input guys.
One thing I learned is that you have to share the vertical controls between the 4 channels. Is that likely to cause problems? The reviewer mentions this at about 2 mins here in this summary review: https://www.youtube.com/watch?v=ETCOhzU1O5A&feature=youtu.be&t=120
One thing I learned is that you have to share the vertical controls between the 4 channels. Is that likely to cause problems? The reviewer mentions this at about 2 mins here in this summary review: https://www.youtube.com/watch?v=ETCOhzU1O5A&feature=youtu.be&t=120
John Bennett
Senior Member
It will presumably just mean you have to press the channel button, then adjust the amplitude. Would make the scope a bit bigger to have four sets of amplitude dials. Many scopes are like this, it's not a biggie.
I was going to post that the DS1074 plus looks worth the extra £50-ish as you get 70MHz and a 16-channel logic analyser. Except the logic analyser cable is an extra and another £226 on-top! . So it soon spirals out-of-control.
I was going to post that the DS1074 plus looks worth the extra £50-ish as you get 70MHz and a 16-channel logic analyser. Except the logic analyser cable is an extra and another £226 on-top! . So it soon spirals out-of-control.
Hahah yes I thnk I'll go for the DS1054Z then - it soon gets expensive.
I can't believe the price of some of their probes!!!
https://www.telonic.co.uk/Rigol-RP1004C-Current-Probe-p/rp1004c.htm
I can't believe the price of some of their probes!!!
https://www.telonic.co.uk/Rigol-RP1004C-Current-Probe-p/rp1004c.htm
John Bennett
Senior Member
Yes, they might do some cheap lower-spec scopes, but that's definitely an industrial-standard(ish) price for a super-high bandwidth current probe (you'd pay that sort-of price for one from Tektronix or Aglient, if not a little bit more).
At least it's reassuring they aim to compete with higher-spec stuff, rather than just being a 'cheap scope' company.
At least it's reassuring they aim to compete with higher-spec stuff, rather than just being a 'cheap scope' company.
Hi John I was reading up on Oscopes (also as a beginner) trying to get my head around the scopeworld and I see the value in your above comment about probing the o/e ,clock, input and output of a chip, so a 4 channel would be the smart choice. I have had a little success with my HP probe an eprom burner testing ttl's burning pals and gals but have come across quite a stubborn board in Chase HQ so thinking its time to delve into a scope. As I'm new to the caper I was hoping you could aid in clarifying what would be the process in using a scope to test signal integrity namely the horizontal sweep speed settings (I understand the volts per division). Lets say I'm probing a LS161 counter a 2048 ram ic and a mask rom, using chase hq as my test bunny, what is the way you would tackle the above using a I feel a 100mhz scope is what im going to go .TBH I would assume most if not all ttl related stuff would have the same type test settings that you would advise. Have done a write up on using a scope by chance mate- if not any suggestions on a good read?
Any help with the above is super appreciated!
Cheers
Y
Last edited:
John Bennett
Senior Member
Hi,
I should start by saying I don't know how you'd be able to do much with an old analogue oscilloscope - they rely on repetitive signals for you to be able to see anything useful. You could get it to trigger on the clock outputs of a 74LS161 and look for square-waves, where triggering on each Qx output, adjusting the horizontal timebase so you could see a full square wave, then counting the divisions on the scope would give you the period. You'd see the period doubling as you went through the Q outputs in sequence. In fairness, a bad chip isn't going to have the wrong frequency on a pin, it'll either be dead or a weak waveform.
But trying to look at something which isn't repetitive on an analogue scope would just show a blur.
If you had a digital oscilloscope, it's a lot easier as they can 'pause' the result for you to look at. So you could simply do a 1-shot trigger on a rising/falling edge of a signal line and see that it's got activity and that there's nice edges and signals that go sufficiently high and low (TTL is often like 0.7V to 4V, or even less range than that).
If it's a bus, where multiple devices can drive the track, then you're into the realm of individually testing every device on the bus. So for each component, you'd want to do something like trigger on the edge of the output enable and put the other probes on inputs and outputs and check the chip is behaving correctly - i.e. for a 74LS244 the outputs should follow the inputs when the device is enabled.. You've got to interpret the datasheet for each chip and know how the inputs and outputs relate.
For RAM and ROM, you can really just check the pins are togging high and low when the output is enabled, you're into logic analyser territory to check the data is valid.
What's up with this particular ChaseHQ? I don't fix many boards, but I've done a few of them in the past.
Hi John
Thank you for taking the time out mate to reply, much appreciated. I had a read of the above as suggested, single shot triggering will definitely give me the info in analyzing the waveform- only thing is I'm not sure what how to interpret it. So clarifying above the clock frequency doesn't change when a chip is going bad I just need to see what the voltages are doing?. TBH as mentioned all I've been doing is probing control, clock lines and looking for inactive or shorted pins across the board.
