Digital storage oscilloscope

Max N · Post by **Max N** » Fri Dec 30, 2011 1:10 pm

Andrew linked to this 'scope on another thread:

http://www.rapidonline.com/Tools-Equipm ... pes-400399

I'm tempted but I thought I would ask if anyone else had any recommendations.
I got the MJ 'building valve amplifiers' book for Xmas, so I'll have a read of that and maybe try and summarise what he says. I think it will be a useful thread for beginners who are looking to get a 'scope.

Post by **Nick** » Fri Dec 30, 2011 1:27 pm

I would echo my post on the other thread.

IslandPink · Post by **IslandPink** » Fri Dec 30, 2011 6:59 pm

I can't remember what was said on the other thread, but I do remember that I bought a nice-looking 'UNI-T' Multimeter from Maplins for about £60, and it stopped working within a couple of years, I think the switching contacts wore out , and maybe there were other problems too .

If anyone has one and used it for a reasonable amount of time, then that would be helpful back-up on this. I might be tempted myself ( soon ) to buy a better-quality make , used, from one of the 2nd hand dealers .

Paul Barker · Post by **Paul Barker** » Fri Dec 30, 2011 7:06 pm

I have three seriously abused Fluke mulitmeters over many years and there is non of that nonsense happening. All three work like new and look like a dogs dinner.

conversely my £100 all singing dancing multimeter from Maplin went poof quietly one day when I was on the phone to one of our number and absent mindedly exposed it to a voltage a long way past it's rating. No noise no smoke no fuss. just stone cold dead. And no it isn't the fuse. It is a dead parrott.

The Flukes do nothing clever, but they do the simple multimeter tasks with impunity in all weathers and conditions repeatedly without the slightest variation.

Especially you notice when the newby starts giving bizaar figures you know right away he isn't using quality.

Andrew · Post by **Andrew** » Fri Dec 30, 2011 9:26 pm

They are a bargain in my opinion, remember the scope is only 8 bit or so however, which is very fine for scoping but the FFT will be a low resolution; use 6dB per bit as a guide. Even the most expensive Agilent scopes don't have huge bit depths, its all about speed, basically so that you can get to >10GigHz waveforms etc without violating Nyquist. The bit depth just divides the screen top to bottom, so often, on the cheaper scopes 8 bit is deemed sufficient, sometimes you get 10 or even up to 14.

The trick of increasing the resolution is all explained in the back of Morgan Jones's second book, take one of these 'scopes and a old HP 333 or similar, like the one I sold to Neal, and use that as a front end to the scope, should give you you a very respectable FFT with reading in the high 90 dB's; once noise etc is taken into account.

The main reason I like these scopes however is the magic, find me the waveform button, even a numpty like me can use one...

Andrew

Max N · Post by **Max N** » Fri Dec 30, 2011 10:23 pm

Andrew wrote:The trick of increasing the resolution is all explained in the back of Morgan Jones's second book, take one of these 'scopes and a old HP 333 or similar, like the one I sold to Neal, and use that as a front end to the scope, should give you you a very respectable FFT with reading in the high 90 dB's; once noise etc is taken into account.
Andrew

Do you mean 'Valve Amplifiers' 2nd edition, or 'Building Valve Amplifiers'?
I'm looking through the latter and I can't find the section

Post by **Nick** » Fri Dec 30, 2011 10:34 pm

They are a bargain in my opinion, remember the scope is only 8 bit or so however, which is very fine for scoping but the FFT will be a low resolution; use 6dB per bit as a guide. Even the most expensive Agilent scopes don't have huge bit depths, its all about speed, basically so that you can get to >10GigHz waveforms etc without violating Nyquist. The bit depth just divides the screen top to bottom, so often, on the cheaper scopes 8 bit is deemed sufficient, sometimes you get 10 or even up to 14.

You can average though if the signal is repetitive to give extra bit depth.

Max N · Post by **Max N** » Fri Dec 30, 2011 10:53 pm

'Building Valve Amplifiers' was published in 2004, so lets assume MJ was writing in 2003. So some of his information may be out of date. Nevertheless it may be useful. At that time the scope that he decided to buy was a TDS3032. This featured adjustable persistence or 'Digital Phosphor' which he said gave the best of the old analogue CRT and the new digital LCD scopes. The specs of this scope were (or are, they are still available and are >£5K as far as I can see)

Bandwidth 300MHz
Channels 2
Sample rate on each channel 2.5GS/s
Maximum record length 10K points
Vertical resolution 9-bits
Vertical sensitivity 1mV-10V
Vertical accuracy 2%
Max input voltage (1MOhm) 150V RMS
Position range +/- 5 div
BW limit 20,150MHz
Input coupling AC, DC, GND
Input impedance selections 1MOhm in parallel with 13pF, or 50Ohm
Timebase 2ns - 10s/div
Timebase accuracy 200 ppm
Display Color LCD 640 x 480