So ATM I've tracked down a couple of dead counters which were having an affect on the object sprites now I've got them back to on screen to a "normal size" they are all corrupted something I assume is going on with the addressing logic wrong sprites everywhere. Funny story reading up on these boards they have hidden logos of old cigarette images within the masks so I've got a pack of cigarettes cruising around!. The mask roms check out are verified and the control lines are also doing there respective jobs, as do the pals and proms. I've been lucky enough to be given a spare working video board so I've been using this as reference when checking out working non working chips.
Interpreting the data when triggering the controls lines, I/O pins and clock pins and in what horizontal time base, is my next step. Is there a set timebase with ttl stuff that is ideal? I've been given a little 20mhz hantek usb scope just to play about with an understand this isn't going to give me the right signal shot but once I get my head around analyzing the data then I'm ready to get the Rigol as suggested.
Is it the difference in timing with respect to I/O's when the chip is clocked or enabled or just analyzing the signal qualities with whats coming in and whats going out with respect to voltage levels that will tell me what's up. I'm sort of across propagation times and am OK with reading the datasheets now just trying to put it altogether mate. Any further assistance and clarification would be again appreciated and not sure if you may have an image on a scope that may help identify what to look for with a bung chip.
Thanks again mate for reaching back out!
Y
John Bennett
Senior Member
You're really into manually doing what things like SLICE and some logic analysers do - taking a chip, looking at the datasheet and interpreting the state tables in there, then looking to see if the chip behaves in the same manner - so every IC different you look at will require you to think of a way to test it.
I don't think you're going to often see a problem in a propagation delay, it's usually weak or dead outputs.
I think 20MHz is a start, but 100MHz is definitely better as you easily get signals into the 10MHz or above range in 80's stuff, which ends up looking like a sine-wave on low-bandwidth equipment (although to be fair, if you ever see a 50MHz waveform on a 1985 game PCB, it probably DOES look like a sine-wave).
I was lucky with the bad graphics on the Chase HQ board I looked at - it was intermittent, so I could tie-in the disruption on the screen with visible changes on the oscilloscope (as sprites went off, the pulse trains were going dead at the same point ). I then used the schematic to work backwards to find the counters that were at fault. I think Chase HQ has a layer of indirection (?) with the sprites, so that rather than a graphics IC directly manipulating the address pins of the graphics ROMs, instead it calls 'spritemap' ROMs which contain the addresses of the sprites needed to draw a larger object (like a car or a building), and these ROMs then go on to toggle the sprite addresses.
github.com
I don't think you're going to often see a problem in a propagation delay, it's usually weak or dead outputs.
I think 20MHz is a start, but 100MHz is definitely better as you easily get signals into the 10MHz or above range in 80's stuff, which ends up looking like a sine-wave on low-bandwidth equipment (although to be fair, if you ever see a 50MHz waveform on a 1985 game PCB, it probably DOES look like a sine-wave).
I was lucky with the bad graphics on the Chase HQ board I looked at - it was intermittent, so I could tie-in the disruption on the screen with visible changes on the oscilloscope (as sprites went off, the pulse trains were going dead at the same point ). I then used the schematic to work backwards to find the counters that were at fault. I think Chase HQ has a layer of indirection (?) with the sprites, so that rather than a graphics IC directly manipulating the address pins of the graphics ROMs, instead it calls 'spritemap' ROMs which contain the addresses of the sprites needed to draw a larger object (like a car or a building), and these ROMs then go on to toggle the sprite addresses.
mame/src/mame/taito/taito_z.cpp at master · mamedev/mame
MAME. Contribute to mamedev/mame development by creating an account on GitHub.
github.com
I see Rigol has quite a few scopes now labeled MSO/DHO/DSO. Any value in having a scope that can do analog and digital on the same probe (i.e. Rigol MSO5074). Trying to discern the advantages of one type over the other (Rigol MSO/DHO/DSO). Thinking the goal would be pcb work up to 50Mhz.
Hey thanks again for above mate, tbh Ive been playing about the last few days with this little hantek and mate Ive got aways to go, I really dont understand what I'm looking out!. I've gone through this board upside down with the logic probe and everything is pulsing as supposed to I started piggy backing ics directly over the top and may have a found few suspects chips as I do see slight improvement's on the dispaly which is not ideal as I'm obviously wanting to comprehend why this IC may be failing. But again without knowing what I'm looking at on the scope in this case these guys are 194's makes it hard to understand what is going on with input V output. Mate I think I'm going to have to do some more homework and with regard to signal analysis, I struggled to get a single shot still image when in auto or normal mode (I was just probing input output pins) and using both probes to compare I/O signals but again way too hard to comprehend when there pulsing flat out. So other than to test voltages at VCC, I m going to need to ask more questions and do some further reading on other forms of fault finding bugs.You're really into manually doing what things like SLICE and some logic analysers do - taking a chip, looking at the datasheet and interpreting the state tables in there, then looking to see if the chip behaves in the same manner - so every IC different you look at will require you to think of a way to test it.
I don't think you're going to often see a problem in a propagation delay, it's usually weak or dead outputs.
I think 20MHz is a start, but 100MHz is definitely better as you easily get signals into the 10MHz or above range in 80's stuff, which ends up looking like a sine-wave on low-bandwidth equipment (although to be fair, if you ever see a 50MHz waveform on a 1985 game PCB, it probably DOES look like a sine-wave).
I was lucky with the bad graphics on the Chase HQ board I looked at - it was intermittent, so I could tie-in the disruption on the screen with visible changes on the oscilloscope (as sprites went off, the pulse trains were going dead at the same point ). I then used the schematic to work backwards to find the counters that were at fault. I think Chase HQ has a layer of indirection (?) with the sprites, so that rather than a graphics IC directly manipulating the address pins of the graphics ROMs, instead it calls 'spritemap' ROMs which contain the addresses of the sprites needed to draw a larger object (like a car or a building), and these ROMs then go on to toggle the sprite addresses.
mame/src/mame/taito/taito_z.cpp at master · mamedev/mame
MAME. Contribute to mamedev/mame development by creating an account on GitHub.
Im thinking maybe if I get a logic pulser and pulse high - reading the voltage output with the Scope i might see an anomaly . (outloud thinking)
Thanks for chiming in mate appreciated.
Last edited:
Hey mate thanks for the advice and chiming in, the point of my original post was to get an understanding of how you would actually compare signals when you have failing IC's and how to set up a scope in order to test these IC's and most importantly what you need to look for that will let you know if it is failing. ( In a beginner type answer as I'm still pretty green). As one newbie to another when you probe in circuit for example a 377 or 166 what to you do with the scope/s. I have a little 20mhz hantek that was gifted to me so I can play about with the settings but can't get a trigger auto or normal image locked in obviously I'm doing something wrong!. TBH I was convinced about a digital scope however looking up on old repair blogs a lot of blokes seem to just go off the signal not being square on an analog scope as their means to identify suspect chips. Seems a little simpler to me for a newbie. Obviously the focus is 80\s 90\s arcade boards so I don't want to get into high end type gear just a good solid understanding on an old crt scope might be the way to go I'm thinking now. What have you found between both scopes in terms of analyzing signals - differences etc and what type of horizontal frequency do you tend to sit on when probing ttl stuff mate?
Cheers
Y
DS1000z series is obsolete,
the one to get now is the DHO804, 4 channels, hackable upto 250MHz!!! and it has touchscreen and HDMI output!
or for 200 extra you can get the DHO914 and plug in a 16 channel logic analyser adapter - but thats a lot of money IMO just for 16 channels
btw, for clarity:
i have a DS1054z hacked to 100MHz,
and a rackmount 50MHz crt scope with full X,Y & Z inputs - if you dont know why then you dont need to know!
and a kikusui tripple trace 50MHz crt scope,
and a 100MHz single channel pocket scope - great upgrade from a logic probe.
next scope will be a DHO series.
I LIKE SCOPES - does it show?
the one to get now is the DHO804, 4 channels, hackable upto 250MHz!!! and it has touchscreen and HDMI output!
Hacking the Rigol DHO800/900 Scope - Page 1
Hacking the Rigol DHO800/900 Scope - Page 1
www.eevblog.com
or for 200 extra you can get the DHO914 and plug in a 16 channel logic analyser adapter - but thats a lot of money IMO just for 16 channels
btw, for clarity:
i have a DS1054z hacked to 100MHz,
and a rackmount 50MHz crt scope with full X,Y & Z inputs - if you dont know why then you dont need to know!
and a kikusui tripple trace 50MHz crt scope,
and a 100MHz single channel pocket scope - great upgrade from a logic probe.
next scope will be a DHO series.
I LIKE SCOPES - does it show?
The Rigol DHO914s and 924s look interesting over the MSO line (12bits instead of MSO 8bit). I'm guessing that the DHO914/24s line is similar to the MSO line but higher bitrate (12 vs 8) and bandwidth as they also support logic analyzer with decoding. Not sure if the "s" features are what you get after the hack, or if the "s" models have different hardware, those eeblog hack threads are massive and I don't want to read through 60+ page threads
Leaning toward the DHO914 with hack if you gain the bandwidth and decoding of the DHO924s.
Leaning toward the DHO914 with hack if you gain the bandwidth and decoding of the DHO924s.