Max N · Post by **Max N** » Fri Dec 30, 2011 11:22 pm

Specs for the UTD2025C from Rapid:

Bandwidth 25MHz
Channels 2
Sample rate on each channel 250MS/s
Maximum record length 2 x 512K
Vertical resolution not stated but Andrew thinks 8-bit
Vertical sensitivity 2mV-10V
Vertical accuracy not stated
Max input voltage 400Vp
Position range not stated
BW limit not stated
Input coupling not stated
Input impedance selections not stated
Timebase 20ns - 50s/div
Timebase accuracy not stated
Display Color LCD 320 x 240

So not quite up to the TDS3032 in some areas but does tick a lot of boxes for the money

Max N · Post by **Max N** » Fri Dec 30, 2011 11:37 pm

Andrew wrote:The trick of increasing the resolution is all explained in the back of Morgan Jones's second book, take one of these 'scopes and a old HP 333 or similar, like the one I sold to Neal, and use that as a front end to the scope, should give you you a very respectable FFT with reading in the high 90 dB's; once noise etc is taken into account.
Andrew

Do you mean using the notch filter in the distortion analyzer to remove the fundamental, then passing the residual (distortion) to the 'scope?
As long as one of the fundamentals is a good bit above the rest, you can assume that the THD reading corresponds to that fundamental, then scale the others relative to that one.
I have a distortion meter that has a scope output for the residual, so I could do this

Only works at 400Hz or 1kHz, but that'll do me.

Paul Barker · Post by **Paul Barker** » Sat Dec 31, 2011 8:46 am

Max N wrote: At that time the scope that he decided to buy was a TDS3032.

He put that scope on my 212 amp and showed me that my AVO voltage meter was 150v inacurate. I was cooking all my valves. If you go back to my posts of the era preceding the meeting with Morgan at Chris's flat you will see how many times I mention that the valves are way off the page of the loadlines. Woops!

It boiled down to the fact that when working with 1600v one better get ones equipment callibrated or get out of the kitchen.

Andrew · Post by **Andrew** » Sat Dec 31, 2011 9:38 am

Do you mean using the notch filter in the distortion analyzer to remove the fundamental, then passing the residual (distortion) to the 'scope?
As long as one of the fundamentals is a good bit above the rest, you can assume that the THD reading corresponds to that fundamental, then scale the others relative to that one.

Yes, exactly that, your meter or an old HP 333A/334A or equivalent will do it, all of them have output to connect up a scope so you see what remains after you have notched out the fundamental, just do the scaling from there.

A very cost effective way to get a decent FFT.

Andrew

Andrew · Post by **Andrew** » Sat Dec 31, 2011 9:41 am

PS I'm only guessing its 8-bit, might be more, I only suspect 8-bit as its a common off the shelf value for a digital system, but you never know with an ADC.

I wouldn't worry about it, I mean how many divisions do you need on that small screen? An 8-bit ADC gives max top to bottom of 256 values, 10 square grid @ 1 v per grid means ~40mV per bit, but the screen is only ~100mm high so that means 0.4mm per bit. Could your eye resolve less than half a millimetre difference?

The only big deal with that scope is the nominal 25Mhz but if you have a older analogue scope you may not need to pay for the extra bandwidth.

Having said that, my old HP scope, an early digital, is spec'd at 100Mhz but is actually has a much much slower ADC and uses sampling tricks to get the bandwidth up to what the front end can handle. I haven't noticed the lack of single shot bandwidth as a problem for analogue work since most if the waveforms repeat.

Andrew

Post by **Nick** » Sat Dec 31, 2011 10:46 am

Not that I mind what scope you get, but you are comparing one scope that the cost of a pair of probes for is more than the cost of the entire other scope (which includes probes, not a great pair, but very usable).

Its max input voltage is 400Vp BTW.

ed · Post by ed » Sat Dec 31, 2011 12:58 pm

Hi Max

I bought a utd2062c a while back....I havn't really had a chance to thrash it yet but what I have seen is rock solid. In comparison to my old farnell 40 mhz it's like having the hand that was tied behind my back being released.......

100% recommendation so far....

I was going to get the 2025 that Nick recommmended but the diff in the price between that and the 60mhz was £10...so I thought if I ever want to sell it on then the higher spec would surely be easier....

I bought from the guy that has the distribution contract for uni-t, check him out he may still have deals

diet: labtronix in loughborough is the distributor/dealer

audio-talk

Digital storage oscilloscope

#1 Digital storage oscilloscope

#2

#3

#4

#5

#6

#7

#8

#9

#10

#11

#12

#13

#14

#15